Anemia

BASIC AND CLINICAL ONCOLOGY

Editor

Bruce D. Cheson, M.D.National Cancer Institute

National Institutes of Health

Bethesda, Maryland

1. Chronic Lymphocytic Leukemia: Scientific Advances and Clinical Devel-opments, edited by Bruce D. Cheson

2. Therapeutic Applications of Interleukin-2, edited by Michael B. Atkinsand James W. Mier

3. Cancer of the Prostate, edited by Sakti Das and E. David Crawford4. Retinoids in Oncology, edited by Waun Ki Hong and Reuben Lotan5. Filgrastim (r-metHuG-CSF) in Clinical Practice, edited by George

Morstyn and T. Michael Dexter6. Cancer Prevention and Control, edited by Peter Greenwald, Barnett S.

Kramer, and Douglas L. Weed7. Handbook of Supportive Care in Cancer, edited by Jean Klastersky,

Stephen C. Schimpff, and Hans-Jörg Senn8. Paclitaxel in Cancer Treatment, edited by William P. McGuire and Eric

K. Rowinsky9. Principles of Antineoplastic Drug Development and Pharmacology,

edited by Richard L. Schilsky, Gérard A. Milano, and Mark J. Ratain10. Gene Therapy in Cancer, edited by Malcolm K. Brenner and Robert C.

Moen11. Expert Consultations in Gynecological Cancers, edited by Maurie

Markman and Jerome L. Belinson12. Nucleoside Analogs in Cancer Therapy, edited by Bruce D. Cheson,

Michael J. Keating, and William Plunkett13. Drug Resistance in Oncology, edited by Samuel D. Bernal14. Medical Management of Hematological Malignant Diseases, edited

by Emil J Freireich and Hagop M. Kantarjian15. Monoclonal Antibody-Based Therapy of Cancer, edited by Michael L.

Grossbard16. Medical Management of Chronic Myelogenous Leukemia, edited by

Moshe Talpaz and Hagop M. Kantarjian

17. Expert Consultations in Breast Cancer: Critical Pathways and ClinicalDecision Making, edited by William N. Hait, David A. August, andBruce G. Haffty

18. Cancer Screening: Theory and Practice, edited by Barnett S. Kramer,John K. Gohagan, and Philip C. Prorok

19. Supportive Care in Cancer: A Handbook for Oncologists: Second Edi-tion, Revised and Expanded, edited by Jean Klastersky, Stephen C.Schimpff, and Hans-Jörg Senn

20. Integrated Cancer Management: Surgery, Medical Oncology, andRadiation Oncology, edited by Michael H. Torosian

21. AIDS-Related Cancers and Their Treatment, edited by Ellen G. Feigal,Alexandra M. Levine, and Robert J. Biggar

22. Allogeneic Immunotherapy for Malignant Diseases, edited by JohnBarrett and Yin-Zheng Jiang

23. Cancer in the Elderly, edited by Carrie P. Hunter, Karen A. Johnson,and Hyman B. Muss

24. Tumor Angiogenesis and Microcirculation, edited by Emile E. Voestand Patricia A. D’Amore

25. Controversies in Lung Cancer: A Multidisciplinary Approach, editedby Benjamin Movsas, Corey J. Langer, and Melvyn Goldberg

26. Chronic Lymphoid Leukemias: Second Edition, Revised and Expanded,edited by Bruce D. Cheson

27. The Myelodysplastic Syndromes: Pathology and Clinical Management,edited by John M. Bennett

28. Chemotherapy for Gynecological Neoplasms: Current Therapy andNovel Approaches, edited by Roberto Angioli, Pierluigi BenedettiPanici, John J. Kavanagh, Sergio Pecorelli, and Manuel Penalver

29. Infections in Cancer Patients, edited by John N. Greene30. Anemia of Chronic Disease, edited by Günter Weiss, Victor R. Gordeuk,

and Chaim Hershko

ADDITIONAL VOLUMES IN PREPARATION

Published in 2005 byCRC PressTaylor & Francis Group 6000 Broken Sound Parkway NWBoca Raton, FL 33487-2742

© 2005 by Taylor & Francis Group, LLCCRC Press is an imprint of Taylor & Francis Group

No claim to original U.S. Government worksPrinted in the United States of America on acid-free paper10 9 8 7 6 5 4 3 2 1

International Standard Book Number-10: 0-8247-5972-9 (Hardcover) International Standard Book Number-13: 978-0-8247-5972-8 (Hardcover)

This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. A wide variety of references are listed. Reasonable efforts have been made to publish reliable data and information, but the author and the publisher cannot assume responsibility for the validity of all materials or for the consequences of their use.

No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers.

For permission to photocopy or use material electronically from this work, please access www.copyright.com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC) 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe.

Library of Congress Cataloging-in-Publication Data

Catalog record is available from the Library of Congress

Taylor & Francis Group is the Academic Division of T&F Informa plc.

Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com

and the CRC Press Web site at http://www.crcpress.com

3050_Discl Page 1 Wednesday, February 2, 2005 2:02 PM

Contents

Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii

Introduction: Anemia of Chronic Disease:The Enigma and the ChallengeChaim Hershko, Victor R. Gordeuk and Gu€nter Weiss . . xvii

PART I: BASICS

1. Regulation of Iron Metabolism . . . . . . . . . . . . . 1Kostas PantopoulosBiology of Iron . . . . 1Iron’s Toxicity . . . . 3Iron Distribution in the Body . . . . 5Biomedical Aspects of Iron Metabolism . . . . 5Mechanisms for Dietary Iron Absorption and ErythroidIron Utilization . . . . 8

Regulation of Dietary Iron Absorption and Erythroid IronUtilization . . . . 13

Cellular Iron Uptake . . . . 17Intracellular Iron Storage . . . . 22

iii

Posttranscriptional Regulation of Cellular IronMetabolism by the IRE=IRP System . . . . 24

Iron Regulatory Proteins, IRP1,and IRP2 . . . . 28

Other Regulatory Mechanisms . . . . 33Conclusions . . . . 34

2. Erythropoietin and Erythropoiesis . . . . . . . . . 61Eric Metzen and Wolfgang JelkmannIntroduction . . . . 61Chemical Structure of EPO . . . . 62Sites and Control of EPO Production . . . . 64Assay of Circulating Epo and Interpretation of

Results . . . . 68Action of EPO on Hemopoietic Cells . . . . 69Pathophysiological Aspects . . . . 72

3. Sepsis and Systemic Inflammatory ResponseSyndrome . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87Herbert TilgIntroduction . . . . 87Definitions . . . . 88Pathogenesis . . . . 89Epidemiology . . . . 97Categorical Definitions and Patient Risk . . . . 98New Concepts in the Treatment of Sepsis . . . . 99Conclusions . . . . 100

PART II: PATHOPHYSIOLOGY

4. Disturbances of Iron Homeostasis . . . . . . . . . . 105Victoriano Mulero and Jeremy H. BrockNormal Iron Metabolism . . . . 105The Imbalance of Iron Homeostasis During

Inflammation . . . . 111Disturbance of Iron Trafficking in Macrophages Infected

with Intracellular Pathogens . . . . 116

iv Contents

5. Inhibition of Erythropoiesis by InflammatoryCytokines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127Robert T. Means, Jr.Introduction . . . . 127Inhibition of Erythroid Progenitors by Specific

Cytokines . . . . 130Effects of Erythropoietin and Other Colony Stimulating

Factors on Cytokine Inhibition ofErythropoiesis . . . . 135

Implications for Therapy . . . . 136

6. Endogenous Erythropoietin in the Anemia ofChronic Disorders . . . . . . . . . . . . . . . . . . . . . . . 145Yves BeguinIntroduction . . . . 145Effects of Cytokines on Erythropoietin

Production . . . . 148Interpretation of Serum EPO Levels . . . . 150Serum Erythropoietin in Anemia of Chronic

Disorders . . . . 156Serum Erythropoietin in Cancer . . . . 161Serum Erythropoietin and Chemotherapy . . . . 167Serum Erythropoietin as Predictor of Response to

rHuEPO . . . . 173

7. Erythrophagocytosis and Decreased ErythrocyteSurvival . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201J. J. M. MarxIntroduction . . . . 201The Physiology of Erythrocyte Aging . . . . 202Methods for Estimation of Erythrocyte Life

Span . . . . 204Causes of Decreased Erythrocyte Survival and Life

Span . . . . 205Effect of Inflammation on Erythrocyte

Survival . . . . 206Uptake of Effete Erythrocytes by the Macrophage

System . . . . 207

Contents v

Erythrocyte Destruction and Release of Iron fromHemoglobin and Macrophages . . . . 209

Major Proteins Involved in Iron Release from Hemoglobinand Macrophages . . . . 213

Iron Release from Macrophages inInflammation . . . . 219

8. New Regulator Molecules in Anemia of ChronicDiseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229Olivier Loreal and Pierre BrissotIntroduction . . . . 229Hepcidin: A Key Molecule in Iron Metabolism . . . . 230Hepcidin: A Key Molecule in Anemia of Chronic

Diseases . . . . 233Hepcidin Expression Regulators . . . . 234Molecular Function of Hepcidin . . . . 240Conclusions . . . . 244

PART III: ACD RATIONALE: IS ANEMIA PART OFTHE BODY’S PHYSIOLOGICIAL RESPONSE TOINFLAMMATION?

9. Iron Withholding as a Defense Strategy . . . . . 255Eugene D. WeinbergIntroduction . . . . 255Historical Development of the Concept of Iron

Withholding Defense . . . . 256Components of Iron Withholding Defense . . . . 260Invader Factors Associated with Iron Withholding

Defense . . . . 265Host Factors Associated with Impaired Iron Withholding

Defense . . . . 270Ecological Aspects of Iron Withholding

Defense . . . . 274

10. Iron, Iron Genes, and the Immune System . . . 281C. S. Cardoso, G. Weiss and M. De SousaIntroduction . . . . 281Iron Homeostasis: A Brief Summary . . . . 282

vi Contents

Interplay Between Iron Metabolism and CytokineActivities . . . . 307

Conclusion . . . . 312

PART IV: DIAGNOSIS OF ACD

11. Clinical Approach to the Patient with Anemia ofChronic Disease . . . . . . . . . . . . . . . . . . . . . . . . . 335Victor R. GordeukIntroduction . . . . 335Chronic Inflammatory Process . . . . 336Hypoproliferative Anemia of Normocytic or Microcytic

Morphology . . . . 338Absence of Other Causes of a Hypoproliferative

Anemia . . . . 339Changes in Iron Metabolism . . . . 340Diagnosis of Iron Deficiency in the Setting of

Inflammation . . . . 341Hepcidin . . . . 344Conclusion . . . . 344Appendix . . . . 345

12. Usefulness of Old and New Diagnostic Tests inACD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349Kari Punnonen and Allan RajamakiDistinguishing the Anemia of Chronic Disease from Other

Forms of Anemia . . . . 349Red Blood Cell Morphology and Traditional Classification

of Anemias on the Basis of Mean CorpuscularVolume . . . . 351

The Traditional Markers of Iron Status IncludingFerritin, Transferrin, and Serum Iron in Diagnosis ofACD . . . . 352

Soluble Transferrin Receptor and TfR-F Index in theDifferential Diagnosis of IDA and ACD . . . . 355

Analysis of Hemoglobin Synthesis and Red BloodCells . . . . 358

Novel Markers of Iron Turnover . . . . 359Summary . . . . 360

Contents vii

PART V: THERAPY

13. Treatment of ACD: An Introduction . . . . . . . . 365Gu€nter Weiss

14. Human Recombinant Erythropoietin . . . . . . . 367Hanspreet Kaur, Alan Lichtin, and Deepjot SinghIntroduction . . . . 367Mechanism of Action . . . . 368Serum Erythropoietin Levels as Guidelines forTherapy . . . . 369

Erythropoietin Therapy for HIV Infection=HIVTreatment Related Anemia . . . . 370

Erythropoietin Therapy of Anemia inPatients with Rheumatoid Arthritis . . . . 372

Treatment of Anemia in InflammatoryBowel Disease . . . . 373

Erythropoietin in Patients with Malignancy orChemotherapy . . . . 373

Side Effects . . . . 374Summary . . . . 376

15. Iron Therapy and the Anemia of ChronicDisease . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 381Victor R. GordeukIntroduction . . . . 381Ineffectiveness and Potential Harm of Routine IronTherapy in the Anemia of Chronic Disease . . . . 382

Iron Deficiency . . . . 383Iron Therapy for Patients with Combined Iron Deficiencyand the Anemia of Chronic Disease . . . . 388

Iron Therapy in Anemia of Chronic Disease PatientsReceiving Erythropoietin . . . . 393

16. Blood Transfusions . . . . . . . . . . . . . . . . . . . . . . 397Eleftherios C. VamvakasRBC Transfusion Therapy in Chronic Anemia . . . . 399Established Noninfectious Risks of ABT . . . . 402

viii Contents

Infectious Risks of ABT . . . . 407Purported Noninfectious Risks of ABT . . . . 417Summary . . . . 426

17. Iron and Erythropoietin . . . . . . . . . . . . . . . . . . 437Lawrence T. GoodnoughIntroduction . . . . 437Prevalence of Anemia of Chronic Disease . . . . 438Anemia and Adverse Outcomes . . . . 438Management of Anemia . . . . 443Erythropoietin, Iron, and Erythropoiesis . . . . 443Erythropoietin Response to Anemia . . . . 445Iron-Restricted Erythropoiesis . . . . 447Laboratory Evaluation of Iron Metabolism . . . . 454Iron Therapy Strategies . . . . 460Current Issues in Erythropoietin Therapy . . . . 461Conclusion . . . . 466

PART VI: CONTROVERSIES IN ACD THERAPY

18. Positive Effects of Correction of Anemia inMalignant Diseases . . . . . . . . . . . . . . . . . . . . . . 489Gudrun Pohl and Heinz LudwigIntroduction . . . . 489Indications for Anemia Treatment . . . . 498Treatment Options . . . . 499Positive Effects of Correction of ACD . . . . 509Conclusion . . . . 535

19. Putative Negative Effects of the Correction ofAnemia in ACD . . . . . . . . . . . . . . . . . . . . . . . . . 559Gu€nter WeissIntroduction: ACD as a Defense Strategy of the

Body . . . . 559Severe ACD Is Associated with a Poor Prognosis: Does

this Imply that Anemia Correction May BeBeneficial? . . . . 563

Contents ix

Potential Hazards of Specific TherapeuticRegimen . . . . 566

General Consideration Concerning Anemia Correction inACD . . . . 576

PART VII: SPECIFIC CONDITIONS OF ACD

20. Anemia of Chronic Disease in HematologicDisorders and Oncology . . . . . . . . . . . . . . . . . . 593Robert T. Means and Gordon D. McLarenFrequency of Anemia of Chronic Disease in Hematologic

and Oncologic Disorders . . . . 593Pathogenesis of ACD in Hematologic and Oncologic

Disorders . . . . 595Diagnosis of ACD in Cancer Patients . . . . 597Treatment . . . . 598

21. Anemia in Cancer Patients UndergoingSurgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607D.OfnerIntroduction . . . . 607Patients and Methods . . . . 608Results and Discussion . . . . 608

22. Iron Status, Anemia of Chronic Disease, andInfection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615Johan R. BoelaertTuberculosis . . . . 616HIV Infection . . . . 620HIV=Tuberculosis Coinfection . . . . 623Hepcidin as Key Mediator of Infection-Related Anemia of

Chronic Disease . . . . 624

23. ACD in Inflammatory Rheumatic Diseases . . . 633J. P. Kaltwasser and U. ArndtIntroduction and Pathogenesis . . . . 633Incidence of Anemia . . . . 636

x Contents

Differential Diagnosis of Anemia . . . . 636Laboratory Diagnosis . . . . 637Clinical Impact of ACD in RheumaticDisorders . . . . 640

Treatment of ACD . . . . 641Summary and Possible Therapeutic FutureOptions . . . . 647

24. Anemia in Intensive Care Patients . . . . . . . . . 659Albert van de WielIntroduction . . . . 659Mechanisms . . . . 660Therapy . . . . 664Summary . . . . 665

25. Anemia in Renal Disease . . . . . . . . . . . . . . . . . 671Jorge Luis Ajuria, Paul L. Kimmel, and Robert S. SiegelIntroduction . . . . 671Epidemiology of Renal Disease in the UnitedStates . . . . 673

Anemia and Chronic Renal Disease . . . . 674Pathophysiology . . . . 675Erythropoietin and the Kidney . . . . 675Erythropoietin Production in Renal Disease . . . . 677Erythropoietin Production in Acute RenalFailure . . . . 678

Uremia and Erythropoiesis . . . . 679Erythropoiesis in Uremia . . . . 680Erythropoiesis and Inflammation . . . . 681Renal Failure and Myelofibrosis . . . . 682Shortened Erythrocyte Life Span . . . . 683Membrane Alterations in UremicErythrocytes . . . . 685

Abnormal Hemostasis . . . . 688Trends in rHuEPO Therapy . . . . 690Management of Anemia . . . . 691Clinical Sequelae of Anemia in Chronic RenalDisease . . . . 694

Contents xi

Clinical Benefits of rHuEPO Therapy . . . . 695Conclusion . . . . 697

26. Anemia of Chronic Disease in Inflammatory BowelDiseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 727Christoph GascheImportance of ACD in IBD . . . . 728Specific Pathogenesis of Anemia in IBD . . . . 729Current Therapy . . . . 731Unanswered Issues . . . . 734

Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .745

xii Contents

Contributors

Jorge Luis Ajuria Department of Medicine, Division of RenalDiseases and Hypertension, George Washington University MedicalCenter, Washington D.C., U.S.A.

U. Arndt Abteilung Rheumatologie, Medizinische Klinik II,Zentrum der Inneren Medizin, der J.W. Goethe—Universitat,Frankfurt, Germany

Yves Beguin National Fund for Scientific Research, Division ofHematology, Department of Medicine, University of Liege; andLaboratory of Cell and Gene Therapy CHU Sart-Tilman; andCenter for Cellular and Molecular Therapy University of Liege,Liege, Belgium

Johan R. Boelaert Unit of Renal and Infectious Diseases,Algemeen Ziekenhuis St-Jan, Brugge, Belgium

Pierre Brissot INSERM U-522 and Service des Maladies duFoie, University Hospital Pontchaillou, Rennes, France

xiii

Jeremy H. Brock Department of Immunology, WesternInfirmary, University of Glasgow, Glasgow, U.K.

C.S. Cardoso Molecular Immunology, Institute for Molecularand Cell Biology, Oporto, Portugal

M. De Sousa Molecular Immunology, Institute for Molecular andCell Biology, Oporto, Portugal

Christoph Gasche Department of Medicine, Division ofGastroenterology and Hepatology, Medical University of Viennaand General Hospital Vienna, Vienna, Austria

Lawrence T. Goodnough Departments of Pathology andMedicine, Stanfold University, Palo Alto, California, U.S.A.

Victor R. Gordeuk Department of Medicine, Center for SickleCell Disease, Howard University, Washington D.C., U.S.A.

Chaim Hershko Hebrew University, Hadassah Medical School,Jerusalem and Ben Gurion University, Faculty of Medicine, BeeBeer Sheva, Israel

Wolfgang Jelkmann Institute of Physiology, University ofLuebeck, Ratzeburger Allee, Luebeck, Germany

J.P. Kaltwasser Abteilung Rheumatologie, Medizinische KlinikII, Zentrum der Inneren Medizin, der J.W. Goethe—Universitat,Frankfurt, Germany

Hanspreet Kaur Department of Hematology=Oncology,Cleveland Clinic Foundation, Cleveland, Ohio, U.S.A.

Paul L. Kimmel Department of Medicine, Division of RenalDisease and Hypertension, George Washington University MedicalCenter, Washington D.C., U.S.A.

Alan Lichtin Department of Hematology=Oncology, ClevelandClinic Foundation, Cleveland, Ohio, U.S.A.

Olivier Loreal INSERM U-522 and Service des Maladies duFoie, University Hospital Pontchaillou, Rennes, France

xiv Contributors

Heinz Ludwig Department of Medicine I and Medical Oncology,Wilhelminenspital, Vienna, Austria

J.J.M.Marx Eijkman–Winkler Centre for Microbiology, InfectiousDiseases and Inflammation, University Medical Centre Utrecht,Utrecht, The Netherlands

Gordon D. McLaren Hematology Oncology Division,Department of Medicine, University of California, Irvine and VALong Beach Healthcare System, Long Beach, California, Ohio,U.S.A.

Robert T. Means, Jr. Hematology=Oncology Division,Department of Medicine, Ralph H. Johnson VA Medical Center, andthe Medical University of South Carolina, Charleston, SouthCarolina, U.S.A.

Eric Metzen Institute of Physiology, University of Luebeck,Ratzeburger Allee, Luebeck, Germany

Victoriano Mulero Department of Cell Biology, Faculty ofBiology, University of Murcia, Murcia, Spain

D. Ofner Department of Surgery, Division of General andTransplant Surgery, Innsbruck University Medical School,Innsbruck, Austria

Kostas Pantopoulos Lady Davis Institute for Medical Research,Sir Mortimer B. Davis Jewish General Hospital and Department ofMedicine, McGill University, Montreal, Quebec, Canada

Gudrun Pohl Department of Medicine I and Medical Oncology,Wilhelminenspital, Vienna, Austria

Kari Punnonen Department of Clinical Chemistry andLaboratory Hematology, Kuopio University Hospital, Kuopio,Finland

Allan Rajamaki Department of Clinical Chemistry andLaboratory Hematology, Turku University Hospital, Turku,Finland

Contributors xv

Herbert Tilg Department of Medicine, Division ofGastroenterology and Hepatology, University Hospital Innsbruck,Innsbruck, Austria

Robert S. Siegel Department of Medicine, Division of Hematologyand Oncology, George Washington University Medical Center,Washington D.C., U.S.A.

Deepjot Singh Department of Hematology=Oncology, UniversityHospitals of Cleveland, Cleveland, Ohio, U.S.A.

Eleftherios C. Vamvakas Division of Medical, Scientific andResearch Affairs, Canadian Blood Services; and Department ofPathology and Laboratory Medicine, University of Ottawa Facultyof Medicine, Ottawa, Canada

Albert van de Wiel Department of Internal Medicine, MeanderMedical Center, Amersfoort, The Netherlands

Eugene D. Weinberg Department of Biology and Program inMedical Sciences, Indiana University, Bloomington, Indiana, U.S.A.

Gunter Weiss Department of General Internal Medicine, ClinicalImmunology and Infectious Diseases, Medical University ofInnsbruck, Innsbruck, Austria

xvi Contributors

Introduction

Anemia of Chronic Disease: TheEnigma and the Challenge

CHAIM HERSHKO

Hebrew University, Hadassah MedicalSchool, Jerusalem andBen Gurion University,Bee Beer Sheva, Israel

VICTOR R. GORDEUK

Department of Medicine, Center forSickle Cell Disease,

Washington D.C., U.S.A.

GUNTER WEISS

Department of General InternalMedicine, Clinical Immunology and

Infections Diseases, Medical Universityof Innsbruck, Innsbruck, Austria

The anemia associated with chronic disease (ACD) is charac-terized by abnormal iron distribution, decreased red cell lifespan, and impaired erythropoietin response. The ACD con-tinues to intrigue clinicians and basic scientists alike, eversince thedefinitionof thisentity (1)summarized inseveralmajorreviews (2–4). The characteristic combination of decreased

xvii

serum iron, decreased serum transferring, and normal orincreased serum ferritin distinguishes it from iron deficiencyanemia (IDA). The main features of abnormal iron handlinginvolving impaired reutilization of iron derived from senes-cent nonviable erythrocytes decreased erythrocyte survivaland a relative failure of the marrow to compensate forincreased red cell loss have all been defined as early as1957 in the remarkable pioneering studies of Freireich etal. (5). Apart from IDA, ACD is the second most common ane-mia of mankind and, its prevalence among hospitalizedpatients exceeds even that of IDA. Although initially desig-nated anemia of infection (1), it is clear that the same entitymay be encountered in chronic diseases in which inflamma-tion is caused by noninfections conditions such as rheumatoidarthritis and other connective tissue disorders, malignantdisease, or trauma. The common denominator of these condi-tions is inflammation, mediated by cytokines. It is also clearthat acute injury such as trauma or severe infection or eventyphoid vaccination may result within hours or days in hypo-ferremia indistinguishable from that of ACD, and hence theterm chronic is not necessarily an essential feature of ACD.Thus, anemia of inflammation would be a much more appro-priate definition. However, the term ACD is now widelyaccepted, it is mostly encountered in chronic disorders, andany further preoccupation with semantics may be futile.

Understanding the abnormalities of iron homeostasis inACD is inseparable from understanding the normal mechan-isms of iron handling. Iron is an essential component of pro-teins that play a key role in respiration, energy production,detoxification of harmful oxygen species and cell replication.Despite the abundance of iron in nature, the solubility of itsstable ferric form is extremely low. Hence, living organismswere compelled to develop efficient mechanisms for irontransport and storage.

In recent years, a number of key mechanisms have beendescribed which are responsible for adaptation to changingenvironmental conditions (6). Production of the iron storageprotein ferritin and the transferrin receptor (TfR) protein isreciprocally regulated by a translational mechanism in which

xviii Hershko et al.

the iron regulatory protein (IRP) is reversibly bound to theiron response elements (IRE) of their respective mRNAs. Asimilar iron-dependent translational mechanism may affectthe expression of divalent metal transporter I (DMTI) respon-sible for the uptake of ferrous iron from the brush border ofduodenal enterocytes, and ferroportin (IregI) responsible forthe export of ferrous iron through the basolateral membraneof the same cells. The brush border ferric reductase convertsferric to ferrous iron for use by DMTI, and Hephaestin, atransmembrane-bound ferroxidase, converts ferrous to ferriciron, creating a concentration gradient of errous iron acrossthe cell membrane facilitating iron egress. At low iron condi-tions, the translation of TfR, DMT1, and ferroportin isenhanced, with the opposite occurring at high iron conditions.In addition, a new protein, Hepcidin, has been describedrecently and is probably the most important regulator of ironhomeostasis (7). Hepcidin functions as an inhibitor of ironabsorption and presumably of iron release from macrophages.Its production is increased by iron overload and inflammationand is suppressed by iron deficiency. Thus, in iron deficiencypowerful compensatory mechanisms involving increasedactivity of iron transport proteins and inhibition of Hepcidinare activated in order to restore normal iron balance. How-ever, these mechanisms are only partly effective, and irondeficiency anaemia (IDA) is one of the most common nutri-tional deficiencies in the global population.

The timeliness of the present volume on ACD is under-scored by a number of recent developments. The discoveryof Hepcidin and its inter-relation with the genes for HFE,hemojuvelin (8), and possibly transferrin receptor-2 revolutio-nized our understanding of the abnormal iron homeostasis ofACD. These recent discoveries offer new insights into theenigma of increased ferritin synthesis in ACD preceding thedevelopment of hypoferremia (9), and the inter-relation ofIRE, IRP NO, and cytokines (10) in the pathogenesis ofACD. The chapters covering the regulation of iron metabo-lism, the systemic inflammatory response, and the newlydescribed regulatory molecules of iron metabolism will pro-vide a comprehensive insight into the molecular mechanisms

Anemia of Chronic Disease xix

involved in ACD. The implications of the unique combinationof hypoferremia and cellular siderosis in inflammation will bediscussed in the chapters covering iron and immunity, ironwithholding as a defense strategy, and the two chapters onthe positive and negative effects on infectious and malignantdisease of ACD and of its correction. Failure of the erythro-poietin response is central to the development of anemia inACD and the introduction of recombinant erythropietin tothe management of ACD has been the most important recentdevelopment in the treatment of its anemia. The reader willfind a wealth of information on these aspects in the chapterson erythropoietin and erythropoiesis, the inhibition of ery-throid progenitor cell proliferation, human recombinant ery-thropoietin, and the inter-relation of iron and erythropoietinadministration in ACD. However, the effects of anemia cor-rection on the clinical course of the underlying disease bythe different therapeutic measures are not known so far,and major attempts should be undertaken to clarify this mostimportant issue depending on the nature of the underlyingdisease.

Further understanding of the mechanism of anemia willbe offered in the chapters on iron-limited erythropoiesis, ery-throphagocytosis, and decreased red cell survival. Finally, onthe practical side, special chapters will cover the issues ofdiagnostic tests in ACD, the use of iron, blood transfusions,and new therapeutic options in the treatment of ACD. Forthe specialists, particular chapters have been devoted toACD in systemic infection, cancer, rheumatic and autoim-mune disorders, anemia in intensive care patients, chronicrenal disease, inflammatory bowel disease, and the anemiaassociated with transplantation.

It is believed that this volume will satisfy the need for anup-to-date compilation of knowledge in the field of ACD. It isintended to be used by students, clinicians, and investigatorsalike. It should be remembered, however, that much is still tobe learned about this common and enigmatic clinical entity.Although Hepcidin is clearly a central player in the dramaof ACD, it does not appear to interact with iron directly andthe manner in which it is able to influence the function of

xx Hershko et al.

other well-defined proteins of iron transport and storage is atthis stage unknown. Finally, one should always keep in mindthat ACD is a secondary phenomenon and that successfultreatment of the underlying disease responsible for theinflammatory condition is the ultimate goal of treatment.

REFERENCES

1. Cartwright GE, Lauritsen MA, Jones PG, Merrill IM,Wintrobe MM. The anemia of infection. J Clin Invest 1946; 25:65–80.

2. Cartwright GE. The anemia of chronic disorders. SeminHematol 1966; 3:351–368.

3. Lee GR. The anemia of chronic disease. Semin Hematol 1983;20:61–80.

4. Weiss G. Pathogenesis and treatment of anemia of chronic dis-ease. Blood Rev 2002; 16:87–96.

5. Freireich EJ, Miller A, Emerson CP, Ross JF. The effect ofinflammation on the utilization of erythrocyte and transferrinbound radioiron for red cell production. Blood 1957; 12:872–979.

6. Andrews NC. A genetic view of iron homeostasis. SeminHematol 2002; 39:227–234.

7. Ganz T. Hepcidin, a key regulator of iron metabolism and med-iator of anemia of inflammation. Blood 2003; 102:783–788.

8. Papanikolau G, Samuels ME, Ludwig EH, et al. Mutations inHFE2 cause iron overload in chromosome 1q-linked juvenilehemochromatosis. Nature Genet 2003.

9. Konijn AM, Hershko C. Ferritin synthesis in inflammation. I.Pathogenesis of impaired iron release. Brit J Haematol 1977;37:7–16.

10. Weiss G, Werner-Felmayer G, Werner ER, Grunewald K,Wachter H, Hentze MW. Iron regulates nitric oxide synthaseactivity by controlling nuclear transcription. J Exp Med1994; 180:969–976.

Anemia of Chronic Disease xxi

1

Regulation of Iron Metabolism

KOSTAS PANTOPOULOS

Lady Davis Institute for Medical Research,Sir Mortimer B. Davis Jewish

General Hospital andDepartment of Medicine,

McGill University, Montreal, Quebec, Canada

BIOLOGY OF IRON

Iron is an abundant transition metal, characterized by itsredox reactivity to switch between two basal ferrous Fe(II)and ferric Fe(III) states, and by its capacity to form a varietyof co-ordination complexes with organic ligands. Virtually, allliving cells and organisms (with a few minor exceptions)exploit these advantageous chemical properties and utilizeiron, as component of iron-containing metalloproteins, toaccomplish vital structural and metabolic functions. Theseinclude oxygen sensing and transport, and a plethora of elec-tron transfer and catalytic reactions (1). Iron co-ordinates to

PART I: BASICS

1

metalloproteins in a dynamic and flexible mode, either inform of heme or not (2).

In mammals, the vast majority of body iron is utilized inerythroid cells for the synthesis of heme, a very common pros-thetic group composed of protoporphyrin IX and Fe(II), whichgets incorporated into hemoglobin. Heme is also essential com-ponent ofmanycell-specificorubiquitoushemoproteins, suchasmyoglobin, cytochromes, and various enzymes. Thus, nonery-throid cells also synthesize heme, albeit in significantly loweramounts. The heme moiety is essential for oxygen binding andserves as oxygen carrier in hemoglobin and myoglobin and asoxygen activator in oxygenases (for example, in cytochrome oxi-daseandcytochromeP450) orH2O2activator inperoxidases (forexample, in catalase). As a constituent of cytochromes a, b, andc, heme participates in electron transfer reactions duringaerobic respiration. Heme is also involved in the sensing andbiosynthesis of nitric oxide (NO) by the soluble guanylatecyclase and nitric oxide synthases (NOS), respectively.

The tetrapyrrol porphyrin ring of heme is synthesized inall organisms from the universal precursor 5-aminolevulinicacid. In most eukaryotes (except plants), the precursor is gen-erated in the cytosol by the condensation of succinyl-CoA andglycine, which is catalyzed by the 5-aminolevulinic acidsynthase (ALAS) (3). Mammals express two ALAS isoforms, ahouse keeping (ALAS-1) and an erythroid specific (ALAS-2).The terminal step of the heme biosynthetic pathway involvesthe insertion of Fe(II) into protoporphyrin IX, which takesplace in the mitochondria and is catalyzed by ferrochelatase.Heme is then exported to the cytosol for incorporation intohemoproteins. In nonerythroid cells, the rate-limiting step ofthe pathway is the synthesis of 5-aminolevulinic acid. In ery-throid cells, the synthesis of the porphyrin ring is tightly co-ordinated with iron supply, which appears to be rate limiting.Under physiological conditions, iron can only be removed fromheme enzymatically, in a reaction catalyzed by heme oxyge-nases (4). The reaction products are Fe(II), carbon monoxide(CO), which may perform signaling functions, and biliverdin,which is further enzymatically converted to bilirubin, a mole-cule with antioxidant properties.

2 Pantopoulos

Many proteins utilize prosthetic groups of nonheme iron,such as iron–sulfur and iron-oxo clusters, or mononuclear ironcenters. Iron–sulfur clusters are probably the most prevalentforms of nonheme iron and play diverse functional roles,including electron transfer and catalysis (5). Characteristicexamples are the Rieske proteins in complex III of the respira-tory chain, which contain 2Fe–2S clusters and are involved inelectron transfer reactions, and aconitase, which contains a4Fe–4S cluster and catalyzes the isomerization of citrate toiso-citrate in the citric acid cycle. The enzymatic reduction ofribo- to deoxyribonucleotides, a critical step in DNA synthesis,depends on an unusual Fe–O–Fe center within the ribonucleo-tide reductase. Finally, members of the broad family of oxyge-nases, such as cyclooxygenase and lipoxygenase, contain amononuclear iron center to activate substrates.

Iron also directly participates in a mechanism for oxygensensing. Cells of higher eukaryotes respond to reduced oxygenavailability by activation of the hypoxia inducible factor (HIF)that controls the transcription of awide array of genes involvedin erythropoiesis, angiogenesis, cell proliferation=survival,glycolysis, and ironmetabolism. The expression of HIF is regu-lated at the level of protein stability. Under normoxic condi-tions, the oxygen-sensitive subunit HIF-1a is hydroxylated atP402 and P564 within two functionally independent degra-dation domains. This modification provides a recognition sitefor the von Hippel–Lindau (VHL) tumor suppressor protein,a component of anE3 ubiquitin ligase complex. The interactionresults in ubiquitination and degradation of HIF-1a by theproteasome. The hydroxylation of P402 and P564 is carriedout by specific prolyl-4-hydroxylases, which are members ofthe family of 2-oxoglutarate-dependent oxygenases (6). Thereaction has an absolute requirement for iron and dependson oxygen availability. Thus, the prolyl-4-hydroxylases modi-fying HIF-1a essentially function as ‘‘oxygen sensors.’’

IRON’S TOXICITY

When present in excess, iron may turn into a potent bioha-zard. Iron’s toxicity is based on its ability to catalyze the

Regulation of Iron Metabolism 3

generation of hydroxyl radicals (OH) in the presence of super-oxide (O2

�) and hydrogen peroxide (H2O2) (7), according toFenton and Haber=Weiss chemistry (Fig. 1). Iron can also cat-alyze the generation of organic radicals from organicperoxides. The superoxide anion and H2O2, also known as‘‘reactive oxygen intermediates’’ (ROIs), are generated asbyproducts of aerobic respiration in mitochondria, or duringenzymatic reactions in peroxisomes, the endoplasmic reticu-lum or the cytoplasm. Reactive oxygen intermediates are alsoproduced by the membrane-bound NADPH oxidase complex(8), which was first discovered in phagocytic neutrophilsand macrophages, but appears to operate in many cell types.This enzyme complex assembles in response to infection andgenerates high levels of superoxide in a ‘‘respiratory burst’’to kill bacteria. The superoxide is spontaneously and=or enzy-matically dismutated to H2O2. Both give rise to more potentoxidants such as peroxynitrite (ONOO�) and hypochlorite(OCl�), which amplify the bactericidal (and cytotoxic) activityof phagocytic cells. The former is generated by the sponta-neous reaction of superoxide and NO, while the latter issynthesized from hydrogen peroxide and chloride in a reac-tion catalyzed by myeloperoxidase.

Free radicals and other reactive species are considered ashighly aggressive compounds because they attack cellularmacromolecules and promote oxidation of proteins, peroxida-tion of membrane lipids and damage of nucleic acids. A multi-tude of enzymatic and nonenzymatic antioxidant mechanismshave evolved to protect cells against ‘‘oxidative stress’’ (7,9),and a compromise in the antioxidant defence may ultimatelylead to cell death and tissue degeneration (10,11). This is

Figure 1 Iron-catalyzed generation of oxygen radicals.

4 Pantopoulos

encountered during the process of aging (12) and in manypathological conditions, such as inflammation, ischemia–reperfusion injury, diabetes, pulmonary disease, and neurode-generation (13).

IRON DISTRIBUTION IN THE BODY

The human body contains approximately 3–5 g iron (45 and55mg=kg of body weight in adult women and men, respec-tively), fromwhich�60–70% is utilized for heme and hemoglo-bin synthesis in the erythron (Fig. 2) (14,15). Macrophages areinstrumental for the phagocytosis of senescent erythrocytes,the breakdown of heme, and the recycling of iron in the circula-tion for its delivery into expanding erythroblasts. Since thereis no specific mechanism for iron excretion, the recycling ofiron by the reticuloendothelial system is imperative for themaintenance of sufficient iron supply during the course of ery-thropoiesis. Other important organs for iron homeostasis arethe muscle and the liver. The former contains significantamounts of heme iron in myoglobin (�7–8% of body iron) andthe latter provides a storage site to the remaining �20–30%of body iron. Dietary iron is absorbed in the duodenum. Inhealthy individuals, the average daily iron absorption is �1–2mg, which represents �0.05% of total body iron. This minis-cule amount is required to compensate iron losses by cell des-quamation, blood loss, or in sloughed mucosal cells. A dynamicpool of �3mg iron remains constant in plasma, despite a highturnover rate of 30 mg=day.

BIOMEDICAL ASPECTS OF IRON METABOLISM

The tight control of iron metabolism is critical for health,because both iron deficiency as well as iron overload areassociated with disease (15). Depletion of body iron storesdue to insufficient dietary iron supply is linked to impairederythropoiesis and iron deficiency anemia (IDA). Interest-ingly, IDA is the most common nutritional pathologic condi-tion, affecting �20% of world population, and poses a


serious health problem in developing countries (16). On theother hand, defects in iron reutilization lead to the anemiaof chronic disease (ACD), the most frequent anemia amonghospitalized patients in industrial countries. Anemia ofchronic disease develops in chronic inflammatory conditionsassociated with infection, cancer, or autoimmune disorders(17). It is characterized by defective macrophage iron trans-port, which results in hypoferremia and diversion of iron fromcirculation into the reticuloendothelial system. This response

Figure 2 Iron distribution in the adult body. The vast majority(�60–70%) of iron is utilized in hemoglobin within bone marrowcells and circulating erythrocytes and is recycled by reticuloen-dothelial macrophages. A significant portion (�7–8%) of iron isutilized in muscle hemoglobin and the rest (�20–30%) is stored inthe liver parenchyma. Daily absorption of 1–2mg iron from the dietoffsets nonspecific iron losses.

6 Pantopoulos

may be part of the organism’s strategy to deplete pathogensfrom an essential nutrient. The pathophysiology of ACD willbe discussed extensively in Chapter 3.

At the other end of the spectrum, iron’s toxicity is vividlyillustrated in diseases of primary and transfusional iron over-load, which are associated with various complications, such asdiabetes, arthropathy, tissue fibrosis, and degeneration, andultimately lead to liver and heart failure (15). The most pro-minent example is hereditary hemochromatosis (HH), agenetically heterogeneous group of iron overload disorders,where a chronic imbalance in dietary iron absorption resultsin iron accumulation, primarily within tissue parenchymalcells. Several types of HH have been described, which varyin the clinical symptoms and the underlying molecular basis.The most common form (HH type 1) is characterized by adultonset and predominant manifestations of liver disease. Thedisease is transmitted in an autosomal recessive manner.The defect is due to mutations in the HFE gene (18,19), whichencodes an atypical major histocompatibility complex (MHC)class I protein of 343 amino acids (20). Its exact function isstill not well defined, but it appears to play an important rolein the regulation of dietary iron absorption and iron recyclingby macrophages. The most common disease-associated muta-tion (C282Y) abrogates a disulfide bridge and prevents theinteraction of HFE with b2-microglobulin (b2M), therebyimpairing its processing in the Golgi and its expression onthe cell surface. Another HFE point mutation, H63D, hasbeen associated with a moderate risk for HH, especially inC282Y=H63D compound heterozygotes (21). Further uncom-mon HFE mutations and polymorphisms have been linkedto HH, including missense, frameshift, and nonsense muta-tions (19), and a mutation leading to a splicing defect (22).

It should be noted that HH of type 1 is the most commongenetic disorder, affecting approximately 1:400 individuals ofprimarily Northern European descent. The estimated carrierfrequency of the prevalent C282Y mutation is �1:10, whichexceeds the frequency of cystic fibrosis, muscular dystrophy,and phenylketonuria combined (23,24). However, it hasrecently been questioned whether the high prevalence of this


mutation is accompanied by comparatively high penetrance(25,26). Nevertheless, the hemochromatosis phenotype hasbeen recapitulated in hfeC282Y=C282Y (27) mice. More severe tis-sue iron overload has been observed in hfe�=� (28,29) mice, butthe degree of iron loading (30,31) and the expression of genesinvolved in iron absorption (32) differ among mouse strains.Thus, the genetic background appears to be an important fac-tor in HFE-related hemochromatosis, and systematic geneticanalysis of the various mouse models of disease is expectedto lead to the identification of genemodifiers. In support of thisview, the crossing of hfe�=� mice withmice carrying mutationsin other genes of iron metabolism has yielded various degreesof iron loading in the progeny (33). Interestingly, mice lackingexpression of either b2-microglobulin (34–36) or classical MHCclass I molecules (37) also develop iron overload.

Additional disorders of hereditary iron overload includejuvenile hemochromatosis (HH of type 2) (38), HH of types 3,4, and 5 (26,39), neonatal hemochromatosis (40), African ironoverload (41), aceruloplasminemia (42), hypotransferrinemia=atransferrinemia (39), and heme oxygenase 1 deficiency (39).The above disorders are rare or extremely rare, but studieson the underlying molecular mechanisms and the develop-ment of animal models (43) have provided and continue toprovide valuable insights on the regulation of iron metabolismin the body. The most significant recent findings will bediscussed in the ensuing sections.

MECHANISMS FOR DIETARY IRONABSORPTION AND ERYTHROID IRONUTILIZATION

Iron’s bioavailability is generally limited, because Fe(II) isreadily oxidized in solution to Fe(III), which is virtuallyinsoluble at physiological pH (Kfree Fe(III)¼ 10�18M) (1,44).Therefore, unicellular and more complex multicellular organ-isms have developed various sophisticated strategies toacquire nutritional iron (45). Mammals absorb dietary ironin the gastrointestinal tract (46). The molecular basis of this

8 Pantopoulos

important physiological process remained obscure for longtime and only recently (in the last 5–6 years) we began tounderstand some fundamental mechanistic aspects. The turn-ing point was the identification and characterization of mole-cules involved in the absorption of inorganic iron in mice.These include novel iron transporters and oxidoreductases,which are also conserved in humans. It should, however, benoted that we still do not know much on the absorption ofheme, which in humans may represent a quantitatively moresignificant fraction of nutritional iron.

According to the current model, inorganic iron within theintestinal lumen has to be reduced to soluble Fe(II) prior toabsorption (Fig. 3). This step requires the activity of the duo-denal cytochrome b (Dcytb), a 286 amino acids hemoproteinwith six predicted transmembrane domains, which possessesferric reductase activity (47). Ascorbate may also facilitatethe reduction of iron. Ferrous iron is then transported acrossthe intestinal mucosa by the divalent metal transporter 1

Figure 3 A model for dietary iron absorption in the duodenum.Ferric iron is reduced in the intestinal lumen by Dcytb. Ferrous ironis then transported across the apical membrane of mature entero-cytes by DMT1. The mechanism for iron traffic within the enterocyteis completely unknown. The export of iron across the basolateralmembrane of enterocytes to circulation is mediated by ferroportin.This step is coupled with reoxidation of ferrous to ferric iron by mem-brane-bound hephaestin. Plasma iron is immediately scavenged bytransferrin (Tf). The iron status of the precursor cells in the cryptsreflects the iron absorption capacity of mature enterocytes.


(DMT1), also known as DCT1 (48), Nramp2 (49), or SLC11A2(50). This is member of the solute carrier family of proton-coupled divalent metal ion transporters and is composed of561 amino acids with 12 predicted transmembrane domains.Both Dcytb (47) and DMT1 (51) are expressed in the apicalmembrane of enterocytes at the brush border of the duodenalepithelium. The expression of DMT1 is not confined tomature enterocytes. In nonintestinal cells, DMT1 is involvedin intracellular iron transport across the endosomalmembrane.

Following an ill-defined step of intracellular trafficwithin the enterocyte, Fe(II) is exported to the portal circula-tion by the basolateral transporter ferroportin 1 (52), alsoknown as IREG1 (53), MTP1 (54), or SLC11A3 (55). Thisis a proton-coupled divalent metal ion transporter of 571amino acids with nine predicted transmembrane domains.Ferroportin is also expressed in macrophages and in placentalsyncytiotrophoblasts, and it appears to play a crucial role inthe export of iron from macrophages to the circulation(Fig. 4) and in maternal iron transfer to the fetus. Macro-phage iron is primarily derived from phagocytosed senescenterythrocytes and its export into circulation is an importantstep for reutilization by expanding erythroblasts.

The mechanisms for basolateral iron transport in enter-ocytes and for iron efflux from macrophages show strikingsimilarities. In both cases, the ferroportin-mediated transportof Fe(II) across the cell membrane is followed by its reoxida-tion to Fe(III) and binding to transferrin (Tf). This is a mono-meric glycoprotein of 698 amino acids, which serves as theplasma iron carrier and binds two Fe(III) ions with high affi-nity (Kd¼ 10�23M at neutral pH), keeping them soluble andnonreactive (14). Under physiological conditions, the concen-tration of plasma iron is �18 mM and of Tf �56 mM. Thus, onlyone-third of Tf is saturated with iron, of which �10% is in thediferric form. However, in most types of hemochromatosis(exceptions are discussed below), Tf saturation graduallyreaches almost maximum capacity, and this leads to accumu-lation of redox active low molecular weight iron in plasma andin tissues. Loading of Tf with iron very likely requires the

10 Pantopoulos

activity of ceruloplasmin, a plasma blue copper ferroxidase of1065 amino acids, despite the fact that Tf has intrinsic ferrox-idase activity. Ceruloplasmin’s membrane-bound homologuehephaestin (56), which is expressed in enterocytes, very likelyfacilitates the basolateral transport of dietary iron.

Functional inactivation of genes involved in dietary ironabsorption or in iron delivery to the erythron is associated withdisease. This is evident in microcytic anemia (mk) mice (49)and in Belgrade (b) rats (57), where a G185R mutation withina predicted transmembrane domain of DMT1 impairs ironabsorption (58) due to improper targeting to the apical mem-brane of enterocytes (59). Similarly, a mutation resulting ina premature stop codon (K264X) accounts for the chardonnay(cdy) phenotype in zebrafish (60). Taken together, in theseanimal models, defects in DMT1 activity lead to iron deficiencyand severe hypochromic microcytic anemia.

Defective expression of hephaestin in hemizygous malesex-linked anemia (sla) mice is also associated with microcyticanemia due to inefficient iron delivery to the circulation (56).

Figure 4 A model for iron export from macrophages. Intracellularferrous iron, mainly derived from phagocytozed senescent erythro-cytes, is transported across the plasma membrane of macrophagesand delivered to circulation for reutilization by expanding erythro-blasts. This step is coupled with reoxidation of ferrous to ferric ironby soluble ceruloplasmin. Plasma iron is immediately scavenged bytransferrin (Tf).


The important function of ceruloplasmin in body iron metabo-lism is evident by the severe iron overload associated withhuman aceruloplasminemia (61,62). This disorder is charac-terized by normal Tf saturation, poor response to venesectiontherapy due to ensuing anemia, iron loading of macrophagesand, notably, iron deposition in the brain (42), which doesnot occur in other forms of hemochromatosis. Similar patholo-gical manifestations have been documented in Cp�=� mice,generated by targeted disruption of the ceruloplasmin gene(63). Taken together, the findings with aceruloplasminemicpatients and Cp�=� mice provide strong evidence for a physio-logical function of ceruloplasmin in iron mobilization fromreticuloendothelial cells and storage sites.

In humans, mutations in the SLC11A3 gene encoding fer-roportin result in an autosomal dominant form of hemochro-matosis (HH type IV), characterized by iron loading ofreticuloendothelial cells, normal Tf saturation, and poorresponse to phlebotomy, thus resembling phenotypical hall-marks of aceruloplasminemia (except brain iron accumula-tion) (64). Missense mutations leading to N144H (55) andA77D (65) substitutions were initially described in a Dutchand an Italian pedigree, respectively. Subsequently, a threebase-pair deletion resulting in elimination of V162 has beendocumented in affected individuals from different ethnicbackgrounds (66–69). The A77D substitution maps to the C-terminal end of the predicted transmembrane domain 1, whilethe N144H substitution and the V162 deletion map to theextracellular loop linking domain 1 with domain 2. Interest-ingly, an L167F substitution in the zebrafish, which is alsolocalized within this predicted extracellular loop, underliesthe weissherbst (weh) phenotype of hypochromic anemia (52).The above data are compatible with the idea that thedisease-associated ferroportin mutations are responsible fordefects in the transport of dietary iron across the basolateralmembrane of enterocytes and, moreover, in the mobilizationof iron from macrophages. The latter may more profoundlycontribute to the disease phenotype, but it is unclear howincreased iron absorption and body iron loading occurs in thepresence a defective enterocyte transporter. The existence of

12 Pantopoulos

alternative basolateral iron transporters, besides ferroportin,would provide a reasonable explanation.

The deficient expression of Tf as observed in extremelyrare hypotransferrinemias and atransferrinemias (70,71) isassociated with microcytic anemia, increased dietary ironabsorption, and profound tissue iron overload. Similar symp-toms are observed in hypotransferrinemic (hpx) mice, whichfail to express physiological levels of Tf due to a spontaneoussplicing defect in Tf mRNA (72). The development of severeanemia despite iron overload in hypotransferrinemias andatransferrinemias emphasizes the importance of Tf as thephysiologically relevant donor of iron for erythropoiesis (39).

Heme oxygenase (Hmox1) is involved in the catabolism ofheme from senescent erythrocytes in macrophages and there-fore plays a key role in iron reutilization by the erythron (4).As one would expect, Hmox1�=� mice suffer from microcyticanemia (73). In addition, thesemice display severe pathologicalfeatures, such as growth retardation, chronic inflammation,and tissue iron overload (paradoxically, nonheme iron) in theliver (hepatocytes and Kupffer cells) and in the kidney,despite low Tf saturation. Similar pathology has been observedin the only reported case of human heme oxygenase deficiency(74,75).

REGULATION OF DIETARY IRON ABSORPTIONAND ERYTHROID IRON UTILIZATION

The Role of HFE

It is believed that iron absorption is regulated by signaling toprecursor enterocytes in the crypts of the duodenal epithe-lium (76). These cells undergo a maturation process, whichis associated with migration along the crypt-villus axis.Signals sensed in the crypts program mature enterocytes toabsorb dietary iron from the lumen in response to body irondemands. Macrophages may also respond to similar signalsto regulate the release of iron for erythropoiesis.

The association of HFE mutations with hereditary hemo-chromatosis (20) suggests that HFE, which is primarily


expressed in gastric epithelial cells, circulating monocytesand tissue resident macrophages (77,78), performs an impor-tant regulatory function in systemic iron metabolism. As HFEcan modulate cellular iron metabolism, it is conceivable thatit may also control the iron status of precursor enterocytesand macrophages. This may be critical to program differen-tiating mucosal cells for dietary iron absorption upon matura-tion and to control iron release from macrophages byregulating the expression of genes involved in iron transport(43). Along these lines, analysis of duodenal samples fromhemochromatosis patients with HFE defects showedincreased expression of DMT1 and ferroportin (79,80). How-ever, analysis of hfe�=� mice for expression of DMT1 in theduodenum yielded contradictory results (81,82). A recent ana-lysis of gene expression profiles in duodenal samples fromhfe�=� and hfeC282Y=C282Y mice showed increased expressionof Dcytb mRNA, without any changes in the mRNA levels ofDMT1 and ferroportin (50).

It should be noted that duodenal crypt cells (83,84) andmacrophages (85) from patients with HFE-related hemochro-matosis are not only spared from iron overload, but alsoappear to be are iron-deficient. The viral delivery of wild-typeHFE to cultured monocytes from such patients resulted innormalization of iron loading by Tf (86). Taken together, theseobservations would be consistent with a view where HFEplays a positive role in the loading of duodenal crypt cellsand macrophages with iron. Experimental support comesfrom findings showing impaired capacity of duodenal cryptcells from hfe�=� mice to take up iron from plasma Tf (87).In addition, monocytes from patients with HFE-related hemo-chromatosis are able to release twice as much iron than con-trol cells (88). Finally, expression of chimeric HFE inhibitsiron efflux from THP-1 monocytic=macrophage cells (89).The molecular mechanisms underlying HFE function remainunclear, despite the intense efforts in the last few yearsto characterize this protein. The employment of cell culturemodels has shed little light on the activity of HFEactivity in vivo, which may well be dependent on intercellularsignals.

14 Pantopoulos

The Role of Hepcidin

It has long been proposed that the pathway for iron absorptionis regulated by sensing body iron stores and the requirementof iron for erythropoiesis (90). The nature of the so-called‘‘stores’’ and ‘‘erythroid’’ regulator has remained elusive, buta growing body of evidence suggests that the antimicrobialpeptide hepcidin is a very good candidate to fulfill suchfunctions (91,92).

Hepcidin is a cysteine-rich peptide (Fig. 5) synthesizedpredominantly in the liver and secreted in the plasma. Itwas first isolated from human blood ultrafiltrate (93) andurine (94) and studied for its bacteriostatic properties. In anindependent screen for hepatic genes related to iron metabo-lism, hepcidin was found overexpressed in response to ironoverload (95). Recent genetic data, initiated by a serendipi-tous observation, have provided strong evidence that hepcidinis an important regulator of iron metabolism. The targeteddisruption of the gene for murine transcription factor Usf2(upstream stimulatory factor 2) resulted in the silencing ofthe downstream hepcidin gene. Analysis of the Usf2�=� miceestablished that the lack of hepcidin expression associateswith profound iron overload in tissue parenchymal cells(96,97). Conversely, transgenic mice expressing high levelsof hepcidin suffer from severe iron deficiency anemia; themajority of animals die within a few hours after birth (97).Hepcidin expression in the mouse is negatively regulated byiron deficiency (98), phlebotomy-induced anemia, phenylhy-

Figure 5 The sequence of mature human hepcidin with theexperimentally established (248) cysteine connectivity. Hepcidin issynthesized as a precursor of 84 amino acids. Cleavage of the leadersequence at the positions indicated by arrows gives rise to threemature isoforms of 25, 22, or 20 amino acids, respectively.


drazine-induced hemolytic anemia, and hypoxia (99). On theother hand, turpentine-induced inflammation (99) or interleu-kine (IL)-6 (100) stimulate hepcidin expression. Takentogether, these data have established a function of hepcidinas a novel iron-regulatory humoral factor controlling dietaryiron absorption and utilization for erythropoiesis (Fig. 6).

On the basis of the above findings, it can be easily pre-dicted that misregulation of hepcidin expression would associ-ate with disease. Several lines of evidence support this view.First, increased hepcidin expression was found in hepatic ade-noma tissue from patients with glycogen storage disease thathad developed severe iron refractory anemia (101). In addi-tion, patients with ACD or transfusional iron overload showedincreased urinary secretion of hepcidin (100). A first geneticlink between hepcidin and iron overload was made with thedemonstration that the deletion of 93G or a C166T substitu-tion in the Hamp gene encoding hepcidin are associated witha severe, 1q-unlinked form of juvenile hemochromatosis intwo analyzed pedigrees (102). The deletion of 93G results in

Figure 6 A model for the regulatory functions of hepcidin. Adecrease in body iron stores, a requirement of iron for erythropoi-esis or hypoxia lead to a drop in plasma hepcidin levels, which, inturn, promotes dietary iron absorption and iron release from macro-phages. Excessive body iron or inflammation stimulates hepcidinexpression and accumulation in plasma, which inhibits dietary ironabsorption and iron release from macrophages.

16 Pantopoulos

a frameshift, predicted to yield an extended peptide of179 amino acids with a completely distorted structure. TheC166T substitution generates a premature R56X terminationcodon in the propeptide, resulting in elimination of all maturesequences. These data imply that the relatively more common1q-linked form of juvenile hemochromatosis may be asso-ciated with defects in molecules involved in the hepcidinpathway.

A Common Pathway for Signaling by HFE andHepcidin?

Recent data have shown that the expression of hepcidin isinappropriately decreased in individuals with ‘‘classical’’HFE-related hemochromatosis (103,104) and in hfe�=� mice(50,103,105). Moreover, the crossing of hfe�=� mice withtransgenic mice overexpressing hepcidin corrected iron over-load in the progeny (106). These findings suggest that the lackof feedback regulation in iron absorption and the ensuing pro-gressive iron overload observed in HFE-related hemochroma-tosis may reflect defects in hepcidin signaling. They also raisethe interesting possibility that the liver has a major functionin the regulation of iron absorption and reutilization via HFEand hepcidin, which are expressed in resident macrophages(Kupffer cells) or in parenchymal cells (hepatocytes), respec-tively. It is expected that this rapidly developing area ofresearch will soon provide insights on the pathophysiologyof iron absorption and reutilization, and the molecularfunction of HFE and hepcidin.

CELLULAR IRON UPTAKE

Erythroid and most nonerythroid cells take up iron from Tf.The pathway involves binding of iron-loaded Tf to the cell sur-face transferrin receptor 1 (TfR1), a homodimeric glycoproteinof 180kDa (14,107). Each subunit has a short cytoplasmic tail(residues 1–67), a single transmembrane-spanning domain(residues 68–88), and a large ectodomain (residues 89–760),


which binds one molecule of ligand (108). At neutral pH, difer-ric Tf has an approximately 30- and 500-fold higher bindingaffinity to TfR1 than monoferric and apoTf, respectively(14). The internalization of TfR1 occurs by receptor-mediatedendocytosis (Fig. 7) involving clathrin-coated pits. Acidifica-tion of the endosome to pH �5.5 by the activity of a protonpump results in the release of Fe(III) from TfR1-bound Tf.The apoTf-TfR1 complex recycles on the cell surface, whileFe(III) is reduced to Fe(II) and transported across the endo-somal membrane into the cytosol, where it is utilized for

Figure 7 Schematic representation of the Tf–TfR cycle. Plasmadiferric transferrin (Tf) binds to cell surface TfR (TfR1 or TfR2)and the Tf–TfR complex is internalized by endocytosis. Acidificationof the endosome results in the release of ferric iron from Tf andreduction and subsequent transport of ferrous iron across the endo-somal membrane by DMT1. In the cytoplasm, iron is utilized for thesynthesis of iron-containing proteins and excess is stored in ferritin.The pathway is completed by recycling of the apoTf–TfR complex tothe cell surface and release of apoTf. Association of TfR1 with HFEimpairs the binding of extracellular diferric-Tf and negatively regu-lates the cycle.

18 Pantopoulos

synthesis of iron-containing proteins, or targeted to mitochon-dria for heme synthesis (a major event in erythroid cells). Thetransport of Fe(II) across the endosomal membrane is mostlikely mediated by DMT1 (57,109).

A fraction of cytosolic iron remains bound to lowmolecularweight chelates, which presumably include citrate, ATP, pyro-phosphates, or ascorbate. This fraction of chelatable iron is alsoknown as ‘‘regulatory iron pool’’ or ‘‘labile iron pool’’ (LIP) andreflects the iron status of the cell (110). It can be monitored byfluorescent techniques (111,112). Not much is known aboutiron metabolism in intracellular organelles. Genetic and bio-chemical data suggest that frataxin, a gene mutated inpatients with Friedreich’s ataxia (113), may play an importantrole for the maintenance of iron homeostasis in the mitochon-dria, but its exact function is still a matter of debate (114).

Targeted disruption of the TfR1 gene in mice is embryo-nic lethal and TfR1�=� embryos die before day E12.5 of gesta-tion (115). The early development up to day E12.5 isassociated with severe defects in erythropoiesis and neuro-genesis. Mice retaining a functional TfR1 allele (TfR1þ=�)develop hypochromic microcytic anemia due to iron defi-ciency. These results have confirmed that the Tf–TfR1 cycleis the major and probably, after embryonic day E12.5, theonly route for iron uptake by erythroid cells. They alsoemphasize the importance of the Tf–TfR1 cycle in earlyembryonic development of the nervous system.

The Tf-TfR cycle can be negatively modulated by HFE,because this protein forms complexes with TfR1 (116,117),which impair the capacity of TfR1 for iron uptake (117–119),possibly via competition for Tf binding (120). In vitro, theHFE=TfR1 interaction occurs readily at pH 7.5 (reflecting con-ditions on the cell surface), but is abrogated at pH 6.0 (reflect-ing conditions in the endosome) (121). The crystal structure ofHFE complexed with the extracellular portion of TfR1 showsthat the interaction involves the a1 and a2 domains of HFEand induces conformational changes in TfR1 (122). Overex-pression of HFE in cell lines generates an iron-deficient pheno-type (123–126). However, it is unclear whether this response isrelevant in the context of hemochromatosis.


A second transferrin receptor, known as TfR2, is mostlyexpressed in the liver parenchyma and in some nonhepatic celllines of erythroid=myeloid origin, such as K562 and KG-1(127,128). Interestingly, TfR2 was also reported to beexpressed in intestinal crypt cells and, moreover, to colocalizewith HFE (129). By analogy to TfR1, TfR2 is composed of twohomodimeric subunits, containing a cytoplasmic portion(residues 1–80), a transmembrane-spanning domain (residues81–104), and an ectodomain (residues 105–801). TfR2 andTfR1 subunits may also form heterodimers (130). The deducedamino acid sequence of human TfR2 ectodomain displays 45%identity and 66% similarity with the respective ectodomain ofTfR1 (127). Two TfR2 transcripts have been detected: a�2.9 kb full-length (TfR2-a) and a shorter �2.5 kb TfR2-b,which gives rise to a presumably intracellular, N-terminallytruncated protein lacking amino acids 1–541. Iron uptakeexperiments show that TfR2 can internalize 55Fe-Tf, similarto TfR1. However, the experiments withTfR1�=� andTfR1þ=�

mice (115) indicate that TfR2 cannot compensate for TfR1insufficiency, at least in the context of erythropoiesis and earlyneuronal development.

An unexpected hint to the physiological function ofTfR2 came with the discovery that a nonsense mutationin TfR2 mRNA generating a premature termination codon(Y250X) is associated with hereditary iron overload (131).The disease underlying TfR2 disfunction is now classifiedas HH of type III and has clinical manifestations similarto HFE-related hemochromatosis. Targeted generation ofthe orthologous Y245X mutation has recapitulated this phe-notype in mouse (132). Other disease-associated TfR2 muta-tions were subsequently documented. These include asecond premature termination codon (E60X) resulting by anucleotide insertion-induced frameshift, which only affectsthe TfR2-a transcript, and a M172K substitution, whichinactivates the predicted initiation codon of the TfR2-b tran-script (133). In addition, the deletion of four amino acids(AVAQ 594–597) (134) and a Q690P substitution, whichmay affect the binding of Tf (135), have also been associatedwith disease. The mechanism for disease pathogenesis is

20 Pantopoulos

still unclear, but it is tempting to speculate that the func-tional inactivation of TfR2 may initiate responses leadingto reduced expression of hepcidin.

The findings discussed above underscore the importantphysiological role of the Tf–TfR route in cellular iron uptake.However, it should be noted that there are additional, yet stillincompletely characterized mechanisms for iron acquisition.Experiments with cultured cells have provided good evidencefor ‘‘nontransferrin-bound iron’’ (NTBI) uptake mechanisms(136–140), some of which may also involve DMT1 (141). Addi-tional arguments are provided by the fact that atransferrine-mia is not lethal and, furthermore, affected individualsaccumulate iron in tissues (39). Furthermore, despite theembryonic lethality associated with the targeted disruptionof TfR1 in mice, embryonal development progresses until upto E12.5 (115). Thus, it is likely that alternative pathwaysto the well-established Tf–TfR route may be predominantduring early embryonic stages.

Recent experiments showed that the neutral gelatinase-associated lipocalin (NGAL), a neutrophil-derived proteininvolved in the delivery of various small molecules to cells(142), binds to iron-loaded enterobactin (143). Moreover, themurine homologue of NGAL,M24p3, can specifically transportiron to the cytoplasm of target cells and promote mesenchymalto epithelial cell differentiation (144). Siderophores are lowmolecular weight iron-chelates, generated and released bybacteria or fungi to scavenge extracellular iron, which is thentaken up by the microorganisms via binding to specific recep-tors (45). Even though it is still not clear whether mammalscan produce siderophores themselves, the above findings sug-gest that they can utilize siderophore-based mechanisms foriron acquisition. They also raise the interesting possibilitythat the NGAL system may represent a major pathway forNTBI uptake. Siderophores bind to iron with extremely highaffinities (45). For example, the Kd of enterobactin for Fe(III)is 10�49M (143), which corresponds to a 1026 times higher affi-nity than that of Tf (Kd ¼ 10�23). Considering that NGALexpression is stimulated during inflammation (142), it hasbeen speculated that this protein may contribute to the drop


in plasma iron encountered under these conditions (145),which is a hallmark of ACD.

INTRACELLULAR IRON STORAGE

Excess of intracellular iron is sequestered and detoxified in fer-ritin, which serves as the major iron storage protein (Fig. 7).Iron overloaded cells also store iron in hemosiderin, a degrada-tion product of ferritin. Ferritin is composed of 24 subunits ofH- and L-chains (containing 190 and 175 amino acids, respec-tively) (146,147). These assemble to a symmetric shell-likestructure forming a cavity of �80 A, with a potential to storeup to 4500 Fe(III) ions in form of ferric oxy-hydroxide phos-phate. Iron incorporation into ferritin requires a ferroxidaseactivity associated with H-subunits and a nucleation centerassociated with L-subunits. The composition of ferritin H-and L-subunits differs in various tissues. For example, H-ferri-tin is enriched in the heart, while L-ferritin predominates inthe liver.

Iron stored in ferritin can be mobilized during iron defi-ciency for metabolic needs, but the mechanism is incompletelycharacterized. Iron mobilization may be coupled with ferritinturnover in lysosomes (148), but it is also possible that ferritinsubunits play an active role in iron release via structuralrearrangements (149). The targeted disruption of the geneencoding H-ferritin associates with embryonic lethality.Fth�=� mice die in utero between embryonic days E3.5 andE9.5 (150), suggesting an important function of H-ferritin inearly development. Heterozygous Fthþ=� mice do not showany apparent abnormalities (151).

H- and L-ferritins are exclusively expressed in the cytosol.A mitochondrial homologue of ferritin, encoded by an unusualintronless nuclear gene, has recently been described (152). Theprotein is synthesized in a precursor form of 242 amino acids,which is targeted to mitochondria by an N-terminal leadersequence of 57 amino acids. The mature protein has ferroxi-dase activity and assembles into functional ferritin shells.There is no evidence suggesting a role of mitochondrial ferritin

22 Pantopoulos

as an obligatory intermediate in mitochondrial iron utilization(153). Its expression in normal mitochondria is weak. How-ever, mitochondrial ferritin is highly abundant in iron-loadedring erythroblasts (sideroblasts) frompatientswith sideroblas-tic anemia (154). This finding suggests a function of mitochon-drial ferritin as a sink for iron deposition in the context of thisdisorder.

A secreted, glycosylated form of ferritin circulates inplasma (14), but its origin and exact function are unknown.Plasma ferritin does not appear to be involved in iron trans-port or storage, because its iron content is low. The concentra-tion of serum ferritin, which is usually <200mg=L in womenand <300mg=L in men, is a useful marker for body iron stores(1 mg=L of serum ferritin corresponds to approximately 8mgstorage iron). However, its usefulness as a marker is limited,because serum ferritin concentration increases independentlyof body iron stores during pathological conditions associatedwith acute or chronic inflammation, such as liver disease,cancer, or ACD. Another characteristic example is the heredi-tary hyperferritinemia-cataract syndrome (HHCS), whereserum ferritin levels increase up to 20-fold in the absence ofiron overload (155). The disease is caused by misregulationin the expression of L-ferritin. The molecular basis of HHCSwill be discussed in the next section. Interestingly, Fthþ=�

mice also display hyperferritinemia (7- to 10-fold increase inserum L-ferritin) in the absence of iron overload (151).

The biomedical relevance of ferritin function is also evi-dent from some recent findings. First, a misregulation inthe expression of H-ferritin (it will also be discussed in thenext section) has been associated with a dominantly inheritedform of hemochromatosis (156). In addition, a frameshiftmutation in L-ferritin gene, predicted to alter 22 residues atthe C-terminus of the polypeptide and extend it by 4 addi-tional amino acids, is a causative defect in a dominantadult-onset basal ganglia disease (157). This rare neurode-generative disorder, which may be related to nonreversiblestorage of metabolically active iron within the mutated ferri-tin, has been termed ‘‘neuroferritinopathy’’ and highlights the


important role of ferritin in the context of brain ironmetabolism (158,159).

POSTTRANSCRIPTIONAL REGULATIONOF CELLULAR IRON METABOLISMBY THE IRE=IRP SYSTEM

The expression of TfR1 and ferritin is controlled at the post-transcriptional level by the IRE=IRP system (160,161), whichwas first described in the late 1980s. The mRNAs encodingTfR1 and ferritin (both H- and L-chains) contain ‘‘iron respon-sive elements’’ (IREs) in their untranslated regions (UTRs).These are hairpin structures of about 30 nucleotides (Fig. 8),

Figure 8 The IRE consensus motif. It consists of a hexanucleotideloop (50-CAGUGN-30) and a stem, interrupted by a bulge with anunpaired C residue. Base pairing between C1 and G3 is functionallyimportant. N6 could be any nucleotide but not G, which wouldpotentially disrupt C1–G3 interaction by C1–G6 pairing. The bulgemay consist of an asymmetric tetranucleotide as in many ferritinmRNAs (left), or a single C residue (right).

24 Pantopoulos

phylogenetically conserved in vertebrates and some insectsand bacteria (162). A typical IRE forms a 50CAGUGN-30 loop(the underlined C and G interact by hydrogen bonding) and astem with moderate stability (DG ��7kcal=mol), interruptedby a bulge with an unpaired C residue.

TfR1 mRNA contains five IREs in its long and complex 30

UTR, while H- and L-ferritin mRNAs contain a single IRE intheir 50 UTRs, respectively. In iron-deficient cells, IREs pro-vide a high affinity (Kd� 10�12M) binding site for ‘‘iron regula-tory proteins,’’ IRP1 or IRP2. The IRE=IRP complexes protectthe otherwise unstable TfR1 mRNA from targeted nucleolyticdegradation, and specifically inhibit ferritin mRNA transla-tion. As a result, iron-starved cells increase their capacity totake up essential iron by the Tf–TfR1 route and minimize its‘‘waste’’ in ferritin stores. Conversely, in iron-replete cells,IRE-binding activity is diminished, thereby triggering oppo-site responses that prevent further uptake and promotesequestration of redundant iron (Fig. 9). The family of IRE-containing mRNAs has expanded over the years. It appearsnow that a variety of proteins of iron and energy metabolismare subjected to regulation by the IRE=IRP system. Theseinclude the erythroid-specific isoform of 5-aminolevulinatesynthase (ALAS-2), a key enzyme in erythroid heme synthesis,mammalian mitochondrial (m-) aconitase and insect Ipsubunit of succinate dehydrogenase (SDH), both enzymes ofthe citric acid cycle, and the iron transporters DMT1 andferroportin (163).

The mechanism underlying the regulation of TfR1mRNA stability by IRE=IRP interactions is not well under-stood. TfR1 mRNA remains, so far, the only identified mRNAwith multiple IREs. Early experiments showed that the mini-mum requirement for regulation is defined by a combinationof three IREs together with additional non-IRE sequences(164). The pathway for TfR1 mRNA degradation involves aninitial endonucleolytic cleavage and is not associated withalterations in the length of the poly(A) tail (165). Trans-actingfactors still remain to be identified.

The function of IRE as a translational regulatory ele-ment is much better defined. Initial hints were provided by


the analysis of mRNA sequences from several vertebrate fer-ritins, which revealed that the IRE is localized close to the cap(166). The cap-proximal segment is critical for interactions ofthe mRNA with the small (40S) ribosomal subunit and therecruitment of translation initiation factors to assemble a43S preinitiation complex. Iron regulatory protein-bindingto ferritin IRE inhibits ferritin mRNA translation by steri-cally abrogating the stable association of the 40S ribosomewith the initiation factor eIF-4F (167,168). The translationof ALAS-2 (169) and m-aconitase (170,171) mRNAs is regu-lated in a similar manner, but in a somehow narrower range.The quantitative differences in the degree of regulation mayreflect differences in IRE sequences and structure.

The function of DMT1 and ferroportin IREs is as yet illdefined. Divalent metal transporter 1 mRNA contains asingle IRE in its 30 UTR (an alternatively spliced varianthas no IRE). The levels of the IRE-containing isoform are

Figure 9 Regulation of cellular iron metabolism by the IRE=IRPsystem. Decreased iron supply activates binding of IRPs to IREs,resulting in stabilization of TfR1 mRNA and translational inhibi-tion of the mRNAs encoding H- and L-ferritin. These responses leadto increased iron uptake and reduced iron storage. Conversely,increased iron supply inactivates binding of IRPs to IREs, resultingin degradation of TfR1 mRNA and translation of the mRNAs encod-ing H- and L-ferritins. These responses lead to decreased ironuptake and elevated iron storage.

26 Pantopoulos

profoundly increased in the iron deficient duodenum (48),suggesting the involvement of an IRE=IRP mediated mechan-ism. Experiments in cultured cells suggest that the IRE-dependent regulation of DMT1 expression may be cell-typespecific and restricted to intestinal cells (172–174). Interest-ingly, the functionality of the IRE appears to requireelements within a previously undetected alternatively splicedexon at the 50 end of DMT1 mRNA (175). Ferroportin mRNAcontains a single IRE at the 50 UTR and its expression issimilarly stimulated in the iron deficient duodenum (53).However, the contribution of the IRE to this response remainsobscure.

In summary, the IRE=IRP system provides a simplemechanistic basis for the co-ordinate regulation of cellulariron uptake, transport, storage, and erythroid utilization,and also links iron with cellular energy metabolism. The firstcharacterized genetic defect in this system associates with theHHCS. Ironically, this dominantly inherited disease lacksany apparent (in terms of general clinical parameters) effectson body iron metabolism. Its hallmark is a substantial (up to20-fold) increase of serum L-ferritin levels in the absence ofiron overload, which correlates with development of cataract(155). The molecular defect lies in mutations in the IRE ofL-ferritin mRNA that render it functionally inactive, preventbinding of IRPs, and thus abolish its translational regulation.Experimental evidence that the hyperferritinemic phenotypeis due to failure of IRPs to control L-ferritin synthesis wasfirst provided by experiments with an L-ferritin IRE tran-script bearing the disease-associated A32G mutation (176).This mutant fails to compete wild-type L-ferritin IRE forIRP binding, even at 500-fold molar excess.

A series of disease-associated mutations in L-ferritin IREhave been identified in individuals of different ethnic back-grounds. Recently updated lists can be found in Refs. 155,177. The IRE mutations can be classified into four groups(see Fig. 8 for reference):

i. deletions in nucleotides forming part of the stemor the loop,


ii. point mutations in the loop or the unpaired C inthe bulge,

iii. point mutations in the upper stem or othernucleotides of the bulge, or

iv. point mutations in the lower stem.

Type (i) and (ii) mutations are associated with the mostacute phenotypes; serum L-ferritin 1200–2700mg=L andsevere cataract (178). Type (iii) mutations correlate withslightly lower levels of L-ferritin (950–1900mg=L) and mildercataract, while type (iv) lead to only a moderate increase inL-ferritin levels (350–650mg=L) and asymptomatic cataract(178). Analysis of different mutants suggested a correlationbetween the degree of inhibition of IRP1 and IRP2 bindingand the severity of HHCS (179), but a clinical variabilityamong individuals sharing the same mutation suggests theinvolvement of additional factors (155). How hyperferritine-mia contributes to the pathogenesis of cataract is unclear.In lymphoblastoid cell lines, as well as in lens from HHCSpatients (recovered from surgery), overproduction of L-ferritin shifts the H-=L-equilibrium in holo-ferritin, and more-over leads to the accumulation of L-homopolymers (180). Thedevelopment of animal models is expected to shed more lightinto this issue.

A point mutation (A49U) in H-ferritin IRE has beencorrelated with an autosomal dominant iron overload diseasein a Japanese pedigree, with iron deposition primarily inhepatocytes (156). A recent review on hemochromatosistermed this condition ‘‘HH of type 5’’ (26). It appears that thismutation increases the affinity of IRPs, resulting in inhibitionof H-ferritin synthesis, but it remains elusive how this leadsto iron overload.

IRON REGULATORY PROTEINS, IRP1AND IRP2

IRP1 and IRP2 are cytoplasmic polypeptides of 889 and 964amino acids, respectively (181,182). They are both membersof the iron–sulfur cluster isomerase family and display high

28 Pantopoulos

homology to m-aconitase (183). Human IRP1 shares 57%sequence identity and 75% similarity with human IRP2. Inaddition, human IRP1 is 31% identical and 56% similar toporcine m-aconitase. By analogy to the known structure ofm-aconitase (184), IRP1 and IRP2 are projected to containthree compact domains, linked to a fourth domain by a flex-ible hinge region (Fig. 10). A notable difference betweenIRP1 and IRP2 is that the latter contains a cysteine- andproline-rich insertion of 73 amino acids, embedded withindomain 1, which is encoded by a unique exon.

The expression of IRP1 is ubiquitous, while IRP2appeared initially to be expressed in a tissue-specific manner(185). However, it is possible that the abundance of IRP2 intissues and cells may have been underestimated (186), inmany instances due to technical difficulties in detection(187). Both IRP1 and IRP2 bind to naturally occurring IREswith similar affinities (188), but show a preference in bindingtowards engineered IRE variants (189,190). Interestingly, inmurine RAW 264 macrophages, the exclusive induction ofIRP2 is sufficient to regulate the expression of TfR1 mRNA(191) and ferritin synthesis (192), regardless of IRP1 activity.On the other hand, the expression of IRP1C437S, a constitutiveIRP1 mutant, in human RD4 (rhabdomyosarcoma) cells (193)or in H1299 (lung cancer) cells (194) is sufficient to disruptiron-dependent regulation of TfR1 and ferritin.

Mechanisms for Regulation of IRP1

In iron-loaded cells, IRP1 assembles a cubane 4Fe–4S clusterand is thereby converted to a cytosolic (c-) aconitase (Fig. 10a).Three iron atoms co-ordinate to C437, C503, and C506 in thepolypeptide backbone, while the fourth, Fea binds to the sol-vent (H2O) and is directly involved in catalysis (isomerizationof citrate to isocitrate). Both c- and m-aconitases display simi-lar catalytic efficiencies (195,196), but the physiological func-tion of c-aconitase is unclear. Iron starvation promotes aslow disassembly of the 4Fe–4S cluster and the resultingapoIRP1 acquires IRE-binding activity. This may trigger astructural rearrangement leading to a more ‘‘open’’ cleft that


Figure 10 Models for the regulation of IRP1 (a) and IRP2 (b),which, based on homology with mitochondrial aconitase, aredepicted as proteins containing three compact domains (1–3) linkedto a fourth (4) via a flexible hinge region. (a) The post-translationalregulation of bifunctional IRP1 in response to iron, NO, and H2O2 ismediated by an iron–sulfur cluster switch. In iron-replete cells,IRP1 assembles a cubane 4Fe–4S cluster between domains 1–3and 4. Iron starvation, NO, and H2O2 trigger the switch of4Fe–4S– to apoIRP1, resulting in the conversion of cytosolic aconi-tase to IRE-binding protein. (b) Regulation of IRP2 at the levelof protein stability. IRP2, which contains a 73 amino acid inser-tion within domain 1, is stable in iron-starved, hypoxic, or NO-treated cells. Increased iron supply or exposure to the nitrosoniumcation (NOþ) lead to IRP2 ubiquitination and degradation by theproteasome.

30 Pantopoulos

allows mRNA binding (197,198). Thus, in contrast to m-aconi-tase, IRP1 is a bifunctional protein, controlled by an unusualiron–sulfur cluster switch (160,181).

Besides iron starvation, exposure of cells to NO andhydrogen peroxide (H2O2) also stimulates conversion ofIRP1 from c-aconitase to IRE-binding protein (199,200).These findings warrant consideration both from the mechan-istic, as well as the pathophysiological point of view. A plausi-ble model that the 4Fe–4S cluster of IRP1 is directly removedby NO and ROIs may not reflect the complexity of the system.Nitric oxide functions both as an inter- and intracellular effec-tor to IRP1 and elicits a relatively slow (> 4h) activation ofIRE-binding (201–203). It has been proposed that NO mayinduce IRP1 indirectly by modulating the pool of intracellularchelatable iron, independently, or in addition to destabilizingits 4Fe–4S cluster (182,201,203). In contrast, extra- but notintracellular H2O2 triggers a rapid induction of IRE-bindingwithin 30–60min (204,205). The latter appears to be a resultof an oxidative stress–response signaling pathway (206) andcorrelates with modulation of downstream genes, such asTfR1 and ferritin (174). Phosphorylation of IRP1 at S138appears to negatively affects the assembly of the cluster(207), but it remains unclear whether this is related to signal-ing by NO or H2O2. Both NO and H2O2 are reactive speciesgenerated at high levels by phagocytic cells during the inflam-matory response. The activation of IRP1 by these stimuli isexpected to promote increased cellular uptake from Tf-boundserum iron and may thus contribute to the shift of extracellu-lar iron to the tissues encountered in ACD.

Mechanisms for Regulation of IRP2

The expression and activity of IRP2 is regulated at the level ofprotein stability (Fig. 10b) and the protein neither assemblesan aconitase-type cluster nor exhibits enzymatic activity. Irondeficiency triggers de novo synthesis and stabilization of IRP2(208,209), while increased iron levels promote its degradationby the proteasome (210,211). It has been proposed that the 73


amino acids segment within domain 1 functions as ‘‘iron sen-sor’’ (211). According to a model (212), iron may directly bindto C168, C174, and C178 and promote their site-specific oxida-tion, which, subsequently, tags the protein for degradationfollowing ubiquitination. However, it should also be notedthat heme has earlier (213) and more recently (214) beenproposed to serve as a signal for IRP2 degradation.

IRP2 is also modulated by iron-independent signals, suchas NO (199,200). Some of the published data appear to be con-flicting. For example, IRP2 was upregulated in B6 fibroblaststransfected with an NOS cDNA (215). However, in other stu-dies employing g-IFN=LPS-stimulated J774 (216) or RAW(217,218) macrophages to induce NOS, IRP2 activity wasdiminished. These seemingly contradictory results may berelated to pleiotropic effects of cytokines, differences in intra-cellular redox status and differential responses of IRP2 to var-ious redox species of NO. Along these lines, the treatment ofRAW cells with sodium nitroprusside, which liberates NOþ

and serves as nitrosylating agent, results in the degradationof IRP2 (192). Interestingly, this is associated with a profoundstimulation in ferritin synthesis, which may also be relevantin the context of ACD (192). IRP2 does not respond to extra-cellular H2O2 (203), but its IRE-binding activity is negativelyregulated by menadione, as is the case with IRP1 (219). Ofparticular interest is the response of IRP2 to hypoxia. Adecrease in oxygen tension or treatment of cells with cobalt(that mimics hypoxic conditions) lead to IRP2 stabilization(220) without activating IRP1 (221).

Targeted Disruption of IRP1 or IRP2

The function of IRP1 and IRP2 in the body has been examinedby gene targeting experiments. IRP1�=� mice lack any dis-cernible phenotype (187). By contrast, IRP2�=� mice displayaberrant iron homeostasis and accumulate iron in the intest-inal mucosa and the central nervous system and, moreover,develop a progressive neurodegenerative disorder (222). Thesedata establish a major regulatory function role of IRP2 in sys-temic iron metabolism. How the loss of IRP2 function leads to

32 Pantopoulos

iron overload in the brain is currently unclear, but it has beenspeculated that neurons may encounter a condition of ‘‘func-tional iron depletion’’ due to accumulation of ferritin and ironsequestration within it (222). Interestingly, this scenario isreminiscent of the ‘‘neuroferritinopathy’’ (158,159).

OTHER REGULATORY MECHANISMS

The IRE=IRP system plays an important role for theregulation of TfR1, ferritin, and other proteins, but additionalregulatory mechanisms are also operative. It is well estab-lished that TfR1 expression is also controlled at the transcrip-tional level (107). This is particularly relevant in the context oferythroid cell differentiation, where IRP-mediated mechan-isms appear to have minor effects (223,224). Transcriptionalregulation of TfR is also evident in other examples. The tetra-cycline-inducible expression of IRP1C437S in H1299 cells wasassociated with a �100-fold increase of IRE-binding activity(194). This resulted in a mere �3-fold stimulation of TfR1mRNA, while an overnight treatment of cells with the ironchelator desferrioxamine activated TfR1 mRNA levels �30-fold by a pathway sensitive to actinomycin D. In the same set-ting, IRP1C437S efficiently inhibited ferritin mRNA translationin low-density cultures, but this effect was completely over-come when cells reached high densities (194).

The promoter of TfR1 contains binding sites for activatorprotein 1/cAMP response element binding protein (AP-1=CREB-like factors, the Ku autoantigen, the DNA-binding activ-ity of transferrin receptor transcriptional control element(TRAC) factor (107), but also for Ets-like factors. The Ets- andAP-1=CREB-like factors bind together to a bipartite erythroidactive element and account for transcriptional activation ofTfR1during erythroid cell differentiation (224). TheTfR1promo-ter also contains hypoxia response elements in form of a HIF-binding site (225,226). Interestingly, HIF-binding sites are alsopresent in the promoters of Tf (227) and ceruloplasmin (228).

Cytokine signaling has important implications for cellulariron metabolism during inflammation. For example, IFN-gnegatively affects the expression of TfR mRNA by a


post-transcriptionalIRP-independentmechanism(215,229,230),and this effect can be antagonized by the anti-inflammatory cyto-kines IL-4 and IL-13 (231). Ferroportin expression is negativelyregulated in the inflammatory state (232). The proinflammatorycytokines TNF-a (233), IL-1 (234), and IL-6 (235) transcription-ally stimulate the expression of ferritin in several cell types.Ferritin can also be transcriptionally activated in response tooxidative stress (and xenobiotics), via ‘‘antioxidant responseelements’’ in the promoter of its genes (236,237).

On the other hand, iron can interfere with cytokine sig-naling and affect the progress of inflammation, which ishighly relevant within the context of ACD (238,239). It is gen-erally believed that iron modulates the immune system andleads to expansion of suppressor T (CD8þ) cells (240). Ironalso modulates the function of macrophages and T-helper(CD4þ) cells (239). The loading of macrophages with ironimpairs IFN-g signaling, and results in reduced expressionof MHC class II antigen, generation of neopterin and produc-tion of TNF-a, respectively (241,242). Iron also negatively reg-ulates the expression of the inducible NOS (243), and reducesgeneration of NO by macrophages, which plays an importantrole in antiviral and antimicrobial defence mechanisms. Inaddition, iron indirectly modulates macrophage activity byshifting the balance between T-helper cells of type 1 (Th1)and 2 (Th2) (239). The former produce IFN-g, IL-2 and TNF-a and TNF-b, while the latter produce the anti-inflammatorycytokines IL-4, IL-5, IL-10, and IL-13 (244). An iron-mediatedshift towards a Th2 response pattern is believed to be unfa-vorable in combating viral or bacterial infection (245–247).

Taken together, in the inflammatory state, complexarrays of IRP-dependent and independent mechanisms regu-late the expression of genes of iron metabolism, which in turnmodulate cellular iron levels and thereby the function ofimmune effector cells in elaborate autoregulatory loops.

CONCLUSIONS

Iron is an essential nutrient, but also a potential biohazardwith limited bioavailability. To satisfy metabolic needs and

34 Pantopoulos

minimize the risk for iron-mediated injury, cells and organ-isms have developed elegant homeostatic mechanisms. Someof these mechanisms are now comprehended, at least to someextent, and a considerable amount of knowledge has beenaccumulated over the past few years. The recent advancesin understanding pathophysiologal aspects of iron absorptionwith the discovery of key molecules implicated in this processwere preceded by elucidating basic principles of cellular ironmetabolism. Unraveling the regulatory pathways for dietaryiron absorption and iron traffic in the body is expected toprovide further insights for the management of pathologicalconditions, such as hemochromatosis, iron deficiency anemia,and the anemia of chronic disease.

ACKNOWLEDGMENTS

KP is a scholar of the Canadian Institutes of Health Research(CIHR) and a researcher of the Canada Foundation forInnovation (CFI).

REFERENCES

1. Aisen P, Enns C, Wessling-Resnick M. Chemistry and biologyof eukaryotic iron metabolism. Int J Biochem Cell Biol 2001;33:940–959.

2. Frausto da Silva JJR, Williams RJP. The Biological Chemis-try of the Elements. The Inorganic Chemistry of Life. Oxford:Clarendon Press, 1982:319–369.

3. Ponka P. Tissue-specific regulation of iron metabolism andheme synthesis: distinct control mechanisms in erythroidcells. Blood 1997; 89:1–25.

4. Maines MD. The heme oxygenase system: a regulator ofsecond messenger gases. Annu Rev Pharmacol Toxicol 1997;37:517–554.

5. Beinert H, Holm RH, Munck E. Iron–sulfur clusters: nature’smodular, multipurpose structures. Science 1997; 277:653–659.


6. Schofield CJ, Zhang Z. Structural and mechanistic studies on2-oxoglutarate-dependent oxygenases and related enzymes.Curr Opin Struct Biol 1999; 9:722–731.

7. Halliwell B, Gutteridge JMC. Antioxidant defences. FreeRadicals in Biology and Medicine. 3rd ed. Oxford: OxfordUniversity Press, 1998:105–245.

8. Hampton MB, Kettle AJ, Winterbourn CC. Inside the neutro-phil phagosome: oxidants, myeloperoxidase, and bacterialkilling. Blood 1998; 92:3007–3017.

9. Cadenas E. Biochemistry of oxygen toxicity. Annu RevBiochem 1989; 58:79–110.

10. Halliwell B, Gutteridge JMC. The role of free radicals andcatalytic metal ions in human disease: an overview. MethodsEnzymol 1990; 186:1–85.

11. Halliwell B, Gutteridge JMC. Oxidative stress: adaptation,damage, repair and death. Free Radicals in Biology and Med-icine. 3rd ed. Oxford: Oxford University Press, 1998:246–350.

12. Finkel T, Holbrook NJ. Oxidants, oxidative stress and thebiology of ageing. Nature 2000; 408:239–247.

13. Halliwell B, Gutteridge JMC. Free radicals, other reactivespecies and disease. Free Radicals in Biology and Medicine.3rd ed. Oxford: Oxford University Press, 1998:617–783.

14. Ponka P, Beaumont C, RichardsonDR. Function and regulationof transferrin and ferritin. Semin Hematol 1998; 35:35–54.

15. Andrews NC. Disorders of iron metabolism. N Engl J Med1999; 341:1986–1995.

16. DeMaeyer E, Adiels-Tegman M. The prevalence of anemia inthe world. World Health Stat Q 1985; 38:302–316.

17. Weiss G. Pathogenesis and treatment of anaemia of chronicdisease. Blood Rev 2002; 16:87–96.

18. Ajioka RS, Kushner JP. Hereditary hemochromatosis. SeminHematol 2002; 39:235–241.

19. Fleming RE, Sly WS. Mechanisms of iron accumulation inhereditary hemochromatosis. Annu Rev Physiol 2002;64:663–680.

36 Pantopoulos

20. Feder JN, Gnirke A, Thomas W, Tsuchihashi Z, Ruddy DA,Basava A, Dormishian F, Domingo R Jr, Ellis MC, Fullan A,Hinton LM, Jones NL, Kimmel BE, Kronmal GS, Lauer P,Lee VK, Loeb DB, Mapa FA, McClelland E, Meyer NC,Mintier GA, Moeller N, Moore T, Morikang E, Prass CE,Quintana L, Starnes SM, Schatzman RC, Brunke KJ, DraynaDT, Risch NJ, Bacon BR, Wolff RK. A novel MHC class I-likegene is mutated in patients with hereditary haemochro-matosis. Nat Genet 1996; 13:399–408.

21. Gochee PA, Powell LW, Cullen DJ, Du Sart D, Rossi E, OlynykJK. A population-based study of the biochemical and clinicalexpression of the H63D hemochromatosis mutation. Gastro-enterology 2002; 122:646–651.

22. Steiner M, Ocran K, Genschel J, Meier P, Gerl H, Ventz M,Schneider ML, Buttner C, Wadowska K, Kerner W,Schuff-Werner P, Lochs H, Schmidt H. A homozygous HFEgene splice site mutation (IVS5þ1 G=A) in a hereditary hemo-chromatosis patient of Vietnamese origin. Gastroenterology2002; 122:789–795.

23. Andrews NC, Levy JE. Iron is hot: an update on the patho-physiology of hemochromatosis. Blood 1998; 92:1845–1851.

24. Bacon BR, Powell LW, Adams PC, Kresina TF, Hoofnagle JH.Molecular medicine and hemochromatosis: at the crossroads.Gastroenterology 1999; 116:193–207.

25. Beutler E, Felitti VJ, Koziol JA, Ho NJ, Gelbart T.Penetrance of 845G!A (C282Y) HFE hereditary haemo-chromatosis mutation in the USA. Lancet 2002; 359:211–218.

26. Bomford A. Genetics of haemochromatosis. Lancet 2002;360:1673–1681.

27. Levy JE, Montross LK, Cohen DE, Fleming MD, AndrewsNC. The C282Y mutation causing hereditary hemochromato-sis does not produce a null allele. Blood 1999; 94:9–11.

28. ZhouXY,TomatsuS,FlemingRE,Parkkila S,WaheedA, JiangJ, Fei Y, Brunt E, RuddyDA, Prass CE, SchatzmanRC,O’NeillR, BrittonRS, BaconBR, SlyWS.HFEgene knockout producesmouse model of hereditary hemochromatosis. Proc Natl AcadSci USA 1998; 95:2492–2497.


29. Bahram S, Gilfillan S, Kuhn LC, Moret R, Schulze JB,Lebeau A, Schumann K. Experimental hemochromatosisdue to MHC class I HFE deficiency: immune status and ironmetabolism. Proc Natl Acad Sci USA 1999; 96:13312–13317.

30. Fleming RE, Holden CC, Tomatsu S, Waheed A, Brunt EM,Britton RS, Bacon BR, Roopenian DC, Sly WS. Mouse straindifferences determine severity of iron accumulation in Hfeknockout model of hereditary hemochromatosis. Proc NatlAcad Sci USA 2001; 98:2707–2711.

31. Sproule TJ, JazwinskaEC, Britton RS, BaconBR, FlemingRE,Sly WS, Roopenian DC. Naturally variant autosomal andsex-linked loci determine the severity of iron overload inbeta 2-microglobulin-deficient mice. Proc Natl Acad Sci USA2001; 98:5170–5174.

32. Dupic F, Fruchon S, Bensaid M, Borot N, Radosavljevic M,Loreal O, Brissot P, Gilfillan S, Bahram S, Coppin H, RothMP. Inactivation of the hemochromatosis gene differentiallyregulates duodenal expression of iron-related mRNAs betweenmouse strains. Gastroenterology 2002; 122:745–751.

33. Levy JE, Montross LK, Andrews NC. Genes that modify thehemochromatosis phenotype in mice. J Clin Invest 2000;105:1209–1216.

34. de Sousa M, Reimao R, Lacerda R, Hugo P, Kaufmann SHE,Porto G. Iron overload in b2-microglobulin-deficient mice.Immunol Lett 1994; 39:105–111.

35. Santos M, Schilham MW, Rademakers LHPM, Marx JJM,de Sousa M, Clevers H. Defective iron homeostasis in b2-microglobulin knockout mice recapitulates hereditary hemo-chromatosis in man. J Exp Med 1996; 184:1975–1985.

36. Rothenberg BE, Voland JR. b2 knockout mice developparenchymal iron overload: a putative role for class I genesof the major histocompatibility complex in iron metabolism.Proc Natl Acad Sci USA 1996; 93:1529–1534.

37. Cardoso EM, Macedo MG, Rohrlich P, Ribeiro E, Silva MT,Lemonnier FA, De Sousa M. Increased hepatic iron in micelacking classical MHC class I molecules. Blood 2002;100:4239–4241.

38 Pantopoulos

38. Camaschella C, Roetto A, De Gobbi M. Juvenile hemochro-matosis. Semin Hematol 2002; 39:242–248.

39. Ponka P. Rare causes of hereditary iron overload. SeminHematol 2002; 39:249–262.

40. Kelly AL, Lunt PW, Rodrigues F, Berry PJ, Flynn DM,McKiernan PJ, Kelly DA, Mieli-Vergani G, Cox TM. Classifi-cation and genetic features of neonatal haemochromatosis: astudy of 27 affected pedigrees and molecular analysis of genesimplicated in iron metabolism. J Med Genet 2001; 38:599–610.

41. Gordeuk VR. African iron overload. Semin Hematol 2002;39:263–269.

42. Nittis T, Gitlin JD. The copper–iron connection: hereditaryaceruloplasminemia. Semin Hematol 2002; 39:282–289.

43. Andrews NC. Iron homeostasis: insights from genetics andanimal models. Nat Rev Genet 2000; 1:208–217.

44. Brock JH. The Biology of Iron. Chichester: John Wiley & SonsLtd, 1989:35–53.

45. Kaplan J. Strategy and tactics in the evolution of iron acquisi-tion. Semin Hematol 2002; 39:219–226.

46. Skikne B, Baynes RD. Iron Absorption. London: W.B. SaundersCompany Ltd, 1994:151–187.

47. McKie AT, Barrow D, Latunde-Dada GO, Rolfs A, Sager G,Mudaly E, Mudaly M, Richardson C, Barlow D, Bomford A,Peters TJ, Raja KB, Shirali S, Hediger MA, Farzaneh F,Simpson RJ. An iron-regulated ferric reductase associatedwith the absorption of dietary iron. Science 2001; 291:1755–1759.

48. Gunshin H, Mackenzie B, Berger UV, Gunshin Y, RomeroMF, Boron WF, Nussberger S, Gollan JL, Hediger MA. Clon-ing and characterization of a mammalian protein-coupledmetal-ion transporter. Nature 1997; 388:482–488.

49. Fleming MD, Trenor CCI, Su MA, Foernzler D, Beier DR,Dietrich WF, Andrews NC. Microcytic anaemia mice have amutation in Nramp2, a candidate iron transporter gene.Nat Genet 1997; 16:383–386.


50. Muckenthaler M, Roy CN, Custodio AO, Minana B, DeGraafJ, Montross LK, Andrews NC, Hentze MW. Regulatorydefects in liver and intestine implicate abnormal hepcidinand Cybrd1 expression in mouse hemochromatosis. NatGenet 2003; 34:102–107.

51. Canonne-Hergaux F, Gruenheid S, Ponka P, Gros P. Cellularand subcellular localization of the Nramp2 iron transporterin the intestinal brush border and regulation by dietary iron.Blood 1999; 93:4406–4417.

52. Donovan A, Brownlie A, Zhou Y, Shepard J, Pratt SJ,Moynihan J, Paw BH, Drejer A, Barut B, Zapata A, Law TC,Brugnara C, Lux SE, Pinkus GS, Pinkus JL, Kingsley PD,Palis J, Fleming MD, Andrews NC, Zon LI. Positional cloning ofzebrafish ferroportin1 identifies a conserved vertebrate ironexporter. Nature 2000; 403:776–781.

53. McKie AT, Marciani P, Rolfs A, Brennan K, Wehr K, BarrowD,Miret S, Bomford A, Peters TJ, Farzaneh F, Hediger MA,Hentze MW, Simpson RJ. A novel duodenal iron-regulatedtransporter IREG1, implicated in the basolateral transfer of ironto the circulation. Mol Cell 2000; 5:299–309.

54. Abboud S, Haile DJ. A novel mammalian iron-regulatedprotein involved in intracellular iron metabolism. J BiolChem 2000; 275:19906–19912.

55. Njajou OT, Vaessen N, Joosse M, Berghuis B, van DongenJW, Breuning MH, Snijders PJ, Rutten WP, Sandkuijl LA,Oostra BA, van Duijn CM, Heutink P. A mutation inSLC11A3 is associated with autosomal dominant hemochro-matosis. Nat Genet 2001; 28:213–214.

56. Vulpe CD, Kuo YM, Murphy TL, Cowley L, Askwith C, LibinaN, Gitschier J, Anderson GJ. Hephaestin, a ceruloplasminhomologue implicated in intestinal iron transport, is defectivein the sla mouse. Nat Genet 1999; 21:195–199.

57. Fleming MD, Romano MA, Su MA, Garrick LM, GarrickMD, Andrews NC. Nramp2 is mutated in the anemic Bel-grade (b) rat: evidence of a role for Nramp2 in endosomaliron transport. Proc Natl Acad Sci USA 1998; 95:1148–1153.

40 Pantopoulos

58. Su MA, Trenor CC III, Fleming JC, Fleming MD, AndrewsNC. The G185R mutation disrupts function of the iron trans-porter Nramp2. Blood 1998; 92:2157–2163.

59. Canonne-Hergaux F, Fleming MD, Levy JE, Gauthier S,Ralph T, Picard V, Andrews NC, Gros P. The Nramp2=DMT1DMT1 iron transporter is induced in the duodenum of micro-cytic anemia mk mice but is not properly targeted to theintestinal brush border. Blood 2000; 96:3964–3970.

60. Donovan A, Brownlie A, Dorschner MO, Zhou Y, Pratt SJ,Paw BH, Phillips RB, Thisse C, Thisse B, Zon LI. The zebra-fish mutant gene chardonnay (cdy) encodes divalent metaltransporter 1 (DMT1). Blood 2002; 100:4655–4659.

61. Yoshida K, Furihata K, Takeda S, Nakamura A, Yamamoto K,Morita H, Hiyamuta S, Ikeda S, Shimizu N, Yanagisawa N. Amutation in the ceruloplasmin gene is associated with sys-temic hemosiderosis in humans. Nat Genet 1995; 9:267–272.

62. Harris ZL, Takahashi Y, Miyajima H, Serizawa M,MacGillivray RT, Gitlin JD. Aceruloplasminemia: molecularcharacterization of this disorder of iron metabolism. Proc NatlAcad Sci USA 1995; 92:2539–2543.

63. Harris ZL, Durley AP, Man TK, Gitlin JD. Targeted gene dis-ruption reveals an essential role for ceruloplasmin in cellulariron efflux. Proc Natl Acad Sci USA 1999; 96:10812–10817.

64. Fleming RE, Sly WS. Ferroportin mutation in autosomaldominant hemochromatosis: loss of function, gain in under-standing. J Clin Invest 2001; 108:521–522.

65. Montosi G, Donovan A, Totaro A, Garuti C, Pignatti E,Cassanelli S, Trenor CC, Gasparini P, Andrews NC,Pietrangelo A. Autosomal-dominant hemochromatosis isassociated with a mutation in the ferroportin (SLC11A3)gene. J Clin Invest 2001; 108:619–623.

66. Roetto A, Merryweather-Clarke AT, Daraio F, Livesey K,Pointon JJ, Barbabietola G, Piga A, Mackie PH, Robson KJ,Camaschella C. A valine deletion of ferroportin 1: a commonmutation in hemochromastosis type 4. Blood 2002; 100:733–734.

67. Wallace DF, Pedersen P, Dixon JL, Stephenson P, Searle JW,Powell LW, Subramaniam VN. Novel mutation in ferroportin1


is associated with autosomal dominant hemochromatosis.Blood 2002; 100:692–694.

68. Devalia V, Carter K, Walker AP, Perkins SJ, Worwood M,May A, Dooley JS. Autosomal dominant reticuloendothelialiron overload associated with a 3-base pair deletion in theferroportin 1 gene (SLC11A3). Blood 2002; 100:695–697.

69. Cazzola M, Cremonesi L, Papaioannou M, Soriani N, KioumiA, Charalambidou A, Paroni R, Romtsou K, Levi S, FerrariM, Arosio P, Christakis J. Genetic hyperferritinaemia andreticuloendothelial iron overload associated with a three basepair deletion in the coding region of the ferroportin gene(SLC11A3). Br J Haematol 2002; 119:539–546.

70. Hayashi A, Wada Y, Suzuki T, Shimizu A. Studies on familialhypotransferrinemia: unique clinical course and molecularpathology. Am J Hum Genet 1993; 53:201–213.

71. Beutler E, Gelbart T, Lee P, Trevino R, Fernandez MA,Fairbanks VF. Molecular characterization of a case of atrans-ferrinemia. Blood 2000; 96:4071–4074.

72. Trenor CC, Campagna DR, Sellers VM, Andrews NC,Fleming MD. The molecular defect in hypotransferrinemicmice. Blood 2000; 96:1113–1118.

73. Poss KD, Tonegawa S. Heme oxygenase 1 is required formammalian iron reutilization. Proc Natl Acad Sci USA1997; 94:10919–10924.

74. Yachie A, Niida Y, Wada T, Igarashi N, Kaneda H, Toma T,Ohta K, Kasahara Y, Koizumi S. Oxidative stress causesenhanced endothelial cell injury in human heme oxygenase-1deficiency. J Clin Invest 1999; 103:129–135.

75. Kawashima A, Oda Y, Yachie A, Koizumi S, Nakanishi I.Heme oxygenase-1 deficiency: the first autopsy case. HumPathol 2002; 33:125–130.

76. Roy CN, Enns CA. Iron homeostasis: new tales from thecrypt. Blood 2000; 96:4020–4027.

77. Parkkila S, Waheed A, Britton RS, Feder JN, Tsuchihashi Z,Schatzman RC, Bacon BR, Sly WS. Immunohistochemistry ofHLA-H, the protein defective in patients with hereditaryhemochromatosis, reveals unique pattern of expression in

42 Pantopoulos

gastrointestinal tract. Proc Natl Acad Sci USA 1997;94:2534–2539.

78. Parkkila S, Parkkila AK, Waheed A, Britton RS, Zhou XY,Fleming RE, Tomatsu S, Bacon BR, Sly WS. Cell surfaceexpression of HFE protein in epithelial cells, macrophages,and monocytes. Haematologica 2000; 85:340–345.

79. Zoller H, Pietrangelo A, Vogel W, Weiss G. Duodenal metal-transporter (DMT-1, NRAMP-2) expression in patients withhereditary haemochromatosis. Lancet 1999; 353:2120–2123.

80. Zoller H, Koch RO, Theurl I, Obrist P, Pietrangelo A, MontosiG, Haile DJ, Vogel W, Weiss G. Expression of the duodenaliron transporters divalent-metal transporter 1 and ferroportin1 in iron deficiency and iron overload. Gastroenterology 2001;120:1412–1419.

81. FlemingRE,MigasMC, ZhouX, Jiang J, Britton RS,Brunt EM,Tomatsu S, Waheed A, Bacon BR, Sly WS. Mechanism ofincreased iron absorption in murine model of hereditary hemo-chromatosis: increasedduodenal expressionof the iron transpor-ter DMT1. Proc Natl Acad Sci USA 1999; 96:3143–3148.

82. Canonne-Hergaux F, Levy JE, Fleming MD, Montross LK,AndrewsNC,GrosP.Expressionof theDMT1(NRAMP2=DCT1)iron transporter in mice with genetic iron overload disorders.Blood 2001; 97:1138–1140.

83. Pietrangelo A, Rocchi E, Casalgrandi G, Rigo G, Ferrari A,Perini M, Ventura E, Cairo G. Regulation of transferrin,transferrin receptor, and ferritin genes in human duodenum.Gastroenterology 1992; 102:802–809.

84. Pietrangelo A, Casalgrandi G, Quaglino D, Gualdi R, ConteD, Milani S, Montosi G, Cesarini L, Ventura E, Cairo G.Duodenal ferritin synthesis in genetic hemochromatosis.Gastroenterology 1995; 108:208–217.

85. Cairo G, Recalcati S, Montosi G, Castrusini E, Conte D,Pietrangelo A. Inappropriately high iron regulatory proteinactivity in monocytes of patients with genetic hemochromato-sis. Blood 1997; 89:2546–2553.

86. Montosi G, Paglia P, Garuti C, Guzman CA, Bastin JM,Colombo MP, Pietrangelo A. Wild-type HFE protein normalizes


transferrin iron accumulation in macrophages from subjectswith hereditary hemochromatosis. Blood 2000; 96:1125–1129.

87. Trinder D, Olynyk JK, Sly WS, Morgan EH. Iron uptake fromplasma transferrin by the duodenum is impaired in the Hfeknockout mouse. Proc Natl Acad Sci USA 2002; 9:5622–5626.

88. Moura E, Noordermeer MA, Verhoeven N, Verheul AF, MarxJJ. Iron release from human monocytes after erythrophago-cytosis in vitro: an investigation in normal subjects andhereditary hemochromatosis patients. Blood 1998; 92:2511–2519.

89. Drakesmith H, Sweetland E, Schimanski L, Edwards J,Cowley D, Ashraf M, Bastin J, Townsend AR. The hemochro-matosis protein HFE inhibits iron export from macrophages.Proc Natl Acad Sci USA 2002; 12:15602–15607.

90. Finch C. Regulators of iron balance in humans. Blood 1994;84:1697–1702.

91. Fleming RE, Sly WS. Hepcidin: A putative iron-regulatoryhormone relevant to hereditary hemochromatosis and theanemia of chronic disease. Proc Natl Acad Sci USA 2001;98:8160–8162.

92. Ganz T. Hepcidin, a key regulator of iron metabolism andmediator of anemia of inflammation. Blood 2003; 102:783–788.

93. Krause A, Neitz S, Magert HJ, Schulz A, Forssmann WG,Schulz-Knappe P, Adermann K. LEAP-1, a novel highlydisulfide-bonded human peptide, exhibits antimicrobialactivity. FEBS Lett 2000; 480:147–150.

94. Park CH, Valore EV, Waring AJ, Ganz T. Hepcidin, a urinaryantimicrobial peptide synthesized in the liver. J Biol Chem2001; 276:7806–7810.

95. Pigeon C, Ilyin G, Courselaud B, Leroyer P, Turlin B, BrissotP, Loreal O. A new mouse liver-specific gene, encoding a pro-tein homologous to human antimicrobial peptide hepcidin, isoverexpressed during iron overload. J Biol Chem 2001;276:7811–7819.

96. Nicolas G, Bennoun M, Devaux I, Beaumont C, GrandchampB, Kahn A, Vaulont S. Lack of hepcidin gene expression and

44 Pantopoulos

severe tissue iron overload in upstream stimulatory factor 2(USF2) knockout mice. Proc Natl Acad Sci USA 2001;98:8780–8785.

97. Nicolas G, Bennoun M, Porteu A, Mativet S, Beaumont C,Grandchamp B, Sirito M, Sawadogo M, Kahn A, Vaulont S.Severe iron deficiency anemia in transgenic mice expres-sing liver hepcidin. Proc Natl Acad Sci USA 2002;99:4596–4601.

98. Frazer DM, Wilkins SJ, Becker EM, Vulpe CD, McKie AT,Trinder D, Anderson GJ. Hepcidin expression inversely cor-relates with the expression of duodenal iron transportersand iron absorption in rats. Gastroenterology 2002;123:835–844.

99. Nicolas G, Chauvet C, Viatte L, Danan JL, Bigard X, DevauxI, Beaumont C, Kahn A, Vaulont S. The gene encoding the ironregulatory peptide hepcidin is regulated by anemia, hypoxia,and inflammation. J Clin Invest 2002; 110:1037–1044.

100. Nemeth E, Valore EV, Territo M, Schiller G, Lichtenstein A,Ganz T. Hepcidin, a putative mediator of anemia of inflamma-tion, is a type II acute-phase protein. Blood 2003; 101:2461–2463.

101. Weinstein DA, Roy CN, Fleming MD, Loda MF, Wolfsdorf JI,Andrews NC. Inappropriate expression of hepcidin isassociated with iron refractory anemia: implications for theanemia of chronic disease. Blood 2002; 100:3776–3781.

102. Roetto A, Papanikolaou G, Politou M, Alberti F, Girelli D,Christakis J, Loukopoulos D, Camaschella C. Mutant antimi-crobial peptide hepcidin is associated with severe juvenilehemochromatosis. Nat Genet 2003; 33:21–22.

103. Bridle KR, Frazer DM, Wilkins SJ, Dixon JL, Purdie DM,Crawford DH, Subramaniam VN, Powell LW, Anderson GJ,Ram GA. Disrupted hepcidin regulation in HFE-associatedhaemochromatosis and the liver as a regulator of body ironhomoeostasis. Lancet 2003; 361:669–673.

104. Gehrke SG, Kulaksiz H, Herrmann T, Riedel HD, Bents K,VeltkampC, StremmelW.Expression of hepcidin in hereditaryhemochromatosis: evidence for a regulation in response to


serum transferrin saturation and non-transferrin-bound iron.Blood 2003; 102:371–376.

105. Ahmad KA, Ahmann JR, Migas MC, Waheed A, Britton RS,Bacon BR, Sly WS, Fleming RE. Decreased liver hepcidinexpression in the hfe knockout mouse. Blood Cells Mol Dis2002; 29:361–366.

106. Nicolas G, Viatte L, Lou DQ, Bennoun M, Beaumont C, KahnA, Andrews NC, Vaulont S. Constitutive hepcidin expressionprevents iron overload in a mouse model of hemochromatosis.Nat Genet 2003; 34:97–101.

107. Ponka P, Lok CN. The transferrin receptor: role in health anddisease. Int J Biochem Cell Biol 1999; 31:1111–1137.

108. Lawrence CM, Ray S, Babyonyshev M, Galluser R, BorhaniDW, Harrison SC. Crystal structure of the ectodomain ofhuman transferrin receptor. Science 1999; 286:779–782.

109. Gruenheid S, Canonne-Hergaux F, Gauthier S, Hackam DJ,Grinstein S, Gros P. The iron transport protein NRAMP2 isan integral membrane glycoprotein that colocalizes withtransferrin in recycling endosomes. J Exp Med 1999;189:831–841.

110. Kakhlon O, Cabantchik ZI. The labile iron pool: characteriza-tion, measurement, and participation in cellular processes.Free Radic Biol Med 2002; 33:1037–1046.

111. Epsztejn S, Kakhlon O, Glickstein H, BreuerW, Cabantchik I.Fluorescence analysis of the labile iron pool of mammaliancells. Anal Biochem 1997; 248:31–40.

112. Petrat F, Rauen U, de Groot H. Determination of the chelata-ble iron pool of isolated rat hepatocytes by digital fluores-cence microscopy using the fluorescent probe, phen greenSK. Hepatology 1999; 29:1171–1179.

113. Koenig M, Mandel JL. Deciphering the cause of Friedreichataxia. Curr Opin Neurobiol 1997; 7:689–694.

114. Puccio H, Koenig M. Friedreich ataxia: a paradigm formitochondrial diseases. Curr Opin Genet Dev 2002;12:272–277.

46 Pantopoulos

115. Levy JE, Jin O, Fujiwara Y, Kuo F, Andrews NC. Transferrinreceptor is necessary for development of erythrocytes and thenervous system. Nat Genet 1999; 21:396–399.

116. Parkkila S, Waheed A, Britton RS, Bacon BR, Zhou XY,Tomatsu S, Fleming RE, Sly WS. Association of the trans-ferrin receptor in human placenta with HFE, the proteindefective in hereditary hemochromatosis. Proc Natl AcadSci USA 1997; 94:13198–13202.

117. Feder JN, Penny DM, Irrinki A, Lee VK, Lebron JA, WatsonN, Tsuchihashi Z, Sigal E, Bjorkman PJ, Schatzman RC.The hemochromatosis gene product complexes with thetransferrin receptor and lowers its affinity for ligand binding.Proc Natl Acad Sci USA 1998; 95:1472–1477.

118. Roy CN, Penny DM, Feder JN, Enns CA. The hereditaryhemochromatosis protein, HFE, specifically regulates trans-ferrin-mediated iron uptake in HeLa cells. J Biol Chem1999; 274:9022–9028.

119. Salter-Cid L, Brunmark A, Li Y, Leturcq D, Peterson PA,Jackson MR, Yang Y. Transferrin receptor is negativelymodulated by the hemochromatosis protein HFE: implica-tions for cellular iron homeostasis. Proc Natl Acad Sci USA1999; 96:5434–5439.

120. Lebron JA, West AP Jr, Bjorkman PJ. The hemochromatosisprotein HFE competes with transferrin for binding to thetransferrin receptor. J Mol Biol 1999; 294:239–245.

121. Lebron JA, Bennet MJ, Vaughn DE, Chirino AJ, Snow PM,Mintier GA, Feder JN, Bjorkman PJ. Crystal structure ofthe hemochromatosis protein HFE and characterization ofits interaction with transferrin receptor. Cell 1998; 93:111–123.

122. Bennett MJ, Lebron JA, Bjorkman PJ. Crystal structure ofthe hereditary haemochromatosis protein HFE complexedwith transferrin receptor. Nature 2000; 403:46–53.

123. Riedel HD, Muckenthaler MU, Gehrke SG, Mohr I, BrennanK, Herrmann T, Fitscher BA, Hentze MW, Stremmel W. HFEdownregulates iron uptake from transferrin and inducesiron-regulatory protein activity in stably transfected cells.Blood 1999; 94:3915–3921.


124. RamalingamTS,West AP, Lebron JA,Nangiana JS,HoganTH,Enns CA, Bjorkman PJ. Binding to the transferrin receptoris required for endocytosis of HFE and regulation of ironhomeostasis. Nat Cell Biol 2000; 2:953–957.

125. Muckenthaler M, Richter A, Gunkel N, Riedel D, Polycarpou-Schwarz M, Hentze S, Falkenhahn M, Stremmel W, AnsorgeW, Hentze MW. Relationships and distinctions in iron regula-tory networks responding to interrelated signals. Blood 2003;101:3690–3698.

126. Wang J, Chen G, Pantopoulos K. The hemochromatosis pro-tein HFE induces an apparent iron-deficient phenotype inH1299 cells that is not corrected by co-expression of b2-microglobulin. Biochem J 2003; 370:891–899.

127. Kawabata H, Yang R, Hirama T, Vuong PT, Kawano S,Gombart AF, Koeffler HP. Molecular cloning of transferrinreceptor 2. A new member of the transferrin receptor-likefamily. J Biol Chem 1999; 274:20826–20832.

128. Fleming RE, Migas MC, Holden CC, Waheed A, Britton RS,Tomatsu S, Bacon BR, Sly WS. Transferrin receptor 2:continued expression in mouse liver in the face of iron over-load and in hereditary hemochromatosis. Proc Natl AcadSci USA 2000; 97:2214–2219.

129. Griffiths WJ, Cox TM. Co-localization of the mammalianhemochromatosis gene product (HFE) and a newly identifiedtransferrin receptor (TfR2) in intestinal tissue and cells.J Histochem Cytochem 2003; 51:613–624.

130. Vogt TM, Blackwell AD, Giannetti AM, Bjorkman PJ, EnnsCA. Heterotypic interactions between transferrin receptorand transferrin receptor 2. Blood 2003; 101:2008–2014.

131. Camaschella C, Roetto A, Cali A, De Gobbi M, Garozzo G,Carella M, Majorano N, Totaro A, Gasparini P. The geneTFR2 is mutated in a new type of haemochromatosismapping to 7q22. Nat Genet 2000; 25:14–15.

132. Fleming RE, Ahmann JR, Migas MC, Waheed A, Koeffler HP,KawabataH,BrittonRS,BaconBR,SlyWS.Targetedmutagen-esis of the murine transferrin receptor-2 gene produces hemo-chromatosis. Proc Natl Acad Sci USA 2002; 99:10653–10658.

48 Pantopoulos

133. Roetto A, Totaro A, Piperno A, Piga A, Longo F, Garozzo G,Cali A, De Gobbi M, Gasparini P, Camaschella C. Newmutations inactivating transferrin receptor 2 in hemochro-matosis type 3. Blood 2001; 97:2555–2560.

134. Girelli D, Bozzini C, Roetto A, Alberti F, Daraio F, ColombariR, Olivieri O, Corrocher R, Camaschella C. Clinical andpathologic findings in hemochromatosis type 3 due to a novelmutation in transferrin receptor 2 gene. Gastroenterology2002; 122:1295–1302.

135. Mattman A, Huntsman D, Lockitch G, Langlois S, BuskardN, Ralston D, Butterfield Y, Rodrigues P, Jones S, Porto G,Marra M, De Sousa M, Vatcher G. Transferrin receptor 2(TfR2) and HFE mutational analysis in non-C282Y iron over-load: identification of a novel TfR2 mutation. Blood 2002;100:1075–1077.

136. Inman RS, Wessling-Resnick M. Characterization of trans-ferrin-independent iron transport in K562 cells. Unique prop-erties provide evidence for multiple pathways of iron uptake.J Biol Chem 1993; 268:8521–8528.

137. Kennard ML, Richardson DR, Gabathuler R, Ponka P,Jefferies WA. A novel iron uptake mechanism mediated byGPI-anchored human p97. EMBO J 1995; 14:4178–4186.

138. Garrick LM, Dolan KG, Romano MA, Garrick MD. Non-transferrin-bound iron uptake in Belgrade and normal raterythroid cells. J Cell Physiol 1999; 178:349–358.

139. Barisani D, Cairo G, Ginelli E, Marozzi A, Conte D. Nitricoxide reduces nontransferrin-bound iron transport in HepG2cells. Hepatology 1999; 29:464–470.

140. Mwanjewe J, Hui BK, Coughlin MD, Grover AK. Treatmentof PC12 cells with nerve growth factor increases iron uptake.Biochem J 2001; 357:881–886.

141. Ludwiczek S, Aigner E, Theurl I, Weiss G. Cytokine mediatedregulation of iron transport in human monocytic cells. Blood2003; 101:4148–4154.

142. Kjeldsen L, Cowland JB, Borregaard N. Human neutrophilgelatinase-associated lipocalin and homologous proteins inrat and mouse. Biochim Biophys Acta 2000; 1482:272–283.


143. GoetzDH,HolmesMA,BorregaardN,BluhmME,RaymondKN,Strong RK. The neutrophil lipocalin NGAL is a bacteriostaticagent that interfereswith siderophore-mediated ironacquisition.Mol Cell 2002; 10:1033–1043.

144. Yang J, Goetz D, Li JY, Wang W, Mori K, Setlik D, Du T,Erdjument-Bromage H, Tempst P, Strong R, Barasch J. Aniron delivery pathway mediated by a lipocalin. Mol Cell2002; 10:1045–1056.

145. Kaplan J. Mechanisms of cellular iron acquisition: anotheriron in the fire. Cell 2002; 111:603–606.

146. Harrison PM, Arosio P. The ferritins: molecular properties,iron storage function and cellular regulation. BiochimBiophys Acta 1996; 1275:161–203.

147. Torti FM, Torti SV. Regulation of ferritin genes and protein.Blood 2002; 99:3505–3516.

148. Radisky DC, Kaplan J. Iron in cytosolic ferritin can berecycled through lysosomal degradation in human fibro-blasts. Biochem J 1998; 336:201–5.

149. Takagi H, Shi D, Ha Y, Allewell NM, Theil EC. Localizedunfolding at the junction of three ferritin subunits. Amechanism for iron release? J Biol Chem 1998; 273:18685–18688

150. Ferreira C, Bucchini D, Martin ME, Levi S, Arosio P,Grandchamp B, Beaumont C. Early embryonic lethality ofH ferritin gene deletion in mice. J Biol Chem 2000;275:3021–3024.

151. Ferreira C, Santambrogio P, Martin ME, Andrieu V, FeldmannG, Henin D, Beaumont C. H ferritin knockout mice: a model ofhyperferritinemia in the absence of iron overload. Blood 2001;98:525–532.

152. Levi S, Corsi B, Bosisio M, Invernizzi R, Volz A, Sanford D,Arosio P, Drysdale J. A human mitochondrial ferritin encodedby an intronless gene. J Biol Chem 2001; 276:24437–24440.

153. Drysdale J, Arosio P, Invernizzi R, Cazzola M, Volz A, CorsiB, Biasiotto G, Levi S. Mitochondrial ferritin: a new playerin iron metabolism. Blood Cells Mol Dis 2002; 29:376–383.

50 Pantopoulos

154. Cazzola M, Invernizzi R, Bergamaschi G, Levi S, Corsi B,Travaglino E, Rolandi V, Biasiotto G, Drysdale J, Arosio P.Mitochondrial ferritin expression in erythroid cells frompatients with sideroblastic anemia. Blood 2003; 101:1996–2000.

155. Roetto A, Bosio S, Gramaglia E, Barilaro MR, Zecchina G,Camaschella C. Pathogenesis of hyperferritinemia cataractsyndrome. Blood Cells Mol Dis 2002; 29:532–535.

156. Kato J, Fujikawa K, Kanda M, Fukuda N, Sasaki K,Takayama T, Kobune M, Takada K, Takimoto R, HamadaH, Ikeda T, Niitsu Y. A mutation, in the iron-responsiveelement of H ferritin mRNA, causing autosomal dominantiron overload. Am J Hum Genet 2001; 69:191–197.

157. Curtis AR, Fey C, Morris CM, Bindoff LA, Ince PG, ChinneryPF, Coulthard A, Jackson MJ, Jackson AP, McHale DP, HayD, Barker WA, Markham AF, Bates D, Curtis A, Burn J.Mutation in the gene encoding ferritin light polypeptidecauses dominant adult-onset basal ganglia disease. NatGenet 2001; 28:350–354.

158. Rouault TA. Iron on the brain. Nat Genet 2001; 28:299–300.

159. Crompton DE, Chinnery PF, Fey C, Curtis AR, Morris CM,Kierstan J, Burt A, Young F, Coulthard A, Curtis A, IncePG, Bates D, Jackson MJ, Burn J. Neuroferritinopathy: awindow on the role of iron in neurodegeneration. Blood CellsMol Dis 2002; 29:522–531.

160. Hentze MW, Kuhn LC. Molecular control of vertebrate ironmetabolism: mRNA-based regulatory circuits operated byiron, nitric oxide and oxidative stress. Proc Natl Acad SciUSA 1996; 93:8175–8182.

161. Rouault T, Harford JB. Translational control of ferritin synth-esis. In: Sonenberg N, Hershey JWB, Mathews MB, eds.Translational Control of Gene Expression. Cold SpringHarbor, NY: Cold Spring Harbor Laboratory Press,2000:655–670.

162. Johansson HE, Theil EC. Iron-response element (IRE) struc-ture and combinatorial RNA regulation. In: Templeton DM,ed. Molecular and Cellular Iron Transport. New York, Basel:Marcel Dekker Inc, 2002:237–253.


163. Theil EC, Eisenstein RS. Combinatorial mRNA regulation:iron regulatory proteins and iso-iron responsive elements(iso-IREs). J Biol Chem 2000; 275:40659–40662.

164. Casey JL, Koeller DM, Ramin VC, Klausner RD, Harford JB.Iron regulation of transferrin receptor mRNA levels requiresiron-responsive elements and a rapid turnover determinantin the 30 untranslated region of the mRNA. EMBO J 1989;8:3693–3699.

165. Binder R, Horowitz JA, Basilion JP, Koeller DM, KlausnerRD, Harford JB. Evidence that the pathway of transferrinreceptor mRNA degradation involves an endonucleolytic clea-vage within the 30 UTR and does not involve poly(A) tailshortening. EMBO J 1994; 13:1969–1980.

166. Theil EC. Regulation of ferritin and transferrin receptormRNAs. J Biol Chem 1990; 265:4771–4774.

167. Muckenthaler M, Gray NK, Hentze MW. IRP-1 binding toferritin mRNA prevents the recruitment of the small riboso-mal subunit by the cap-binding complex eIF4F. Mol Cell1998; 2:383–388.

168. Paraskeva E, Gray NK, Schlager B, Wehr K, Hentze MW.Ribosomal pausing and scanning arrest as mechanisms oftranslational regulation from cap-distal iron-responsiveelements. Mol Cell Biol 1999; 19:807–816.

169. Melefors O, Goossen B, Johansson HE, Stripecke R, GrayNK, Hentze MW. Translational control of 5-aminolevulinatesynthase mRNA by iron-responsive elements in erythroidcells. J Biol Chem 1993; 268:5974–5978.

170. Gray NK, Pantopoulos K, Dandekar T, Ackrell BAC, HentzeMW. Translational regulation of mammalian and drosophilacitric acid cycle enzymes via iron-responsive elements. ProcNatl Acad Sci USA 1996; 93:4925–4930.

171. Schalinske KL, Chen OS, Eisenstein RS. Iron differentiallystimulates translation of mitochondrial aconitase and ferritinmRNAs in mammalian cells. Implications for iron regulatoryproteins as regulators of mitochondrial citrate utilization.J Biol Chem 1998; 273:3740–3746.

52 Pantopoulos

172. Wardrop SL, Richardson DR. The effect of intracellular ironconcentration and nitrogen monoxide on Nramp2 expressionand non-transferrin-bound iron uptake. Eur J Biochem 1999;263:41–49.

173. Gunshin H, Allerson CR, Polycarpou-Schwarz M, Rofts A,Rogers JT, Kishi F, Hentze MW, Rouault TA, Andrews NC,Hediger MA. Iron-dependent regulation of the divalent metalion transporter. FEBS Lett 2001; 509:309–316.

174. Caltagirone A,Weiss G, Pantopoulos K.Modulation of cellulariron metabolism by hydrogen peroxide. Effects of H2O2 on theexpression and function of iron-responsive element-contain-ing mRNAs in B6 fibroblasts. J Biol Chem 2001; 276:19738–19745.

175. Hubert N, Hentze MW. Previously uncharacterized isoformsof divalent metal transporter (DMT)-1: implications for regu-lation and cellular function. Proc Natl Acad Sci USA 2002;99:12345–12350.

176. Beaumont C, Leneuve P, Devaux I, Scoazec J-Y, Berthier M,Loiseau M-N, Grandchamp B, Bonneau D. Mutation in theiron responsive element of the L ferritin mRNA in a familywith dominant hyperferritinaemia and cataract. Nat Genet1995; 11:444–446.

177. Cremonesi L, Paroni R, Foglieni B, Galbiati S, Fermo I,Soriani N, Belloli S, Ruggeri G, Biasiotto G, Cazzola M, Fer-rari F, Ferrari M, Arosio P. Scanning mutations of the 50UTRUTR regulatory sequence of l-ferritin by denaturing high-performance liquid chromatography: identification of newmutations. Br J Haematol 2003; 121:173–179.

178. Cazzola M, Bergamaschi G, Tonon L, Arbustini E, Grasso M,Vercesi E, Baroi G, Bianchi PE, Cairo G, Arosio P. Hereditaryhyperferritinemia-cataract syndrome: relationship betweenphenotypes and specific mutations in the iron-responsive ele-ment of ferritin light-chain mRNA. Blood 1997; 90:814–821.

179. Allerson CR, Cazzola M, Rouault TA. Clinical severity andthermodynamic effects of iron-responsive element mutationsin hereditary hyperferritinemia-cataract syndrome. J BiolChem 1999; 274:26439–26447.


180. Levi S, Girelli D, Perrone F, Pasti M, Beaumont C, CorrocherR, Albertini A, Arosio P. Analysis of ferritins in lymphoblas-toid cell lines and in the lens of subjects with hereditaryhyperferritinemia-cataract syndrome. Blood 1998; 91:4180–4187.

181. Eisenstein RS. Iron regulatory proteins and the molecularcontrol of mammalian iron metabolism. Annu Rev Nutr2000; 20:627–662.

182. Pantopoulos K, Hentze MW. Nitric oxide, oxygen radicals andiron metabolism. In: Ignarro L, ed. Nitric Oxide. San Diego,CA: Academic Press, 2000:293–313.

183. Frishman D, Hentze MW. Conservation of aconitase residuesrevealed by multiple sequence analysis. Implications forstructure=function relationships. Eur J Biochem 1996;239:197–200.

184. Robbins AH, Stout CD. The structure of aconitase. Proteins1989; 5:289–312.

185. Samaniego F, Chin J, Iwai K, Rouault TA, Klausner RD.Molecular characterization of a second iron-responsiveelement binding protein, iron regulatory protein 2. J BiolChem 1994; 269:30904–30910.

186. Erlitzki R, Long JC, Theil EC. Multiple, conserved iron-responsive elements in the 30-untranslated region of transfer-rin receptor mRNA enhance binding of iron regulatoryprotein 2. J Biol Chem 2002; 277:42579–42587.

187. Rouault TA. Post-transcriptional regulation of human ironmetabolism by iron regulatory proteins. Blood Cells Mol Dis2002; 29:309–314.

188. Kim H-Y, Klausner RD, Rouault TA. Translationalrepressor activity is equivalent and is quantitatively pre-dicted by in vitro RNA binding for two iron-responsiveelement-binding proteins, IRP1 and IRP2. J Biol Chem1995; 270:4983–4986.

189. Butt J, Kim HY, Basilion JP, Cohen S, Iwai K, PhilpottCC, Altschul S, Klausner RD, Rouault TA. Differences inthe RNA binding sites of iron regulatory proteins and

54 Pantopoulos

potential target diversity. Proc Natl Acad Sci USA 1996;93:4345–4349.

190. Henderson BR, Menotti E, Bonnard C, Kuhn LC. Optimalsequence and structure of iron-responsive elements. Selec-tion of RNA stem-loops with high affinity for iron regulatoryfactor. J Biol Chem 1994; 269:17481–17489.

191. Kim S, Ponka P. Control of transferrin receptor expressionvia nitric oxide-mediated modulation of iron-regulatoryprotein 2. J Biol Chem 1999; 274:33035–33042.

192. Kim S, Ponka P. Nitrogen monoxide-mediated control offerritin synthesis: implications for macrophage iron homeos-tasis. Proc Natl Acad Sci USA 2002; 99:12214–12219.

193. DeRusso PA, Philpott CC, Iwai K, Mostowski HS, KlausnerRD, Rouault TA. Expression of a constitutive mutant of ironregulatory protein 1 abolishes iron homeostasis in mamma-lian cells. J Biol Chem 1995; 270:15451–15454.

194. Wang J, Pantopoulos K. Conditional de-repression of ferritinsynthesis in cells expressing a constitutive IRP1 mutant. MolCell Biol 2002; 22:4638–4651.

195. Kennedy MC, Mende-Mueller L, Blondin GA, Beinert H.Purification and characterization of cytosolic aconitase frombeef liver and its relationship to the iron-responsive elementbinding protein. Proc Natl Acad Sci USA 1992; 89:11730–11734.

196. Brazzolotto X, Gaillard J, Pantopoulos K, Hentze MW, MoulisJM. Human cytoplasmic aconitase (Iron regulatory protein 1)is converted into its [3Fe–4S] form by hydrogen peroxide invitro but is not activated for iron-responsive element binding.J Biol Chem 1999; 274:21625–21630.

197. Gegout V, Schlegl J, Schlager B, Hentze MW, Reinbolt J,Ehresmann B, Ehresmann C, Romby P. Ligand-inducedstructural alterations in human iron regulatory protein-1revealed by protein footprinting. J Biol Chem 1999;274:15052–15058.

198. Brazzolotto X, Timmins P, Dupont Y, Moulis JM. Structuralchanges associated with switching activities of the human ironregulatory protein 1. J Biol Chem 2002; 277:11995–12000.


199. CairoG,PietrangeloA. Iron regulatory proteins in pathobiology.Biochem J 2000; 352:241–250.

200. Fillebeen C, Pantopoulos K. Redox control of iron regulatoryproteins. Redox Rep 2002; 7:15–22.

201. Weiss G, Goossen B, Doppler W, Fuchs D, Pantopoulos K,Werner-Felmayer G, Wachter H, Hentze MW. Translationalregulation via iron-responsive elements by the nitric oxi-de=NO-synthase pathway. EMBO J 1993; 12:3651–3657.

202. Drapier JC, Hirling H, Wietzerbin J, Kaldy P, Kuhn LC.Biosynthesis of nitric oxide activates iron regulatory factorin macrophages. EMBO J 1993; 12:3643–3649.

203. Pantopoulos K, Weiss G, Hentze MW. Nitric oxide and oxida-tive stress (H2O2) control mammalian iron metabolism bydifferent pathways. Mol Cell Biol 1996; 16:3781–3788.

204. Pantopoulos K, Hentze MW. Rapid responses to oxidativestress mediated by iron regulatory protein. EMBO J 1995;14:2917–2924.

205. Pantopoulos K, Mueller S, Atzberger A, Ansorge W, Strem-mel W, Hentze MW. Differences in the regulation of ironregulatory protein 1 (IRP1) by extra- and intracellular oxida-tive stress. J Biol Chem 1997; 272:9802–9808.

206. Pantopoulos K, Hentze MW. Activation of iron regulatoryprotein-1 by oxidative stress in vitro. Proc Natl Acad SciUSA 1998; 95:10559–10563.

207. Brown NM, Anderson SA, Steffen DW, Carpenter TB,Kennedy MC, Walden WE, Eisenstein RS. Novel role of phos-phorylation in Fe–S cluster stability revealed by phosphomi-metic mutations at Ser-138 of iron regulatory protein 1. ProcNatl Acad Sci USA 1998; 95:15235–15240.

208. Guo B, Yu Y, Leibold EA. Iron regulates cytoplasmic levels ofa novel iron-responsive element-binding protein withoutaconitase activity. J Biol Chem 1994; 269:24252–24260.

209. Henderson BR, Kuhn LC. Differential modulation of theRNA-binding proteins IRP1 and IRP2 in response to iron.IRP2 inactivation requires translation of another protein.J Biol Chem 1995; 270:20509–20515.

56 Pantopoulos

210. Guo B, Phillips JD, Yu Y, Leibold EA. Iron regulates theintracellular degradation of iron regulatory protein 2 by theproteasome. J Biol Chem 1995; 270:21645–21651.

211. Iwai K, Klausner RD, Rouault TA. Requirements for iron-regulated degradation of the RNA binding protein, iron regu-latory protein 2. EMBO J 1995; 14:5350–5357.

212. Iwai K, Drake SK, Wehr NB, Weissman AM, LaVaute T,Minato N, Klausner RD, Levine RL, Rouault TA. Iron-dependent oxidation, ubiquitination, and degradation ofiron regulatory protein 2: implications for degradationof oxidized proteins. Proc Natl Acad Sci USA 1998;95:4924–4928.

213. Goessling LS, Mascotti DP, Thach RE. Involvement of hemein the degradation of iron-regulatory protein 2. J Biol Chem1998; 273:12555–12557.

214. Yamanaka K, Ishikawa H, Megumi Y, Tokunaga F, Kanie M,Rouault TA, Morishima I, Minato N, Ishimori K, Iwai K.Identification of the ubiquitin-protein ligase that recognizesoxidized IRP2. Nat Cell Biol 2003; 5:336–340.

215. Pantopoulos K, Hentze MW. Nitric oxide signaling to iron-regulatory protein (IRP): direct control of ferritin mRNAtranslation and transferrin receptor mRNA stability intransfected fibroblasts. Proc Natl Acad Sci USA 1995;92:1267–1271.

216. Recalcati S, Taramelli D, Conte D, Cairo G. Nitricoxide-mediated induction of ferritin synthesis in J774macrophages by inflammatory cytokines: role of selectiveiron regulatory protein-2 downregulation. Blood 1998;91:1059–1066.

217. Bouton C, Oliveira L, Drapier J-C. Converse modulation ofIRP-1 and IRP-2 by immunological stimuli in murine RAW264.7 macrophages. J Biol Chem 1998; 273:9403–9408.

218. Kim S, Ponka P. Effects of interferon-gamma and lipopolysac-charide on macrophage iron metabolism are mediated bynitric oxide-induced degradation of iron regulatory protein2. J Biol Chem 2000; 275:6220–6226.


219. Gehring N, Hentze MW, Pantopoulos K. Menadione-inducedoxidative stress leads to inactivation of both RNA-bindingand aconitase activities of iron regulatory protein-1. J BiolChem 1999; 274:6219–6225.

220. HansonES, FootLM,LeiboldEA.Hypoxia post-translationallyactivates iron-regulatory protein 2. J Biol Chem 1999;274:5047–5052.

221. Hanson ES, Leibold EA. Regulation of iron regulatory protein1 during hypoxia and hypoxia=reoxygenation. J Biol Chem1998; 273:7588–7593.

222. LaVaute T, Smith S, Cooperman S, Iwai K, Land W,Meyron-Holtz E, Drake SK, Miller G, Abu-Asab M, TsokosM, Switzer R III, Grinberg A, Love P, Tresser N, RouaultTA. Targeted deletion of the gene encoding iron regulatoryprotein-2 causes misregulation of iron metabolism and neuro-degenerative disease in mice. Nat Genet 2001; 27:209–214.

223. Chan RYY, Seiser C, Schulman HM, Kuhn LC, Ponka P.Regulation of transferrin receptor mRNA expression.Distinct regulatory features in erythroid cells. Eur J Biochem1994; 220:683–692.

224. Lok CN, Ponka P. Identification of an erythroid activeelement in the transferrin receptor gene. J Biol Chem 2000;275:24185–24190.

225. Lok CN, Ponka P. Identification of a hypoxia response ele-ment in the transferrin receptor gene. J Biol Chem 1999;274:24147–24152.

226. Tacchini L, Bianchi L, Bernelli-Zazzera A, Cairo G. Transfer-rin receptor induction by hypoxia. HIF-1-mediated transcrip-tional activation and cell-specific post-transcriptionalregulation. J Biol Chem 1999; 274:24142–24146.

227. Rolfs A, Kvietikova I, Gassmann M, Wenger RH. Oxygen-regulated transferrin expression is mediated by hypoxia-inducible factor-1. J Biol Chem 1997; 272:20055–20062.

228. Mukhopadhyay CK, Mazumder B, Fox PL. Role of hypoxia-inducible factor-1 in transcriptional activation of cerulo-plasmin by iron deficiency. J Biol Chem 2000; 275:21048–21054.

58 Pantopoulos

229. Bourgeade M-F, Silbermann F, Kuhn LC, Testa U, Peschle C,Memet S, Thang MN, Besancon F. Post-transcriptional regu-lation of transferrin receptor mRNA by IFNg. Nucl Acids Res1992; 20:2997–3003.

230. Mulero V, Brock JH. Regulation of iron metabolism in mur-ine J774 macrophages: role of nitric oxide-dependent and-independent pathways following activation with gammainterferon and lipopolysaccharide. Blood 1999; 94:2383–2389.

231. Weiss G, Bogdan C, Hentze MW. Pathways for the regulationof macrophage iron metabolism by the anti-inflammatorycytokines IL-4 and IL-13. J Immunol 1997; 158:420–425.

232. Yang F, Liu XB, Quinones M, Melby PC, Ghio A, Haile DJ.Regulation of reticuloendothelial iron transporter MTP1(Slc11a3) by inflammation. J Biol Chem 2002; 277:9786–39791.

233. Miller LL, Miller SC, Torti SV, Tsuji Y, Torti FM. Iron-independent induction of ferritin H chain by tumor necrosisfactor. Proc Natl Acad Sci USA 1991; 88:4946–4950.

234. Wei Y, Miller SC, Tsuji Y, Torti SV, Torti FM. Interleukin 1induces ferritin heavy chain in human muscle cells. BiochemBiophys Res Commun 1990; 169:289–296.

235. Kobune M, Kohgo Y, Kato J, Miyazaki E, Niitsu Y. Interleu-kin-6 enhances hepatic transferrin uptake and ferritinexpression in rats. Hepatology 1994; 19:1468–1475.

236. Tsuji Y, Ayaki H, Whitman SP, Morrow CS, Torti SV, TortiFM. Coordinate transcriptional and translational regulationof ferritin in response to oxidative stress. Mol Cell Biol2000; 20:5818–5827.

237. Pietsch EC, Chan JY, Torti FM, Torti SV. Nrf2 mediates theinduction of ferritin H in response to xenobiotics and cancerchemopreventive dithiolethiones. J Biol Chem 2003;278:2361–2369.

238. Weiss G. Iron and anemia of chronic disease. Kidney IntSuppl 1999; 69:S12–S17.

239. Weiss G. Iron and immunity: a double-edged sword. Eur JClin Invest 2002; 32(suppl 1):70–78.


240. Walker EM Jr, Walker SM. Effects of iron overload on theimmune system. Ann Clin Lab Sci 2000; 30:354–365.

241. Gordeuk VR, Ballou S, Lozanski G, Brittenham GM.Decreased concentrations of tumor necrosis factor-alpha insupernatants of monocytes from homozygotes for hereditaryhemochromatosis. Blood 1992; 79:1855–1860.

242. Weiss G, Fuchs D, Hausen A, Reibnegger G, Werner ER, Wer-ner-Felmayer G, Wachter H. Iron modulates interferon-gamma effects in the human myelomonocytic cell line THP-1.Exp Hematol 1992; 20:605–610.

243. Weiss G, Werner-Felmayer G, Werner ER, Grunewald K,Wachter H, Hentze MW. Iron regulates nitric oxide synthaseactivity by controlling nuclear transcription. J Exp Med 1994;180:969–976.

244. Romagnani S. Th1=Th2 cells. Inflamm Bowel Dis 1999;5:285–294.

245. Thuma PE, Weiss G, Herold M, Gordeuk VR. Serum neop-terin, interleukin-4, and interleukin-6 concentrations incerebral malaria patients and the effect of iron chelationtherapy. Am J Trop Med Hyg 1996; 54:164–168.

246. Menacci A, Cenci E, Boelaert JR, Bucci P, Mosci P, Fed’Ostiani C, Bistoni F, Romani L. Iron overload alters innateand T helper cell responses to Candida albicans in mice.J Infect Dis 1997; 175:1467–1476.

247. Weiss G, Thuma PE, Mabeza G, Werner ER, Herold M,Gordeuk VR. Modulatory potential of iron chelation therapyon nitric oxide formation in cerebral malaria. J Infect Dis1997; 175:226–230.

248. Hunter HN, Fulton DB, Ganz T, Vogel HJ. The solutionstructure of human hepcidin, a peptide hormone with antimi-crobial activity that is involved in iron uptake and hereditaryhemochromatosis. J Biol Chem 2002; 277:37597–37603.

60 Pantopoulos

2

Erythropoietin and Erythropoiesis

ERIC METZEN and WOLFGANG JELKMANN

Institute of Physiology, University of Luebeck,Ratzeburger Allee, Luebeck, Germany

INTRODUCTION

Red blood cells fulfill a vital task by carrying oxygen (O2) fromthe lung to the other organs. Anemic persons suffer from tissuehypoxia. They present with paleness, shortness of breath,angina pectoris, physical weakness, fatigue, and possibly,other symptoms such as headache. Severe cases may requirered cell transfusions from blood donors.

The concentration of red blood cells in the circulatory sys-tem is normally maintained constant. Red blood cells are con-tinually produced and, after about 120 days in the bloodstream, removed by macrophages in bone marrow, liver, andspleen. In adult humans, 2–3� 1011 reticulocytes derive everyday in a process that involves a series of steps of proliferationand differentiation of hemopoietic stem and progenitor cells in

61

the bone marrow. The basal rate of erythropoiesis may rise 10-fold on hypoxic stress. Both the continual and the stress-induced increase in the rate of erythropoiesis are under thecontrol of the glycoprotein hormone erythropoietin (Epo),which is an essential viability and growth factor for theerythrocytic progenitors. Table 1 lists normal values of impor-tant erythrocytic elements in the blood of adult humans.

CHEMICAL STRUCTURE OF Epo

Human Epo is a glycoprotein hormone encoded by a singlecopy gene located on the long arm (q11–q22) of chromosome7 (1–3). It contains five exons, which encode a 193 amino acidprehormone that includes a leader peptide of 27 residues.Taking into account that the carboxy-terminal arginineexpected from the nucleotide sequence is also cleaved priorto secretion, circulating human Epo is composed of 165 aminoacids with a peptide mass of 18.2 kDa (4). The polypeptide

Table 1 Normal Values of Erythrocytic Elements in Blood ofAdult Humans

Parameter Value Unit

Erthrocytes , 4.8 (4.2–5.5) 1012 L�1 blood< 5.3 (4.4–6.3) 1012 L�1 blood

Reticulocytes 0.1 (0.05–0.2) 1012 L�1 bloodHematocrit , 0.42 (0.37–0.47) vol=vol

< 0.47 (0.40–0.54) vol=volHemoglobin , 140 (120–160) g L�1 blood

< 160 (140–180) g L�1 bloodMCV 85 (80–96) flMCH 30 (28–34) pgFerritin 100 (40–160) mg L�1 serumTransferrin 35 (20–45) %saturation

Erythropoietin 10 (6–32) U L�1 serum

Values are medians, with approximate range in parentheses.Abbreviations: MCV, mean corpuscular volume; MCH, mean corpuscularhemoglobin.

62 Metzen and Jelkmann

chain mediates binding to the Epo receptor and suffices for invitro stimulation of erythropoiesis (5). The most importantamino acids interacting with the two different binding sitesof the dimeric Epo receptor have been identified (6). Ithas also been established that Epo—similar to other class Icytokines—forms a compact globular structure consisting offour (A–D) a-helical bundles, which are connected by two longcrossover loops (AB and CD) and one short loop (BC), yieldingan up–up–down–down configuration (7).

For in vivo activity of the hormone, N-linked carbohy-drate chains are necessary. Circulating human Epo possessesthree N-linked (at the asparagines 24, 38, and 83) and oneO-linked (at serine 126) oligosaccharide chains which repre-sent about 40% of the total mass (30.4 kDa) of the molecule(5). It is assumed that hundreds of different endogenous iso-forms exist due to structural permutations of four glycans ofEpo (8). The isoelectric point values of the major fractionsare in the pH range 2–4 (9,10). Studies of the electrophoreticmobility of serum Epo have shown that there are intraindivi-dual diurnal variations and interindividual differences,besides disease-dependent alterations, in the mobility of Epoon electrophoresis (11). Because highly branched N-glycansstabilize the conformation of glycoproteins, the in vivo activityof Epo increases with the degree of branching (12). The term-inal monosaccharide of the N-glycans is generally sialic acid,which is attached to galactose. Epo isoforms having fewer sia-lic acid residues have a greater affinity for the Epo receptorbut a shorter survival in circulation (13). Truly, however,the understanding of the sites and mechanisms of the removalof Epo from circulation is still incomplete (14). Evidence sug-gests that Epo receptor-mediated clearance of the hormone isthe dominant pathway. To a minor degree, Epo may befiltered in the renal glomeruli, and once it is desialylated,metabolized in the liver.

In viewing recombinant human Epo (rhEpo), differentpreparations are available for therapy. The conventionalforms, epoetin alfa and epoetin beta, are produced by Chinesehamster ovary (CHO) cells transfected with the authentichuman Epo gene (15,16). Recently, a new rhEpo product

Erythropoietin and Erythropoiesis 63

(epoetin omega) produced in baby hamster kidney (BHK) cellscame on the market (17–19). The epoetins and endogenoushuman Epo are probably identical with respect to their peptidecore of 165 amino acids and the four glycosylation sites (20,21).However, some differences exist in the molecular compositionof the N-glycans (for review see Ref. 14). Note that epoetinomega exhibits further differences in the structure of its gly-cans compared to the CHO cell-derived epoetins alfa and beta(22). An alternative drug is darbepoetin alfa, a hyperglyco-sylated product, which contains two novel N-linkedcarbohydrate side chains. The amino acid sequence of darbe-poetin alfa differs from that of human Epo at five positionsincluding two additional oligosaccharide attachment sites atasparagine residues in positions 30 and 88 (23). Darbepoetinalfa has an increased molecular mass, sialic acid content,and a negative charge when compared to endogenous Epo orthe epoetins.

SITES AND CONTROL OF Epo PRODUCTION

Epo is mainly of hepatic origin in the fetal stage (24). Afterbirth, the kidneys become the primary production site. Basedon experimental studies in nonhuman mammals, it is thoughtthat almost all circulating Epo originates from peritubularinterstitial fibroblasts in the cortex of the kidneys (25,26).In the liver, the Epo mRNA expressing cells are mainly hepa-tocytes (27). In addition, Epo mRNA is detectable in minoramounts in spleen, lung, testis, and brain. The Epo producedin brain is of special interest (28). It seems to act as a neuro-trophic and neuroprotective factor separate from the systemicaction of Epo on hemopoietic cells. First clinical trials havebeen reported in which rhEpo was administered followingischemic stroke in humans (29).

There are no significant Epo stores in kidney and liver,but the synthesis of the hormone is acutely stimulated whenthe O2 availability decreases. Thus, plasma Epo increaseswithin 1 hr after the initiation of hypoxia. Peak values arereached after 1–2 days in humans. Epo production on acute


hypoxic stress is dynamic, i.e., initially very large and thenmore moderate despite persisting hypoxia (for references,see Ref. 30). Apart from being controlled by the hemoglobinconcentration or, more precisely, the O2 capacity of the blood,Epo synthesis is stimulated when the arterial O2 tension islowered or when the O2 affinity of the blood is increased.Excessive Epo and red cell production due to hypoxemia is amajor pathogenetic factor in the development of chronicmountain sickness. Figure 1 illustrates the negative feedbackcircuit of Epo production based on the O2 supply to thetissues.

Tissue hypoxia stimulates Epo gene expression. Based oncell culture work, major progress has been made in under-standing the nature of the O2 sensor in control of the expres-sion of the Epo gene and other genes that are induced by

Figure 1 Feedback regulation of erythropoiesis linking the rate ofproliferation and differentiation of myeloid erythrocytic progenitors(burst- and colony-forming units-erythroid: BFU-E and CFU-E) tothe renal (and hepatic) O2 supply. (Modified from Ref. 109.)


hypoxia (for review, see Refs. 31, 32). Several cis-acting regu-latory DNA sequences neighbouring the Epo gene have beenidentified. Most studies have focused on the O2-dependenttranscriptional enhancer located 30 to the poly (A)-adenylationsite of the Epo gene. The key regulator is the transcriptionfactor hypoxia-inducible factor 1 (HIF-1), a dimeric proteincomposed of two different subunits, the 100–120kDa HIF-1aand the 91–94kDa HIF-1b (33,34). Under hypoxic conditions,the HIF-1 complex binds to the conserved consensus sequence(A=G)CGTG within the hypoxia response element (HRE)present in O2-controlled genes such as those for vascularendothelial growth factor (VEGF), several glucose transpor-ters, and virtually all glycolytic enzymes. HIF-1a is theO2-labile partner, while HIF-1b is a constitutive nuclear pro-tein that functions in a variety of transcriptional systemswith alternative dimerization partners, e.g., the dioxin recep-tor. HIF-1a possesses a central oxygen-dependent proteolyticdegradation domain (ODD) and two transactivation domains(TAD). In cell culture, HIF-1a is accumulated in an exponen-tial way with lowered O2 tension (35). In the presence of O2,HIF-1a is hydroxylated at the proline residues 402 (36)and 564 in the ODD (37,38). This reaction is catalyzed by spe-cific HIF-1a prolyl hydroxylase domain (PHD) possessingenzymes that belong to the group of a-oxoglutarate andFe (II)-dependent dioxygenases (39,40). Furthermore, anasparaginyl hydroxylase, also termed factor inhibiting HIF-1 (FIH-1) (41,42), has been identified, which hydroxylatesasparagine 803 in the HIF-1a carboxy-terminal TAD, therebylowering the ability of HIF-1a to bind the transcriptional coac-tivator p300=CBP (43,44). Of the three well-characterizedhuman PHDs (1–3), PHD1 is exclusively present in thenucleus, PHD2—as well as FIH-1—is mainly located in thecytoplasm, whereas PHD3 can be found in both cytoplasmand nucleus (45). The existence of fourth human PHD asso-ciated with the endoplasmic reticulum (PHD-4) has beendescribed recently (46). Prolyl hydroxylated HIF-1a is immedi-ately bound by the von-Hippel–Lindau gene product (pVHL),which forms a complex with the E3 ubiquitin ligase (47,48).As a result, prolyl hydroxylated HIF-1a is immediately


polyubiquitinated and degraded by the proteasome (49). Onlyunder hypoxic conditions, HIF-1a is enabled to translocate tothe nucleus and to heterodimerize with HIF-1b. Besides O2

tension, phosphorylation reactionsmediated by the phosphati-dylinositol 3-kinase and the MAPK kinase pathways can mod-ulate HIF-1a stability and its transactivation activity (50). O2

sensing and HIF-1-dependent gene expression are not linkedto an active mitochondrial respiratory chain (51,52).

Compared with the HIF-1 sensitive Epo gene enhancer,the Epo gene promoter has received much less attention.Some evidence has been provided to assume that the tran-scription factor GATA-2 inhibits Epo gene expression by bind-ing to the Epo gene promoter under normoxic conditions(53,54). In addition, the Epo gene promoter and the 50 flank-ing region contain binding sites for nuclear factor kB(NFkB) (55). Both GATA-2 and NFkB are probably involvedin the inhibition of Epo gene expression in inflammation.The proinflammatory cytokines interleukin-1 (IL-1) andtumor necrosis factor a (TNF-a) activate GATA-2 (56) andNFkB (56,57). It has been proposed that IL-1 and TNF-amay contribute to the anemia of chronic disease by activatingGATA-2 and NFkB and thereby suppressing the activity ofthe Epo gene promoter (56). However, the role of NFkB isdiscussed controversially (58).

Truly, the role of HIF-1, GATA-2, and NFkB in the invivo control of Epo gene transcription in the kidney still needsto be explored. Reportedly, Epo mRNA expression in kidneycells follows an all or nothing fashion (59). Previous attemptshave failed to establish renal cell cultures for the study of O2-dependent Epo synthesis. Isolated renal hypoxia, induced byrenal artery constriction, is a minor stimulus to Epo produc-tion, when compared to the strong reaction on systemichypoxia (60). The assumption that the central nervous systemmodulates renal Epo production is supported both by dataobtained following manipulations of hypothalamic functions(61,62) and the physiological circadian rhythms of the plasmaEpo level (63,64). An overview of humoral factors that maymodulate the renal and the hepatic production of Epo is givenelsewhere (30,65). Important mediators include thyroid


hormones (T3 and T4), growth hormone, insulin-like growthfactors, steroidal hormones, prostaglandins, and angiotensinII. Note, too, that the concentration of circulating Epo willnot only depend on the rate of the hormone’s synthesis butalso on its degradation (66). Endogenous Epo was earliershown to be cleared with a half-life of about 2 hr in humans(67). The longer half-life of rhEpo (about 6–8 hr) is probablydue to the more complete glycosylation of the recombinantproducts (68).

ASSAY OF CIRCULATING Epo ANDINTERPRETATION OF RESULTS

Epo activities are commonly expressed in units (U).Historically, 1 U was defined as a dose eliciting the sameerythropoiesis-stimulating effect in experimental animals as5 mmol cobaltous chloride. International standards have beenestablished of impure human urinary Epo (second IRP) (69)and rhEpo (specific activity 130,000 U=mg fully glycosylatedprotein) (70). Radioimmunoassays or enzyme-linked immuno-sorbent assays (ELISA) are commonly used for routine mea-surements of Epo. The concentration of immunoreactiveEpo in healthy nonanemic humans is 6–32 U=L (about10�11mol=L). Possible indications for assay of circulating Epoin clinical routine include the differential diagnosis ofpolycythemias and anemias, the follow-up of paraneoplasticEpo production, and the election of anemic patients for rhEpotherapy. Assay-specific reference intervals need to be takeninto account when commercial assay kits are used. For eachEpo assay, a description of the reference interval (normalrange) should be provided along with information onanalytical performance and the selection of subjects in bloodsampling. Details of the different assay procedures andpossible pitfalls have been described elsewhere (71). Majorparameters that need to be considered in the interpretationof Epo immunoassay data are summarized in Table 2 .

Insufficient Epo production is the primary cause of ane-mia associated with chronic renal failure (72). The lack of


Epo in renal disease is primarily caused by the destruction ofthe Epo-producing cells and the inhibition of Epo synthesisdue to metabolic acidosis and the action of uremia toxinsand proinflammatory cytokines. Apart from renal disease, arelative lack of circulating Epo has been assessed in chronicinflammation, malignancy, and AIDS (73–75). Note that theconcentration of Epo in blood cannot be evaluated in absoluteterms but only in relation to the blood hemoglobin concentra-tion (66,76). Beguin et al.(77) have earlier suggested to calcu-late the so-called observed=predicted log [Epo] ratio (O=Pratio) for each plasma sample from patients, with the pre-dicted level being estimated from a reference group ofpatients with uncomplicated anemia, such as iron deficiency.

ACTION OF Epo ON HEMOPOIETIC CELLS

Blood cell production is sustained by a relatively small poolof self-renewing stem cells in the bone marrow and otherhemopoietic organs. Stem cells are capable of producing pro-genitors with more restricted potential. In the erythrocyticlineage, the most primitive progenitor is the burst-formingunit erythroid (BFU-E), which gives rise to several colony-forming units-erythroid (CFU-E). In the presence of Epo, each

Table 2 Technical and Biological Parameters to Be Considered inthe Interpretation of Epo Values

Principle of assay procedure (competitive vs. two-site immunoassays)and specificity of antibodies (monoclonal vs. polyclonal)

Type of calibrator (second IRP or rDNA derived international standardvs. inhouse standards)

Assay-specific reference intervalAbsence of matrix effects, circulating antibodies, and soluble Epo

receptors (control by recovery tests)Dependence of the level of circulating Epo on the blood

hemoglobin concentration (log O=P ratio)Dynamic response of levels of circulating Epo to acute hypoxic stressCircadian fluctuations of Epo levelsInfluence of proliferative activity of the erythron on Epo metabolism

Modified from Ref. 71.


CFU-E can proliferate and differentiate into small colonies of8–64 nucleated polychromatic normoblasts. This process isinfluenced by a number of other positively or negatively actinghemopoietic growth factors and cytokines (Fig. 2).

Figure 2 Action of erythropoietin (Epo) and selected hemopoieticgrowth factors and inhibitory cytokines on the major differentiationandmaturation steps in erythropoiesis. Erythrocytic progenitors arederived from multipotent stem cells (CFU-GEMM: colony-formingunit producing progenitors of granulocytes, erythrocytes, mono-cytes, and megakaryocytes). In the presence of Epo, committed ery-throcytic progenitors give rise to an increasing number of progeny.SCF, stem cell factor; G-CSF, granulocyte colony-stimulating factor;GM-CSF, granulocyte=monocyte colony-stimulating factor; IL,interleukin; Tpo, thrombopoietin; EDF, erythroid differentiationfactor; IGF, insulin-like growth factor; TGF-b, transforming growthfactor-b; IFNs, interferons a, b, and g; TNF-a, tumor necrosisfactor-a. (Modified from Ref. 110.)


CFU-Es carry the highest density of Epo receptors ontheir surface and are the most Epo sensitive cells. Themature Epo receptor is a 484 amino acid glycoprotein belong-ing to the cytokine class I receptor superfamily (78,79), whichalso comprises the receptors for thrombopoietin, G-CSF, GM-CSF, and various interleukins. All of these receptors possessa single hydrophobic transmembrane sequence, a variablecytoplasmic domain and an extracellular domain with con-served cysteines, and a WSXWS motif, which is importantfor ligand binding and signal transduction (79). Two of themembrane-spanning Epo receptor molecules form a dimerto which one Epo molecule binds. Epo binding induces a con-formational change and a tighter connection of the two recep-tor molecules (7,80,81). In turn, two Jak2 tyrosine kinasemolecules, which are loosely associated to the cytoplasmicregion of the Epo receptor molecules, are activated (81,82).As a result, several tyrosine residues of the Epo receptorare phosphorylated and become docking sites forsignaling proteins containing SH2 (SRC homology 2)domains. Various signal transduction pathways are then acti-vated, including phosphatidylinositol 3-kinase=Akt, Jak2=Stat5, MAPK kinase, and protein kinase C (83,84). The actionof Epo is terminated by the action of the hematopoietic cellphosphatase SHP-1, which initiates Jak2 dephosphorylation(85,86). Mutations of the cytoplasmic carboxy-terminalregions of the Epo receptor and the loss of domains associatedwith SHP-1 may lead to familial erythrocytosis (87). Apartfrom being expressed by erythrocytic tissue, Epo receptormRNA has been demonstrated in nonhemopoietic tissuessuch as endothelium, neuronal cells, and placenta (78). Thefunction of nonhemopoietic receptors is presently investi-gated by tissue-specific gene-knockout techniques (88). Inaddition, fragments of the Epo receptor have been demon-strated as soluble receptor in plasma (89,90).

The primary mechanism by which Epo maintainserythropoiesis is the prevention of programmed cell death(91,92). The normally low concentration of Epo enables onlya small percentage of CFU-E to survive and to proli-ferate while the remainder undergoes apoptosis. When the


concentration of Epo rises in blood, as in anemia, an increas-ing number of progenitor cells escape from apoptosis and pro-liferate. Epo is absolutely essential for the viability,proliferation, and differentiation in the erythrocytic lineage.In addition to its action on erythrocytic progenitors, Epo sti-mulates the growth and maturation of the morphologicallyidentifiable proerythroblasts and normoblasts. The time fromthe proerythroblast to the reticulocyte is 4–6 days andinvolves 4–5 cell divisions. The early, basophilic, normoblast(or erythroblast) is characterized by the absence of nucleoli.At the stage of the late, polychromatic, normoblast, hemoglo-bin is accumulated. Then the nucleus becomes pyknotic and iseventually excluded from the cell. Reticulocytes contain resi-dual RNA. Significant reticulocytosis becomes apparent about3–4 days after an acute increase in plasma Epo.

PATHOPHYSIOLOGICAL ASPECTS

Erythropoiesis is a finely tuned process, which is responsiblefor the maintenance of red cell mass, red blood cell numbers,hemoglobin concentration, and O2 delivery to the tissues.Pathophysiologically, red cell mass may be increased abovenormal (erythrocytosis) or be reduced (anemia).

Primary erythrocytosis results from an excessive activityof the erythrocytic progenitors. Polycythemia vera is charac-terized by a clonal expansion of hematopoietic progenitors ofall myeloid lineages due to an acquired mutation in a singlehematopoietic progenitor (93). Primary familial and congeni-tal polycythemia, also known as autosomal dominant erythro-cytosis, is a less common cause of primary erythrocytosis (87).Secondary erythrocytosis results from an abnormal increasein the rate of the production of Epo due to hypoxemia (resi-dence at high altitude, cyanotic heart disease) or an increasein blood oxygen affinity (hemoglobinopathy, 2,3-bisphospho-glycerate deficiency) (94). The main risks of erythrocytosisinclude heart failure, myocardial infarction, seizures, periph-eral thromboembolic events, and pulmonary embolism. Inorder to prevent hemodynamic and rheological complications,


the patients may have to undergo repeated phlebotomies.Unfortunately, specific erythropoiesis-inhibiting drugs arenot yet available, although Epo receptor antagonists areunder development (95). With some success, clinical trialshave been carried out in which angiotensin-convertingenzyme inhibitors or angiotensin II antagonists were admi-nistered to reduce the rate of erythropoiesis in patients suffer-ing from erythrocytosis after renal transplantation (96) andpersons suffering from chronic mountain sickness (97). Themechanism of action of these drugs is not well understoodwith respect to red cell production.

Anemia may be due to abnormal blood losses, impairedred blood cell production, or increased hemolysis. In patientsshowing symptoms of an absolute or relative lack of Epo,therapy with rhEpo is beneficial (72,98). RhEpo therapyaims at maintaining the patients’ hemoglobin values abovethe transfusion trigger, increasing their exercise tolerance,and improving quality-of-life parameters (99). In contrast,red blood cell transfusions are indicated to avoid or abolishlife-threatening consequences of severe anemia. However,rhEpo therapy is expensive. Therefore, attempts are beingmade to develop alternative drugs for stimulation of erythro-poiesis. First, less expensive generic rhEpo preparations areexpected to come to the market in the near future, when thepresent recombinant drugs will no longer be protected bypatent. Second, both peptide (100) and nonpeptide(101,102) Epo mimetics have been discovered, which arestructurally unrelated to Epo but stimulate the proliferationand differentiation of erythrocytic progenitors in vitro and,partially, in vivo. Third, iron chelators may prove useful toincrease endogenous Epo synthesis (103–105). Fourth, inhi-bitors of hemopoietic cell phosphatase SHP-1 have beendeveloped, which prolong the action of Epo on its target cells(106). Finally, preclinical trials of Epo gene transfer havebeen reported (107,108). However, before Epo gene therapycan be explored in humans, the stability, tissue specificityand regulatory mechanisms of the transgenes requirefurther investigation.


REFERENCES

1. Powell JS, Berkner KL, Lebo RV, Adamson JW. Humanerythropoietin gene: high level expression in stably trans-fected mammalian cells and chromosome localization. ProcNatl Acad Sci USA 1986; 83:6465–6469.

2. Watkins PC, Eddy R, Hoffman N, Stanislovitis P, Beck AK,Galli J, Vellucci V, Gusella JF, Shows TB. Regional assign-ment of the erythropoietin gene to human chromosome region7pter-q22. Cytogenet Cell Genet 1986; 42:214–218.

3. Law ML, Cai GY, Lin FK, Wei Q, Huang SZ, Hartz JH, MorseH, Lin CH, Jones C, Kao FT. Chromosomal assignment of thehuman erythropoietin gene and its DNA polymorphism. ProcNatl Acad Sci USA 1986; 83:6920–6924.

4. Recny MA, Scoble HA, Kim Y. Structural characterization ofnatural human urinary and recombinant DNA-derivederythropoietin. Identification of des-arginine 166 erythropoietin. J Biol Chem 1987; 262:17156–17163.

5. Dordal MS, Wang FF, Goldwasser E. The role of carbohydratein erythropoietin action. Endocrinology 1985; 116:2293–2299.

6. Syed RS, Reid SW, Li C, Cheetham JC, Aoki KH, Liu B, ZhanH, Osslund TD, Chirino AJ, Zhang J, Finer-Moore J, Elliott S,Sitney K, Katz BA, Matthews DJ, Wendoloski JJ, Egrie J,Stroud RM. Efficiency of signalling through cytokine recep-tors depends critically on receptor orientation. Nature 1998;395:511–516.

7. Cheetham JC, Smith DM, Aoki KH, Stevenson JL, HoeffelTJ, Syed RS, Egrie J, Harvey TS. NMR structure of humanerythropoietin and a comparison with its receptor boundconformation. Nat Struct Biol 1998; 5:861–866.

8. Storring PL, Tiplady RJ, Gaines-Das RE, Rafferty B, MistryYG. Lectin-binding assays for the isoforms of human erythro-poietin: comparison of urinary and four recombinant erythro-poietins. J Endocrinol 1996; 150:401–412.

9. Dorado M, Langton AA, Brandan NC, Espada J. Electro-phoretic behavior of erythropoietin in polyacrylamide gel.Biochem Med 1972; 6:238–245.


10. Tam RC, Coleman SL, Tiplady RJ, Storring PL, Cotes PM.Comparisons of human, rat and mouse erythropoietins byisoelectric focusing: differences between serum and urinaryerythropoietins. Br J Haematol 1991; 79:504–511.

11. Wide L, Bengtsson C. Molecular charge heterogeneity ofhuman serum erythropoietin. Br J Haematol 1990; 76:121–127.

12. Takeuchi M, Inoue N, Strickland TW, Kubota M, Wada M,Shimizu R, Hoshi S, Kozutsumi H, Takasaki S, Kobata A.Relationship between sugar chain structure and biologicalactivity of recombinant human erythropoietin produced inChinese hamster ovary cells. Proc Natl Acad Sci USA 1989;86:7819–7822.

13. Egrie JC, Grant JR, Gillies D, Aoki KH, Strickland TW. Therole of carbohydrate on the biological activity of erythropoietin.Glycoconjugate J 1993; 10:263 [Abstr].

14. Jelkmann W. The enigma of the metabolic fate of circulatingerythropoietin (Epo) in view of the pharmacokinetics of therecombinant drugs rhEpo and NESP. Eur J Haematol 2002;69:265–274.

15. Lin FK, Suggs S, Lin CH, Browne JK, Smalling R, Egrie JC,Chen KK, Fox GM, Martin F, Stabinsky Z, Badrawi SM, LaiPH, Goldwasser E. Cloning and expression of the humanerythropoietin gene. Proc Natl Acad Sci USA 1985; 82:7580–7584.

16. Davis JM,ArakawaT, Strickland TW, YphantisDA. Character-ization of recombinant human erythropoietin produced in Chi-nese hamster ovary cells. Biochemistry 1987; 26:2633–2638.

17. Acharya VN, Sinha DK, Almeida AF, Pathare AV. Effectof low dose recombinant human omega erythropoietin(rHuEPO) on anaemia in patients on hemodialysis. J AssocPhysicians India 1995; 43:539–542.

18. Bren A, Kandus A, Varl J, Buturovic J, Ponikvar R, KvederR, Primozic S, Ivanovich P. A comparison between epoetinomega and epoetin alfa in the correction of anemia in hemo-dialysis patients: a prospective, controlled crossover study.Artif Organs 2002; 26:91–97.


19. Sikole A, Spasovski G, Zafirov D, Polenakovic M. Epoetinomega for treatment of anemia in maintenance hemodialysispatients. Clin Nephrol 2002; 57:237–245.

20. Sasaki H, Bothner B, Dell A, Fukuda M. Carbohydrate struc-ture of erythropoietin expressed in Chinese hamster ovarycells by a human erythropoietin cDNA. J Biol Chem 1987;262:12059–12076.

21. Imai N, Kawamura A, Higuchi M, Oh-eda M, Orita T,Kawaguchi T, Ochi N. Physicochemical and biological compar-ison of recombinant human erythropoietin with human urin-ary erythropoietin. J Biochem (Tokyo) 1990; 107:352–359.

22. Skibeli V, Nissen-Lie G, Torjesen P. Sugar profiling provesthat human serum erythropoietin differs from recombinanthuman erythropoietin. Blood 2001; 98:3626–3634.

23. Egrie JK, Browne JK. Development and characterization ofnovel erythropoiesis stimulating protein (NESP). Br J Cancer2001; 84:3–10.

24. Dame C, Fahnenstich H, Freitag P, Hofmann D, Abdul NT,Bartmann P, Fandrey J. Erythropoietin mRNA expression inhuman fetal and neonatal tissue. Blood 1998; 92:3218–3225.

25. Bachmann S, Le Hir M, Eckardt KU. Co-localization oferythropoietin mRNA and ecto-50-nucleotidase immunoreac-tivity in peritubular cells of rat renal cortex indicates thatfibroblasts produce erythropoietin. J Histochem Cytochem1993; 41:335–341.

26. Maxwell PH, Osmond MK, Pugh CW, Heryet A, Nicholls LG,Tan CC, Doe BG, Ferguson DJ, Johnson MH, Ratcliffe PJ.Identification of the renal erythropoietin-producing cellsusing transgenic mice. Kidney Int 1993; 44:1149–1162.

27. Maxwell PH, Ferguson DJ, Nicholls LG, Iredale JP, PughCW, Johnson MH, Ratcliffe PJ. Sites of erythropoietinproduction. Kidney Int 1997; 51:393–401.

28. Marti HH, Wenger RH, Rivas LA, Straumann U, Digicaylio-glu M, Henn V, Yonekawa Y, Bauer C, Gassmann M.Erythropoietin gene expression in human, monkey and mur-ine brain. Eur J Neurosci 1996; 8:666–676.


29. Ehrenreich H, Hasselblatt M, Dembrowski C, Cepek L,Lewczuk P, Stiefel M, Rustenbeck H-H, Breiter N, JacobS, Knerlich F, Bohn M, Poser W, Ruther E, Kochen M,Gefeller O, Gleiter C, Wessel TC, De Ryck M, Itri L, PrangeH, Cerami A, Brines M, Siren A-L. Erythropoietin therapyfor acute stroke is both safe and beneficial. Mol Med 2002;8:495–505.

30. Jelkmann W. Erythropoietin: structure, control of produc-tion, and function. Physiol Rev 1992; 72:449–489.

31. Ratcliffe P. From erythropoietin to oxygen: hypoxia-induciblefactor hydroxylases and the hypoxia signal pathway. BloodPurif 2002; 20:445–450.

32. WengerRH.Cellular adaptation tohypoxia:O2-sensingproteinhydroxylases, hypoxia-inducible transcription factors, andO2-regulated gene expression. FASEB J 2002; 16:1151–1162.

33. Wang GL, Semenza GL. General involvement of hypoxia-inducible factor 1 in transcriptional response to hypoxia. ProcNatl Acad Sci USA 1993; 90:4304–4308.

34. Wang GL, Jiang BH, Rue EA, Semenza GL. Hypoxia-inducible factor 1 is a basic-helix-loop-helix-PAS heterodimerregulated by cellular O2 tension. Proc Natl Acad Sci USA1995; 92:5510–5514.

35. Jiang BH, Semenza GL, Bauer C, Marti HH. Hypoxia-inducible factor 1 levels vary exponentially over a physiologically relevant range of O2 tension. Am J Physiol 1996;271:C1172–C1180.

36. Masson N, William C, Maxwell PH, Pugh CW, Ratcliffe PJ.Independent function of two destruction domains in hypoxia-inducible factor-a chains activated by prolyl hydroxylation.EMBO J 2001; 20:5197–5206.

37. Jaakkola P, Mole DR, Tian YM, Wilson MI, Gielbert J, Gas-kell SJ, Kriegsheim A, Hebestreit HF, Mukherji M, SchofieldCJ, Maxwell PH, Pugh CW, Ratcliffe PJ. Targeting of HIF-ato the Von Hippel–Lindau ubiquitylation complex by O2-regu-lated prolyl hydroxylation. Science 2001; 292:468–472.

38. Ivan M, Kondo K, Yang H, Kim W, Valiando J, Ohh M, SalicA, Asara JM, Lane WS, Kaelin WG Jr. HIFa targeted for


VHL-mediated destruction by proline hydroxylation: implica-tions for O2 sensing. Science 2001; 292:464–468.

39. Epstein ACR, Gleadle JM, McNeill LA, Hewitson KS,O’Rourke J, Mole DR, Mukherji M, Metzen E, Wilson MI,Dhanda A, Tian YM, Masson N, Hamilton DL, Jaakkola P,Barstead R, Hodgkin J, Maxwell PH, Pugh CW, SchofieldCJ, Ratcliffe PJ. C. elegans EGL-9 and mammalian homologsdefine a family of dioxygenases that regulate HIF by prolylhydroxylation Cell. 2001; 107:43–54.

40. Bruick RK, McKnight SL. A conserved family of prolyl-4-hydroxylases that modify HIF. Science 2001; 294:1337–1340.

41. Mahon PC, Hirota K, Semenza GL. FIH-1: a novel proteinthat interacts with HIF-1a and VHL to mediate repressionof HIF-1 transcriptional activity. Genes Dev 2001; 15:2675–2686.

42. Hewitson KS, McNeill LA, Riordan MV, Tian YM, BullockAN, Welford RW, Elkins JM, Oldham NJ, Bhattacharya S,Gleadle JM, Ratcliffe PJ, Pugh CW, Schofield CJ. Hypoxiainducible factor (HIF) asparagine hydroxylase is identical toFactor Inhibiting HIF (FIH) and is related to the cupin struc-tural family. J Biol Chem 2002; 277:26351–26355.

43. Lando D, Peet DJ, Gorman JJ, Whelan DA, Whitelaw ML,Bruick RK. FIH-1 is an asparaginyl hydroxylase enzyme thatregulates the transcriptional activity of hypoxia-induciblefactor. Genes Dev 2002; 16:1466–1471.

44. McNeill LA, Hewitson KS, Claridge TD, Seibel JF, HorsfallLE. Hypoxia-inducible factor asparaginyl hydroxylase (FIH-1) catalyses hydroxylation at the beta-carbon of asparagine-803. Biochem J 2003; 367:571–575.

45. Metzen E, Berchner-Pfannschmidt U, Stengel P, MarxsenJH, Stolze I, Klinger M, Huang WQ, Wotzlaw C,Hellwig-Buergel T, Jelkmann W, Acker H, Fandrey J. Intra-cellular localisation of human HIF-1alpha hydroxylases:implications for oxygen sensing. J Cell Science 2003; 116:1319–1326.

46. Oehme F, Ellinghaus P, Kolkhof P, Smith TJ, RamakrishnanS, Hutter J, Schramm M, Flamme I. Overexpression of PH-4,a novel putative proline 4-hydroxylase, modulates activity of


hypoxia-inducible transcription factors. Biochem Biophys ResCommun 2002; 296:343–349.

47. Maxwell PH, Wiesener MS, Chang GW, Clifford SC, VauxEC, Cockman ME, Wykoff CC, Pugh CW, Maher ER, RatcliffePJ. The tumour suppressor protein VHL targets hypoxia-inducible factors for oxygen-dependent proteolysis. Nature1999; 399:271–275.

48. Ohh M, Park CW, Ivan M, Hoffman MA, Kim TY, HuangLE, Pavletich N, Chau V, Kaelin WG. Ubiquitination ofhypoxia-inducible factor requires direct binding to the b-domain of the von Hippel–Lindau protein. Nat Cell Biol2000; 2:423–427.

49. Salceda S, Caro J. Hypoxia-inducible factor 1a (HIF-1a) pro-tein is rapidly degraded by the ubiquitin-proteasome systemunder normoxic conditions. Its stabilization by hypoxiadepends on redox-induced changes. J Biol Chem 1997;272:22642–22647.

50. Minet E, Michel G, Mottet D, Raes M, Michiels C. Transduc-tion pathways involved in hypoxia-inducible factor-1 phos-phorylation and activation. Free Radic Biol Med 2001;31:847–855.

51. Vaux EC, Metzen E, Yeates KM, Ratcliffe PJ. Regulation ofhypoxia-inducible factor is preserved in the absence of a func-tioning mitochondrial respiratory chain. Blood 2001; 98:296–302.

52. Srinivas V, Leshchinsky I, Sang N, King MP, Minchenko A,Caro J. Oxygen sensing and HIF-1 activation does not requirean active mitochondrial respiratory chain electron-transferpathway. J Biol Chem 2001; 276:21995–21998.

53. Imagawa S, Izumi T, Miura Y. Positive and negative regula-tion of the erythropoietin gene. J Biol Chem 1994; 269:9038–9044.

54. Imagawa S, Yamamoto M, Miura Y. Negative regulation ofthe erythropoietin gene expression by the GATA transcrip-tion factors. Blood 1997; 89:1430–1439.

55. Lee HS, Lin JJ, Kung HF, Huang PL, Lee L, Huang PL. Thehuman erythropoietin-encoding gene contains a CAAT box,


TATA boxes and other transcriptional regulatory elements inits 50 flanking region. Gene 1993; 128:227–236.

56. La Ferla K, Reimann C, JelkmannW, Hellwig-Burgel T. Inhi-bition of erythropoietin gene expression signaling involvesthe transcription factors GATA-2 and NF-kB. FASEB J2002; 16:1811–1813.

57. Jelkmann W, Hellwig-Buergel T. Tumor necrosis factor p55receptor (TNF-RI) mediates the in vitro inhibition of hepaticerythropoietin production. Exp Hematol 1999; 27:224–228.

58. Figueroa YG, Chan AK, Ibrahim R, Tang Y, Burow ME, AlamJ, Scandurro AB, Beckman BS. NF-kB plays a key role inhypoxia-inducible factor-1-regulated erythropoietin geneexpression. Exp Hematol 2002; 30:1419–1427.

59. Koury ST, Koury MJ, Bondurant MC, Caro J, Graber SE.Quantitation of erythropoietin-producing cells in kidneys ofmice by in situ hybridization: correlation with hematocrit,renal erythropoietin mRNA, and serum erythropoietinconcentration. Blood 1989; 74:645–651.

60. Pagel H, Jelkmann W, Weiss C. A comparison of the effects ofrenal artery constriction and anemia on the production oferythropoietin. Pflugers Arch 1988; 413:62–66.

61. Halvorsen S. The central nervous system in regulation oferythropoiesis. Acta Haematol 1966; 35:65–79.

62. Mirand EA. Extra-renal and renal control of erythropoietinproduction. Ann N Y Acad Sci 1968; 149:94–106.

63. Wide L, Bengtsson C, Birgegard G. Circadian rhythm of ery-thropoietin in human serum. Br J Haematol 1989; 72:85–90.

64. Klausen T, Poulsen TD, Fogh-Andersen N, Richalet JP,Nielsen OJ, Olsen NV. Diurnal variations of serum erythro-poietin at sea level and altitude. Eur J Appl Physiol OccupPhysiol 1996; 72:297–302.

65. Fisher JW. Erythropoietin: physiology and pharmacologyupdate. Proc Soc Exp Biol Med 2003; 216:358–369.

66. Cazzola M, Guarnone R, Cerani P, Centenara E, Rovati A,Beguin Y. Red blood cell precursor mass as an independent


determinant of serum erythropoietin level. Blood 1998;91:2139–2145.

67. Embury SH, Garcia JF, Mohandas N, Pennathur DR, ClarkMR. Effects of oxygen inhalation on endogenous erythropoie-tin kinetics, erythropoiesis, and properties of blood cells insickle-cell anemia. N Engl J Med 1984; 311:291–295.

68. Weikert S, Papac D, Briggs J, Cowfer D, Tom S, Gawlitzek M,Lofgren J, Mehta S, Chisholm V, Modi N, Eppler S, Carroll K,Chamow S, Peers D, Berman P, Krummen L. EngineeringChinese hamster ovary cells to maximize sialic acid contentof recombinant glycoproteins. Nature Biotechnol 1999;17:1116–1121.

69. Annable L, Cotes PM, Mussett MV. The second internationalreference preparation of erythropoietin, human, urinary, forbioassay. Bull World Health Organ 1972; 47:99–112.

70. Storring PL, Gaines DR. The International Standard forRecombinant DNA-derived Erythropoietin: collaborativestudy of four recombinant DNA-derived erythropoietins andtwo highly purified human urinary erythropoietins. J Endo-crinol 1992; 134:459–484.

71. Jelkmann W. Biochemistry and assays of Epo. In: JelkmannW, ed. Erythropoietin: Molecular Biology and Clinical Use.Johnson City, TN: FP Graham Publishing Co., 2003:35–63.

72. Adamson JW, Eschbach JW. Treatment of the anemia ofchronic renal failure with recombinant human erythropoie-tin. Annu Rev Med 1990; 41:349–360.

73. Means RT, Krantz SB. Progress in understanding the patho-genesis of the anemia of chronic disease. Blood 1992;80:1639–1647.

74. JelkmannW. Proinflammatory cytokines lowering erythropoie-tin production. J Interferon Cytokine Res 1998; 18:555–559.

75. Spivak JL. The blood in systemic disorders. Lancet 2000;355:1707–1712.

76. Cazzola M, Beguin Y. New tools for clinical evaluation oferythron function in man. Br J Haematol 1992; 80:278–284.


77. Beguin Y, Loo M, R’Zik S, Sautois B, Lejeune F, Rorive G,Fillet G. Early prediction of response to recombinant humanerythropoietin in patients with the anemia of renal failure byserum transferrin receptor and fibrinogen. Blood 1993;82:2010–2016.

78. D’Andrea AD, Zon LI. Erythropoietin receptor. Subunit struc-ture and activation. J Clin Invest 1990; 86:681–687.

79. Youssoufian H, Longmore G, Neumann D, Yoshimura A,Lodish HF. Structure, function, and activation of the erythro-poietin receptor. Blood 1993; 81:2223–2236.

80. Livnah O, Stura EA, Middleton SA, Johnson DL, Jolliffe LK,Wilson IA. Crystallographic evidence for preformed dimers oferythropoietin receptor before ligand activation. Science1999; 283:987–990.

81. Remy I, Wilson IA, Michnick SW. Erythropoietin receptoractivation by a ligand-induced conformation change. Science1999; 283:990–993.

82. Witthuhn BA, Quelle FW, Silvennoinen O, Yi T, Tang B,Miura O, Ihle JN. JAK2 associates with the erythropoietinreceptor and is tyrosine phosphorylated and activated follow-ing stimulation with erythropoietin. Cell 1993; 74:227–236.

83. Klingmuller U. The role of tyrosine phosphorylation in prolif-eration and maturation of erythroid progenitor cells—signalsemanating from the erythropoietin receptor. Eur J Biochem1997; 249:637–647.

84. Constantinescu SN, Huang LJ, Nam H, Lodish HF. Theerythropoietin receptor cytosolic juxtamembrane domaincontains an essential, precisely oriented, hydrophobic motif.Mol Cell 2001; 7:377–385.

85. Yi T, Zhang J, Miura O, Ihle JN. Hematopoietic cell phospha-tase associates with erythropoietin (Epo) receptor afterEpo-induced receptor tyrosine phosphorylation: identificationof potential binding sites. Blood 1995; 85:87–95.

86. Klingmuller U, Lorenz U, Cantley LC, Neel BG, Lodish HF.Specific recruitment of SH-PTP1 to the erythropoietin recep-tor causes inactivation of JAK2 and termination of prolifera-tive signals. Cell 1995; 80:729–738.


87. Kralovics R, Prchal JT. Genetic heterogeneity of primaryfamilial and congenital polycythemia. Am J Hematol 2001;68:115–121.

88. Suzuki N, Ohneda O, Takahashi S, Higuchi M, Mukai HY,Nakahata T, Imagawa S, Yamamoto M. Erythroid-specificexpression of the erythropoietin receptor rescued its nullmutant mice from lethality. Blood 2002; 100:2279–2288.

89. Baynes RD, Reddy GK, Shih YJ, Skikne BS, Cook JD. Serumform of the erythropoietin receptor identified by a sequence-specific peptide antibody. Blood 1993; 82:2088–2095.

90. Harris KW, Winkelmann JC. Enzyme-linked immunosorbentassay detects a potential soluble form of the erythropoietinreceptor in human plasma. Am J Hematol 1996; 52:8–13.

91. Koury MJ, Bondurant MC. The molecular mechanism oferythropoietin action. Eur J Biochem 1992; 210:649–663.

92. De Maria R, Zeuner A, Eramo A, Domenichelli C, Bonci D,Grignani F, Srinivasula SM, Alnemri ES, Testa U, PeschleC. Negative regulation of erythropoiesis by caspase-mediatedcleavage of GATA-1. Nature 1999; 401:489–493.

93. Spivak JL. Polycythemia vera: myths, mechanisms, andmanagement. Blood 2002; 100:4272–4290.

94. Blacklock HA, Royle GA. Idiopathic erythrocytosis—a declin-ing entity. Br J Haematol 2001; 115:774–781.

95. Burns S, Arcasoy MO, Li L, Kurian E, Selander K, EmanuelPD, Harris KW. Purification and characterization of theyeast-expressed erythropoietin mutant Epo (R103A), a speci-fic inhibitor of human primary hematopoietic cell erythropoi-esis. Blood 2002; 99:4400–4405.

96. Wang AY-M, Yu AW-Y, Lam CW-K, Yu LM, Li PK-T, Goh J,Lui SF. Effects of losartan or enalapril on hemoglobin, circu-lating erythropoietin, and insulin-like growth factor-1 inpatients with and without posttransplant erythrocytosis.Am J Kidney Dis 2002; 39:600–608.

97. Plata R, Cornejo A, Arratia C, Anabaya A, Perna A, DimitrovBD, Remuzzi G, Ruggenenti P. Angiotensin-converting-enzyme inhibition therapy in altitude polycythaemia: aprospective randomised trial. Lancet 2002; 359:663–666.


98. Jelkmann W. Use of recombinant human erythropoietin asan antianemic and performance enhancing drug. Curr Phar-maceut Biotechnol 2000; 1:11–31.

99. Crawford J, Cella D, Cleeland CS, Cremieux PY, DemetriGD, Sarokhan BJ, Slavin MB, Glaspy JA. Relationshipbetween changes in hemoglobin level and quality of lifeduring chemotherapy in anemic cancer patients receivingepoetin alfa therapy. Cancer 2002; 95:888–895.

100. Wrighton NC, Balasubramanian P, Barbone FP, KashyapAK, Farrell FX, Jolliffe LK, Barrett RW, Dower WJ.Increased potency of an erythropoietin peptide mimeticthrough covalent dimerization. Nat Biotechnol 1997;15:1261–1265.

101. Qureshi SA, Kim RM, Konteatis Z, Biazzo DE, Motamedi H,Rodrigues R, Boice JA, Calaycay JR, Bednarek MA, Griffin P,Gao YD, Chapman K, Mark DF. Mimicry of erythropoietin bya nonpeptide molecule. Proc Natl Acad Sci USA 1999;96:12156–12161.

102. Goldberg J, Jin Q, Ambroise Y, Satoh S, Desharnais J, CappsK, Boger DL. Erythropoietin mimetics derived from solutionphase combinatorial libraries. J Am Chem Soc 2002;124:544–555.

103. Wanner RM, Spielmann P, Stroka DM, Camenisch G, Came-nisch I, Scheid A, Houck DR, Bauer C, Gassmann M, WengerRH. Epolones induce erythropoietin expression via hypoxia-inducible factor-1 alpha activation. Blood 2000; 96:1558–1565.

104. Ren X, Dorrington KL, Maxwell PH, Robbins PA. Effects ofdesferrioxamine on serum erythropoietin and ventilatorysensitivity to hypoxia in humans. J Appl Physiol 2000;89:680–686.

105. Kling PJ, Dragsten PR, Roberts RA, Dos-Santos B, BrooksDJ, Hedlund BE, Taetle R. Iron deprivation increaseserythropoietin production in vitro, in normal subjects andpatients with malignancy. Br J Haematol 1996; 95:241–248.

106. Barbone FP, Johnson DL, Farrell FX, Collins A, MiddletonSA, McMahon FJ, Tullai J, Jolliffe LK. New epoetin mole-


cules and novel therapeutic approaches. Nephrol DialTransplant 1999; 14(suppl 2):80–84.

107. Bartholomew A, Patil S, Mackay A, Nelson M, Buyaner D,Hardy W, Mosca J, Sturgeon C, Siatskas M, Mahmud N,Ferrer K, Deans R, Moseley A, Hoffman R, Devine SM.Baboon mesenchymal stem cells can be genetically modifiedto secrete human erythropoietin in vivo. Hum Gene Ther2001; 12:1527–1541.

108. Rinsch C, Dupraz P, Schneider BL, Deglon N, Maxwell PH,Ratcliffe PJ, Aebischer P. Delivery of erythropoietin byencapsulated myoblasts in a genetic model of severe anemia.Kidney Int 2002; 62:1395–1401.

109. Golde DW, Gasson JC. Hormones that stimulate the growthof blood cells. Sci Am 1988; 259:62–71.

110. Schobersberger W, Hobisch-Hagen P, Fuchs D, Hoffmann G,Jelkmann W. Pathogenesis of anaemia in the critically illpatient. Clin Intens Care 1998; 9:111–117.


3

Sepsis and Systemic InflammatoryResponse Syndrome

HERBERT TILG

Department of Medicine, Division ofGastroenterology and Hepatology, University

Hospital Innsbruck, Innsbruck, Austria

INTRODUCTION

Sepsis is the leading cause of critically ill patients in theWestern world and its incidence continues to increase. Thisclinical syndrome complicates severe infection and is charac-terized by systemic inflammation and widespread tissueinjury. Epidemiological analyses have shown that tissueinjury secondary to activation of the inflammatory systemmay also complicate noninfectious disorders (e.g., acute pan-creatitis). The term systemic inflammatory response syndrome(SIRS) is used in this context to refer to the consequences

87

of a dysregulated host inflammatory response when infectionis not seen (1,2).

Due to these facts, clinicians distinguish between anunderlying disease (infection or pancreatitis) and the host’sresponse (sepsis or SIRS). This distinction is important clini-cally since it is the latter, not the initial disease, which isresponsible for the multiple organ dysfunction syndrome(MODS) seen in those patients. The MODS is usually respon-sible for the high mortality associated with these syndromes.

Whereas our pathophysiological interpretation in thelast two decades focused primarily on the role of proinflamma-tory mediators and their inhibition (2–4), recent insights sug-gest that therapies aimed at reversing the accompanyingperiods of immunosuppression might be more appropriate.Modern pathophysiological concepts and treatment optionsare discussed in detail in this chapter.

DEFINITIONS

To understand and differentiate the various clinicalsyndromes associated with inflammation, a clear and repro-ducible definition used both for clinical research and dailypractice is necessary.

Infection

Infection describes a microbial infection characterized by aninflammatory response in the presence of micro-organismsor the invasion of normally sterile host tissue by organisms.

Bacteremia

Bacteremia is defined by the presence of bacteria in the blood.

Systemic Inflammatory Response Syndrome (SIRS)

Systemic inflammatory response syndrome (SIRS) is a diffuseand widespread inflammatory response to a variety of severeclinical insults (noninfectious or no infectious agent identi-fied). The syndrome is defined by the presence of two or more

88 Tilg

of the following: temperature > 38�Cor< 36�C;heart rate > 90beats=min; respiratory rate > 20 breaths=min or PaCo2< 32mmHg; or white blood cell count > 12,000, < 4000 cells=mm3, or> 10% immature (band) forms.

Sepsis

Sepsis is the systemic response to infection. Thus, in sepsis,clinical aspects of SIRS are accompanied by evidence of defi-nite infection.

Severe Sepsis

Severe sepsis is defined in case of associated organ dysfunc-tion, hypoperfusion, or hypotension. The manifestations ofhypoperfusion may include lactic acidosis, oliguria, or anacute change in mental status.

Septic Shock

Septic shock includes sepsis with hypotension despite suffi-cient volume replacement combined with perfusion abnormal-ities that may include, but are not limited to, lactic acidosis,oliguria, or an acute alteration in mental status.

Multiple Organ Failure

Multiple organ failure (MOF) refers to the presence of alteredorgan function in an acutely ill patient such that homeostasiscannot be maintained without intervention. Increasingabnormalities in the following organ-specific parameterscorrelate with a higher mortality: PO2=FiO2 ratio, serumcreatinine, platelet count, Glasgow coma score, and serumbilirubin.

PATHOGENESIS

Sepsis has been considered as a malignant intravascular typeof inflammation reflecting exaggerated inflammatoryresponses. In all cases of injury (infectious and noninfectious),

Sepsis and SIRS 89

proinflammatory and anti-inflammatory mediators arereleased simultaneously and=or synthesized by the host (2).Besides many involved mediators (mainly cytokines), manydifferent cell types are involved in this response includingneutrophils, lymphocytes, monocytes=macrophages, dendriticcells, and endothelial cells. Whereas the proinflammatoryresponse is initiated primarily to fight against invadingmicro-organisms and=or tissue injury, the counteractinganti-inflammatory response of the body finally leads to anergyand immunosuppression (Fig. 1) (5,6), which in itself is

Figure 1 Many immune cells including macrophages, dendriticcells, neutrophils, and T cells are involved in responses to patho-gens (bacteria and viruses). The pathophysiology of SIRS=sepsis isclosely related to the various cytokines produced by all involved celltypes. Whereas proinflammatory cytokines such as IL-1 and TNFmediate early aspects of SIRS=sepsis, the later stage of sepsis(anergy and immunosuppression) is primarily regulated by theanti-inflammatory and immunoregulatory cytokines IL-10 andTGF-beta.

90 Tilg

also associated with clinical features=mortality of thesesyndromes.

Normal Inflammatory Responses

Inflammation is initiated to limit and control infection.Although the initiating insults are different (infectious andnoninfectious), the body’s responses are qualitatively verysimilar. One of the first events is the expression of adhesionmolecules on endothelial cells to allow activated neutro-phils=lymphocytes to leave the circulation and to enter theinsulted tissue. Released products by activated immune cellsfurther enable local vasodilatation, hyperemia, and increasedmicrovascular permeability. These processes (adherence, che-motaxis, phagocytosis, and bacterial killing) are mainly regu-lated by the synthesis and release of cytokines bymonocytes=macrophages. Cytokines involved in these earlyprocesses include mainly the proinflammatory cytokinestumor necrosis factor (TNF) and interleukin-1 (IL-1) (7,8).Release of these mediators further recruits more neutrophilsand macrophages in a paracrine fashion. The net effect=goalis clearance of invading micro-organisms and=or debris,which is followed by tissue repair. This system in general isvery balanced and is terminated either by the clearance ofbacteria and—probably the most important feature—by theconcurrent release of anti-inflammatory, counteracting cyto-kines such as IL-10 (5). In some cases, the release of theseproinflammatory mediators is overwhelming, causing sys-temic presence of these soluble mediators. In this case, againdependent on the amount of these proinflammatory media-tors, we observe a generalized response to a ‘‘local problem’’,which is referred to as sepsis when it occurs in associationwith infection, and SIRS when it is induced by noninfectiousconditions.

Proinflammatory Cytokines

Not surprisingly, the first cytokines identified in the early1980s were the two key proinflammatory cytokines, namelyTNF and IL-1. Both are of equal importance and share many

Sepsis and SIRS 91

biological principles. Another aspect, underlying their impor-tance in the biology of inflammation, is the fact that they arecontrolled very tightly by the body by the parallel release ofTNF- and IL-1-neutralizing factors, mainly constituting theirsoluble receptors. Evidence supporting a role for TNF and IL-1 in sepsis=SIRS arises from several aspects: (i) elevated cir-culating levels epists in sepsis; (ii) infusion of both mediatorscauses sepsis in animal models; and (iii) neutralization of bothcytokines protects animals from lethal challenge with endo-toxin, the key upregulating factor of both cytokines (7,9,10).Furthermore, various other cytokines such as IL-10 have, astheir key biological function, to counteract and down-regulate actions of TNF and IL-1. However, effects ofcytokines such as IL-6 and IL-10 are not solely anti-inflammatory, as they also exhibit various different functions,which are discussed later.

Complement Activation

The complement system consists of more than 30 plasma andmembrane proteins that all organize and enhance innateimmunity functions. Furthermore, they are involved in theupregulation of various humoral immune responses. This sys-tem works by depositing complement components on patholo-gic targets and by promoting inflammation. Reflecting one ofthe most potent cascade systems in the innate immunesystem, a rapid amplification of the system is possible(11,12). Antagonizing key components of the complementsystem (C5a, C5a receptor) enables decreased inflammation,vascular permeability, andmortality in animalmodels (13,14).

Apoptosis

The precise mechanisms of cell injury and resulting organdysfunction in sepsis are not fully understood. Besidesischemic events (microthrombosis) and direct cytopathicinjury, programmed cell death (apoptosis) seems to be of criti-cal importance. This is the principal mechanism by whichsenescent or dysfunctional cells are eliminated by the body.

92 Tilg

In addition, cell death via apoptosis is the dominant processleading to the termination of inflammation once infectionhas subsided. Experimental studies in animals and criticallyill patients have demonstrated that increased apoptosis oflymphoid organs and some parenchymal tissues contributesto the immune suppression, anergy, and organ dysfunctionobserved in sepsis syndromes (15,16). A recent study byHotchkiss et al. (17) demonstrated increased apoptosis in lym-phocytes (spleen and lymph nodes) and gastrointestinalepithelial cells as well as pronounced lymphopenia in patientswho died of sepsis. In addition, marked increases in activatedcaspase-3 and reduced Bcl-2 expression were seen in these tis-sues. In contrast, patients dying from nonseptic causes didnot show an increase in apoptosis in any of these cell popula-tions. Earlier clinical reports are consistent with these find-ings, showing an increased frequency of lymphopenia inpatients dying from sepsis. While lymphoid cells in sepsisare undergoing accelerated apoptosis, spontaneous neutrophilapoptosis associated with sepsis or SIRS is delayed. Thisdecreased apoptosis of neutrophils is thought to be importantin enhancing tissue injury in SIRS and other injuries seen insepsis by promoting a dysbalanced tissue load of neutrophilsand uncontrolled release of toxic metabolites to other cellssuch as endothelial cells. Delayed neutrophil apoptosis hasbeen associated with severe clinical sepsis. During sep-sis=SIRS, apoptosis can be triggered by the release of corticos-teroids or so called ‘‘death’’ cytokines such as TNF or Fasligand. In experimental animals, treatment with inhibitorsof apoptosis can prevent lymphoid cell apoptosis and mightalso improve outcome. Currently, although clinical trials withantiapoptotic agents remain difficult (administration, tissuespecificity: lymphocytes vs. neutrophils), inhibition of lympho-cyte apoptosis represents an attractive therapeutic modalityin the future.

Bacterial Factors

Various bacterial factors contribute to activation of immunecells. These factors include: endotoxin, cell wall components

Sepsis and SIRS 93

of bacteria (peptidoglycan, muramyl dipeptide, and lipote-choic acid), staphylococcal enterotoxin B, toxic shock syn-drome toxin, Pseudomonas exotoxin A, and others (18).They all interact with various surface molecules of immunecells (mainly monocytes=macrophages and lymphocytes) andcause the release of cytokines. There is substantial evidencethat endotoxin is one of the key toxins responsible for manyof the clinical syndromes observed (7,19). Besides the activa-tion of the cytokine cascade, endotoxin activates thecomplement, coagulation, and fibrinolytic system (2,20).Endotoxemia is observed in patients with sepsis syndromesand elevated plasma levels are associated with shock andmultiple organ failure.

Anti-inflammatory Responses

The interaction between proinflammatory and anti-inflam-matory cytokines can be viewed as the fight between oppos-ing influences. Whereas TNF and interferon gamma (IFNg)reflect the so-called Th-1 cytokines, most anti-inflammatoriesbelong to the group of Th-2 cytokines. One of these key anti-inflammatory cytokines is IL-10. Interleukin-10 is producedby monocytes=macrophages and Th-2 lymphocytes andantagonizes the generation of Th-1 cytokines (TNF, IFN g,IL-1, IL-2, and IL-18). Via these actions, IL-10 is not onlyable to control inflammation, but more importantly inducesimmunosuppression, which might be of considerable impor-tance in terms of SIRS=sepsis (5). Interleukin-10 circulatesin the blood of patients with sepsis syndromes, and increasedconcentrations have been associated with an adverse out-come. On the other hand, experimental animal studies havedemonstrated that endogenously produced and=or exogen-ously administered IL-10 might positively affect outcome insepsis. However, as mentioned, endogenous IL-10 productionand systemic administration can also exacerbate T-celldysfunction, decrease T-cell apoptosis, reduce antimicrobialfunction, and increase mortality in other less acute bacterialmodels of sepsis or after thermal injury. One of the key func-tions of IL-10 in sepsis might be its effects on neutrophil

94 Tilg

apoptosis. Proinflammatory cytokines prolong life of neutro-phils, and this may contribute to the adverse outcomes ofincreased neutrophil activation, which occurs in the lungsand other tissues. Interleukin-10 increases apoptosis ofneutrophils in sepsis; the unresolved issue, however, iswhether this might also reduce innate immune responsesand attenuate neutrophil antimicrobial functions. So, themajor question remains: To what extent is an anti-inflamma-tory response initiated by the body beneficial and=ordetrimental? Different scenarios might result from thecombined effects of the host’s anti-inflammatory responses:(i) In cases of tight balance where the initial infectious insultis overcome, homeostasis will be restored; (ii) the initial insultis so severe that it directly leads to SIRS and organ dysfunc-tion; and (iii) most patients who survive the initial severeinsult enter an immunosuppressive state caused by amassive anti-inflammatory reaction associated with immuneparalysis.

From Excessive Proinflammatory Response toImmunosuppression: A Disorder Due toUncontrolled Inflammation?

Even Lewis Thomas speculated that ‘‘It is our response to theinvading pathogen that makes the disease. Our arsenals forfighting off bacteria are so powerful . . . that we are more indanger from them than the invaders.’’ In favor of this think-ing is also our recent clinical experience that neutralizationof endotoxin by specific antibodies and neutralizing key cyto-kines such as TNF or IL-1 failed to improve outcome in sepsistrials. The theory that death in sepsis was attributable to anaccelerated and overwhelmingly active immune system wasmainly based on studies in animals that reflected a model thatdoes not correlate with clinical reality in humans. Anotherimportant aspect to mention is the fact that circulating levelsof TNF and IL-1 are detected altogether in a minority ofpatients. In certain forms of sepsis, e.g., meninogococcemia,circulating levels of TNF are high and correlate withmortality. This phenomenon seems to be present more in

Sepsis and SIRS 95

young ‘‘healthy’’ people in whom the immune system reactshyperinflammatory. The individual cytokine response isdetermined by many factors, including the virulence of theorganism, the size of the inoculum, and the patient’s coexist-ing conditions, age, and probably polymorphisms in genes forvarious cytokines. Therefore, the ‘‘young and healthy’’ patientmight react completely different from the elderly malnour-ished patient who may show a prolonged hypoinflammatoryresponse. These aspects have to be kept in mind when inter-preting older studies using, e.g., anti-TNF antibody treatmentin patients with sepsis (3,4). A better understanding of thisdistinct pathophysiology has again initiated a debate on themerits of inhibiting cytokines in patients with sepsis. Recentsubgroup analysis of patients being treated with anti-TNFantibodies has revealed evidence that a subgroup of patientshad improved survival (4). Also, a meta-analysis of clinicaltrials with anti-inflammatory agents showed that eventhough such a treatment was generally harmful, again a sub-group of patients (around 10%) benefited (21). So, in clinicalreality, there seems to be a small group of patients with sepsisin whom the ‘‘cytokine storm’’ is part of their disease andtherefore anticytokine strategies might be beneficial. How-ever, it is very clear that in the majority of patients, thisconcept is not beneficial.

Mechanisms of Immune Suppression in Sepsis

Patients with sepsis have several features of immunosuppres-sion: loss of delayed hypersensitivity, inability to clear infec-tion, and a predisposition to nosocomial infections. Asmentioned above, in many patients, the clinical situation ofsepsis may be associated with an anti-inflammatory immuno-suppressive state. This is also in accordance with a recentfinding demonstrating that the application of the proinflam-matory cytokine IFNg reversed this situation in patients withsepsis, restoring macrophage TNF production and improvingsurvival (22). As discussed previously, mechanisms ofimmune suppression in sepsis may be caused mainly by theTh-2 cytokines IL-4 and IL-10 (5). Anergy is a state of

96 Tilg

nonresponsiveness to antigen. Increased synthesis of, e.g., IL-10 suppresses T-cell function and thereby might lead toanergy seen in sepsis. Another aspect in anergy is apoptosis(11,12). As mentioned, a large number of lymphocytes andgastrointestinal epithelial cells die by apoptosis during sepsis.Apoptotic cells induce anti-inflammatory cytokines andthereby further enhance anergy, impairing the response topathogens. In summary, advanced sepsis is clearly associatedwith immunosuppression involving Th-2 cytokines, anergy,and apoptosis. This recently arising new concept in sepsis cer-tainly demands new treatment strategies, which must becompletely different from simply neutralizing a proinflamma-tory cytokine.

EPIDEMIOLOGY

More than any disease, sepsis and SIRS reflect the responseto the aggressive and modern medicine practiced in the Wes-tern world (23). Besides the recent observed increase in lifeexpectancy, increasing incidence=prevalence of these syn-dromes is mainly due to the high number of immunocompro-mised patients (transplantation; cancer; end-stage liver,renal and cardiovascular disease). The frequency of thesesyndromes is expected to increase further and current datasuggest that, in the United States, these diseases cause morethan 100,000 deaths annually. The lack of national databasesmakes it impossible, however, to establish the exact incidenceof sepsis=SIRS in most countries. Most retrospective analysesthroughout the Western world suggest that severe sepsis ispresent in at least 6–10% of all intensive care unit admissionsand sepsis is suspected in 20%.

Sepsis is most frequently observed in middle-aged andelderly patients. Many patients with established sepsis haveunderlying diseases such as malignancies, renal or hepaticfailure, or advanced cardiovascular disease. More impor-tantly, a primary site of infection can only be identified inas many as 10% of patients with criteria for severe sepsisand=or SIRS.

Sepsis and SIRS 97

CATEGORICAL DEFINITIONS ANDPATIENT RISK

Despite the advantage of new categorical definitions (SIRS,severe sepsis, and septic shock), they still identify and definepatients with considerably different mortalities. Various clin-ical characteristics have a major impact on mortality:

i. Abnormal host response to infection: Failure todevelop fever is associated with increased mortality.Another feature associated with increased mortalityis leukopenia (a white blood cell count less than4000 cells=mm3). Both aspects might reflect anabnormal host response.

ii. Site of infection: Sepsis originating from the urogen-ital tract has the lowest mortality rates, whereassepsis arising from the gastrointestinal or pulmon-ary tract has a considerably higher mortality rate.

iii. Positive blood cultures: The presence or absence of apositive blood culture has altogether no impact onmortality rates, again supporting the concept thathost factors are of major importance. However,nosocomial bloodstream infections have a worse out-come than community-acquired infections. Thiscould also reflect a more severe underlying disease.

iv. Antimicrobial therapy: The influence of timely anti-biotic therapy is uncertain, although some reportssuggest a beneficial effect of the early treatmenton survival.

v. Underlying disease: This probably reflects the mostrelevant prognostic factor. Risk factors for mortalityinclude comorbid conditions such as renal and hepa-tic failure, malignancies, immunosuppressive thera-pies, advanced cardiovascular diseases, and humanimmunodeficiency virus (HIV) infection. Anotherclearly established feature is the fact that thereexists a continuum of severity from sepsis to septicshock and multiorgan failure.

98 Tilg

NEW CONCEPTS IN THE TREATMENT OF SEPSIS

Many patients developing sepsis do so without exhibitingfever or inducing an appropriate acute-phase response withelevated C-reactive protein levels. Even more important, thelack of an apparent acute-phase response in patients withsepsis is associated with high mortality and might reflectthe immunosuppressive state of sepsis (Th-2 cytokines donot induce C-reactive protein in hepatocytes).

Activated Protein C

Recombinant human activated protein C is the first anti-inflammatory agent that has been effective in the treatmentof sepsis (24). This drug prevents formation of thrombinthereby inhibiting platelet activation, neutrophil recruitment,and mast-cell degeneration. Furthermore, this pleiotropicmolecule blocks cytokine synthesis and has also antiapoptoticfeatures. Altogether, this new drug inherits several featuresthat might be responsible for its efficacy in severe sepsis,including inhibition of thrombin generation and cytokinesand antiapoptotic properties. However, it is currently notknown how it affects the immunosuppressive state ofadvanced sepsis discussed in the previous paragraph. Dueto its high costs, this treatment is currently limited in manycountries to very severely affected and young patients.

Other Therapies

Intensive insulin therapy recently has been shown to reducethe rate of death from multiple organ failure among patientswith sepsis, regardless of whether they had diabetes. The pro-tective mechanism of insulin in sepsis is unknown (25). Bloodglucose levels should be maintained around 80–110 mg=dL.Early aggressive treatment with colloid or crystalloid infu-sions, vasoactive agents, and transfusions of red cells increaseoxygen delivery and survival (26). Corticosteroids at highdoses should not be used in patients with sepsis (27).Low-dose steroid treatment was effective recently in one

Sepsis and SIRS 99

study in patients with septic shock, but this finding needs tobe confirmed by others (28).

Potential Therapies for Patients with Sepsis=SIRS

Besides the use of activated protein C, there is currently noproven ‘‘immune’’ therapy for these conditions. It would beof considerable interest to define hyperimmune=hypoimmunepatients to stratify for potential treatments. Measurement ofcirculating cytokines=acute-phase proteins (CRP) could beone such approach. In case of a hypoimmune state (Th-2 dri-ven), immunostimulatory treatments such as IFNg could be ofinterest (15). This approach, however, needs to be testedin appropriate large clinical trials. In case of a hyperi-mmune state, blockade of cytokines could be the strategy ofchoice.

CONCLUSIONS

The paradigm that ‘‘sepsis and SIRS are diseases based on acytokine storm’’ has changed in the last years. The conceptof immunosuppression has been introduced and will lead toa more detailed treatment concept of patients with sepsisand SIRS. As proven repeatedly in medical history, only abetter understanding of the pathophysiology of diseases willlead finally to new and better treatments.

REFERENCES

1. Bone RC, Balk RA, Cerra FB, et al. Definitions for sepsis andorgan failure and guidelines for the use of innovative therapiesin sepsis. Chest 1992; 101:1644–1655.

2. Hotchkiss RS, Karl IE. The pathophysiology and treatment ofsepsis. N Engl J Med 2002; 348:138–150.

3. Abraham E, Wunderink R, Silverman H, et al. Efficacy andsafety of monoclonal antibody to human tumor necrosis factor

100 Tilg

alpha in patients with sepsis syndrome: a randomized,controlled, double-blind, multicenter clinical trial. JAMA1995; 273:934–941.

4. Reinhart K, Karzai W. Anti-tumor necrosis factor therapy insepsis: update on clinical trials and lessons learned. Crit CareMed 2001; 29:S121–S125.

5. Oberholzer A, Oberholzer C, Moldawer LL. Interleukin-10: acomplex role in the pathogenesis of sepsis syndromes and itspotential as an anti-inflammatory drug. Crit Care Med 2002;30:S58–S63.

6. Gogos CA, Drosou E, Bassaris HP, Skoutelis A. Pro-versusanti-inflammatory cytokine profile in patients with severesepsis: a marker for prognosis and future therapeutic options.J Infect Dis 2000; 181:176–180.

7. Beutler B, Milsark IW, Cerami A. Passive immunizationagainst cachectin=tumor necrosis factor protects mice fromlethal effects of endotoxin. Science 1985; 229:869–872.

8. Oberholzer A, Oberholzer C, Moldawer LL. Sepsis syndromes:understanding the role of innate and acquired immunity.Shock 2001; 16:83–96.

9. Tracey KJ, Beutler B, Lowry SF, et al. Shock and tissue injuryinduced by recombinant human cachectin. Science 1986;234:470–472.

10. Dinarello CA. Biologic basis for interleukin-1 in disease. Blood1996; 87:2095–2147.

11. Walport MJ. Complement. First of two parts. N Engl J Med2001; 344:1058.

12. Walport MJ. Complement. Second of two parts. N Engl J Med2001; 344:1140.

13. Furebring M, Hakansson LD, Venge P, et al. Expression of theC5a receptor (CD88) on granulocytes and monocytes inpatients with severe sepsis. Crit Care 2002; 6:363.

14. Huber-Lang MS, Riedeman NC, Sarma JV. Complement-induced impairment of innate immune system during sepsis.J Immunol 2002; 169:3223.

Sepsis and SIRS 101

15. Voll RE, Herrmann M, Roth EA, Stach C, Kalden JR,Girkontaite I. Immunosuppressive effects of apoptotic cells.Nature 1997; 390:350–351.

16. Oberholzer C, Oberholzer A, Clare-Salzler M, Moldawer LL.Apoptosis in sepsis: a new target for therapeutic exploration.FASEB J 2001; 15:879–892.

17. Hotchkiss RS, Chang KC, Swanson PE, et al. Caspase inhibi-tors improve survival in sepsis: a critical role of the lympho-cyte. Nat Immunol 2000; 1:496–501.

18. Pugin J. Recognition of bacteria and bacterial products by hostimmune cells in sepsis. In:Vincent JL, ed. Yearbook of Inten-sive Care and Emergency Medicine. Berlin: Springer-Verlag,1996:11.

19. Suffredini AF, Fromm RE, Parker MM. The cardiovascularresponse of normal humans to the administration of endotoxin.N Engl J Med 1989; 321:280.

20. Tapper H, Herwald H. Modulation of hemostatic mechanismsin bacterial infectious diseases. Blood 2000; 96:2329.

21. Zeni F, Freemann BF, Natanson C. Anti-inflammatory thera-pies to treat sepsis and septic shock: a reassessment. Crit CareMed 1997; 25:1095–1100.

22. Docke WD, Randow F, Syrbe U, et al. Monocyte deactivation inseptic patients: restoration by IFN-gamma treatment. NatMed 1997; 3:678–681.

23. Angus DC, Linde-Zwirble WT, Lidicker J, et al. Epidemiologyof severe sepsis in the United States: analysis of incidence,outcome, and associated costs of care. Crit Care Med 2001;29:1303–1310.

24. Bernard GR, Vincent J-L, Laterre P-F, et al. Efficacy andsafety of recombinant human activated protein C for severesepsis. N Engl J Med 2001; 344:699–709.

25. Van den Berghe G, Wouters B, Weekers F, et al. Intensiveinsulin therapy in critically ill patients. N Engl J Med 2001;345:1359–1367.

102 Tilg

26. Rivers E, Nguyen B, Havstad S, et al. Early goal-directed ther-apy in the treatment of severe sepsis and septic shock. N EnglJ Med 2001; 345:1368–1377.

27. Abraham E, Evans T. Corticosteroids and septic shock. JAMA2002; 288:862–871.

28. Annane D, Sebille V, Charpentier C, et al. Effect of treatmentwith low doses of hydrocortisone and fludrocortisone on mor-tality in patients with septic shock. JAMA 2002; 288:862–871.

Sepsis and SIRS 103

4

Disturbances of Iron Homeostasis

VICTORIANO MULERO

Department of Cell Biology, Facultyof Biology, University of Murcia,

Murcia, Spain

JEREMY H. BROCK

Department of Immunology,Western Infirmary,

University of Glasgow,Glasgow, U.K.

NORMAL IRON METABOLISM

Molecular Mechanisms of Iron Absorption

Iron homeostasis is mainly controlled by the absorption ofiron from the diet. When iron levels in the body are low, therate of iron absorption is increased, and when iron levelsare replete there is a reduction in the rate of iron absorptionand excess iron is excreted when enterocytes are sloughed offevery 2–3 days. The epithelial cell layer of the duodenum isresponsible for sensing changes in body iron demands andthen adapting to meet them. Within the crypts of the intestineare multipotent precursor cells, which only act as sensors of

PART II: PATHOPHYSIOLOGY

105

body iron needs, but upon differentiation into enterocytes,they are capable of transporting iron (1,2).

Iron is found in the diet as ionic (nonheme) iron andhaem iron. The mechanism of absorption of haem iron hasyet to be fully elucidated. Absorption of ionic iron is a multi-step process involving the uptake of iron from the intestinallumen across the apical cell surface of the villus enterocytesand the transfer out of the enterocyte across the basolateralmembrane to the circulation. Recently, a number of newgenes involved in iron metabolism have been identifiedwhich are allowing the molecular mechanisms of iron absorp-tion to be elucidated (Fig. 1). The most extensively character-ized uptake pathway is via the divalent metal transporterNramp2 (also named SLC11A2, DMT1, and DCT1) thatcan transport ferrous iron as well as a number of other metalions including copper, cobalt, zinc, and lead (3). Evidence for

Figure 1 A model for the pathways of iron absorption by theenterocyte. The figure shows uptake of ionic iron and haem ironfrom the gut lumen and transfer of iron to blood. [From Ref. 2,Copyright (2000) BMJ Publishing Group.]

106 Mulero and Brock

the role of Nramp2 in iron absorption is supported by studiesin mk mice and the Belgrade rat, which have a G185R muta-tion in Nramp2 that inhibits iron uptake across the brushborder leading to iron deficiency (4,5). A ferric reductase,called Dcytb, may be an important element in this ironabsorption pathway, as it may act upstream to Nramp2.Notably, Dcytb is highly expressed in the brush border ofthe duodenum, is regulated by iron status, and induces ferricreductase activity when expressed in Xenopus oocytes andcultured cells (6).

The basolateral membrane of enterocytes mediates thetransfer of the iron transported into the enterocytes to therest of the body by a membrane-bound protein called ferropor-tin (also known as SLC11A3, IREG1, and MPT1) (7–9). Extra-cellular ferrous iron is then oxidized by the multicopperoxidase hephaestin, bound by plasma transferrin (Tf) andpassed initially through the portal system of the liver, whichis the major site of iron storage (10). Hepatocytes take upTf-bound iron (TBI) via the classical Tf receptor 1 (TfR1)(11) but likely in greater amounts by the recently identifiedhomologous protein TfR2 (12,13).

Besides ferroportin, other proteins related to ironmetabolism expressed on the basolateral membrane of intest-inal crypt cells are responsible for sensing body iron stores.These include the TfR1 (11), the hemochromatosis protein(HFE), which forms a complex with b2 microglobulin andTfR1 (14), and the TfR2, which does not interact with theHFE in vitro (15). The role of HFE and TfR2 in mediatingTBI uptake is largely unclear. Mutations of the HFE genelead to hereditary hemochromatosis (HH), a common inher-ited disease characterized by chronic excessive intestinalabsorption of dietary iron that is subsequently deposited,with associated toxicity, in a variety of parenchymal tissues(16). The HFE has been shown to compete with Tf for bind-ing TfR1, and several cell lines engineered to express HFEtake up less TBI and demonstrate reduced ferritin (Ft) andenhanced TfR1 levels (17–19). Therefore, cells expressingdysfunctional HFE should take up excess iron, as occurs inmany tissues of patients with HH. Paradoxically, intestinal

Disturbances of Iron Homeostasis 107

crypt cells in HH behave as though they are relatively iron-deficient (20,21). An elegant study seems to reconcile theseobservations, since HFE is also able to inhibit iron releasefrom macrophages, and probably from intestinal crypt cells,by a mechanism independent of its ability to compete withTf for binding to TfR1 (22).

Molecular Regulation of Iron Metabolism

Intracellular iron homeostasis is post-transcriptionally con-trolled by cytoplasmic iron regulatory proteins (IRP1 andIRP2), which regulate several mRNAs containing iron-responsive elements (IREs) in their untranslated regions(23). The IRP1 is a bifunctional protein that can act eitheras a cytoplasmic aconitase or as an IRE-binding protein(24). In iron replete cells, IRP1 bears a 4Fe–S cluster andshows aconitase activity, but in iron-depleted cells reversibledisassembly of the cluster converts IRP1 to its RNA-bindingform. The IRP2, despite conservation of the cluster-ligatingcysteines at the active site, is unable to assemble a Fe–S clus-ter in vitro and therefore is unable to exhibit aconitase activ-ity (25). Unlike the regulation of IRP1 by iron, loss of IREbinding of IRP2 in iron replete cells is due to iron-dependentoxidation, ubiquitinylation, and degradation by the protea-some (26).

The IRP binding to the IREs in the 50 untranslatedregions of Ft and 5-aminolaevulinate synthase mRNAsrepresses their translation (27), whereas binding of IRP tomultiple IREs in the 30 untranslated region of TfR mRNA con-fers stability against targeted endonucleolytic degradation(28). Several iron transporters also contain an IRE consensussequence motif in their mRNAs, but its role needs to be clar-ified. In vitro and in vivo approaches have demonstrated thatiron deprivation results in increased mRNA and proteinexpression of Nramp2 and ferroportin, but the effect islargely transcriptional rather than post-transcriptional(29–31). Whatever the case, these data provide evidence forthe expected negative feed-back loop between intracellulariron availability and transmembrane iron transport.


A Pivotal Role for Macrophages inIron Metabolism

Iron metabolism and macrophage physiology are closely con-nected. Macrophages, through processing of hemoglobin-ironfrom senescent erythrocytes, are responsible for iron supplyto peripheral tissues, including the bone marrow (32). More-over, changes in macrophage iron content can affect the func-tion of these cells in the inflammatory response. For example,iron plays a critical role in macrophage-mediated cytotoxicityby contributing to the production of highly toxic hydroxylradicals via the Fenton reaction (33) and by controlling theproduction of nitric oxide (NO) after activation by immunolo-gical stimuli (34).

Two transporters of ferrous iron belonging to the Nrampfamily have been shown to participate in iron handling bymacrophages. Nramp2, besides being the major Tf-indepen-dent iron uptake system at the apical pole of intestinal cells,also localizes to early recycling endosomes in most cell typeswhere it seems to transport iron into the cytoplasm (35,36).Nramp1 (also called Slc11a1) localizes to late endosomes=lysosomes (37,38), is more restricted in cellular distribution,principally to cells of the myeloid lineage, and is associatedwith infectious and autoimmune disease susceptibility (39).Whereas Nramp2 is a symporter of protons and divalentcations (3), the direction of cation transport by Nramp1 isquite controversial (40–42). Kuhn et al. (40) found higherrates of iron uptake in isolated phagosomes from wild-typeNramp1 macrophages, compared to phagosomes from mutantNramp1 macrophages, and proposed that the Nramp1 trans-ports iron into phagosomes, where, together with a low pH, itgenerates reactive oxygen species (ROS) via the Fenton reac-tion (43) to kill invading microorganisms. Most recently,employing the Xenopus oocyte-cRNA model, Goswami et al.(42) provided evidence that Nramp1 is a highly pH-dependentantiporter that fluxes divalent cations in either directiondepending on the pH on either side of the membrane. In sharpcontrast, the results obtained by Gros’ group (41) employing adivalent-cation-sensitive fluorophore covalently attached to


zymosan particles demonstrated that Nramp1 is a symp-orter which extrudes Mn2þ from the intraphagosomal spaceof isolated peritoneal macrophages from wild-type Nrampmice.

Whatever the case, in terms of iron uptake and metabo-lism within macrophages, the location of Nramp2 is likely toinfluence TfR-mediated entry of iron into cells, whereasNramp1 would be expected to influence degradation and ironrecycling from effete erythrocytes entering macrophages byphagocytosis. In fact, we have found using uptake of iron via59FeTf–antiTf immune complexes as a model of iron recyclingvia erythrophagocytosis that when iron is delivered to lateendosomes=lysosomes via this phagocytic pathway, but notvia TfR in the early recycling endosomes, macrophages stablytransfected with the wild-type allele ofNramp1 recycle 2.4-foldmore iron to the medium than mutant macrophages (44). Nota-bly, release of iron is inhibitable byNG-monomethyl-l-arginine(NMMA), indicating that NO provides a crucial signal for thisiron release. These studies suggest that Nramp1 plays animportant role in recycling of iron acquired by macrophagesby phagocytosis, implying a role in degradation and recyclingof iron from effete erythrocytes.

Although the mechanisms involved in cellular ironrelease from macrophages are not fully understood, it hasgenerally been assumed that iron released by macrophagesmust first enter the cytoplasm from the phagosome for subse-quent export across the cytoplasmic membrane, perhaps by aprotein analogous to the recently described ferroportin (7–9).Direct evidence for the involvement of ferroportin in ironmetabolism has not been obtained so far, but a recent studyhas shown that ferroportin expression in macrophages ofthe spleen, liver, and bone marrow is downregulated duringacute (LPS-injected mice) and chronic inflammation (Leish-mania donovani-infected mice) (45), suggesting that ironsequestration in the macrophage that accompanies inflamma-tion is due to downregulation of ferroportin. Interestingly,theresponse of ferroportin to LPS requires signaling throughthe LPS-receptor and Toll-like receptor 4 (TLR4), asferroportin expression is not altered in mice lacking this


receptor. In addition, TNFa may not be required for the LPSeffect, since mice lacking TNF receptor 1a respond appropri-ately to LPS with downregulation of ferroportin, despitehyporesponsiveness to TNFa signaling (45).

Iron might also be released from activated macrophagesdirectly via a lysosomal secretory pathway involving Nramp1,as we (44,46) and others (47) have previously suggested. Giventhe potentially damaging influence of high cytoplasmic iron onmRNA stability (48), this may prove the safest route for recy-cling of iron from effete red cells. The NO is known to enhancesecretory and=or excretory mechanisms in other cells (49),which could account for its ability to enhance iron release fromactivated macrophages. Direct extracellular secretion of ironfrom phagosomal contents would also imply that most, if notall, iron entering a cell by phagocytosis never enters the cyto-plasm. Loading the cells with iron via Tf–antiTf immune com-plexes had little effect on Ft expression, compatible with thisproposal.

THE IMBALANCE OF IRON HOMEOSTASISDURING INFLAMMATION

Action of Cytokines

Several cytokines have been shown to affect iron homeostasisby various mechanisms. Proinflammatory cytokines, such asIL-1b and TNFa, induce hypoferremia by modulating macro-phage iron metabolism via induction of Ft biosynthesis(50,51). Recent in vitro evidence has also suggested that theanti-inflammatory cytokines IL-4 and IL-13 may contributeto the observed diversion of iron traffic in ACD by increasingiron uptake from Tf and storage into Ft in activated macro-phages (52). Similarly, administration of IL-10 to chronicactive Crohn’s disease patients leads to anemia (53). Parallelin vitro studies performed with the human monocytic THP-1cell line show that IL-10 decreased the IRE-binding activity ofIRPs to the 50 UTR of Ft mRNA, suggesting that IL-10 contri-butes in vivo to hyperferritinemia and limited iron availabil-ity to erythroid progenitor cells.


All these studies have focused on the effects of cytokineson TfR and Ft expression. Nevertheless, a recent study byWeiss’ group has also evaluated the effects of INFg and IL-10 on the expression of two pivotal macrophage iron transpor-ters, Nramp2 and ferroportin, in THP-1 monocytes (54). Theyfound that IFNg and LPS increase the cellular expression ofNramp2 and stimulate the uptake of non-Tf bound iron(NTBI) into cells, and at the same time downregulate theexpression of ferroportin mRNA and decrease iron releasefrom monocytes. They also confirmed previous finding thattreatments of cells with IFNg=LPS reduced TfR mRNA levels,surface expression, and iron uptake. All these effects werepartly counteracted by preincubation of the cells with IL-10.These results demonstrate that IFNg and LPS increase theuptake of NTBI via stimulation of Nramp2 expression andcause retention of the metal within monocytes by downregu-lating ferroportin synthesis. This contrasts with earlierfindings obtained with bone marrow-derived mouse macro-phages pulsed with NTBI (46), where iron flux (uptake andrelease) was unaffected either by NO or directly by IFNg=LPSstimulation. These differences may actually be the result ofdifferent expression patterns for Nramp2 and ferroportinbetween human and mouse cells and=or between primarycells and transformed cell lines.

The Role of ROS

The IRP binding activity is normally regulated by cellular ironlevels, but other signals such as NO and oxidative stress (i.e.,H2O2 and peroxynitrite) can modulate the activity of both IRPsand thus influence cellular iron metabolism (55–57).

The ROS are generated within cells as byproducts ofbiological oxidations, including electron transfer reactionsin the respiratory chain. In addition, ROS are released dur-ing the respiratory burst of professional phagocytes. AmongROS, H2O2 and peroxynitrite, which is derived from NO andO2

�, have been shown to be able to activate IRPs and thisresults in the reduction of cellular Ft content and in aparallel increase in the cell surface expression of TfR and


stimulation of Tf-mediated iron uptake into cells (55,56,58),suggesting that macrophage activation would lead to anincrease of intracellular iron concentration. Moreover, H2O2-treated cells display an increased capacity to sequester ironin Ft, despite a reduction in the Ft pool, which results in a rear-rangement of iron intracellular distribution (58). Importantly,while treatment of cells with exogenous H2O2 results in therapid activation of IRP-1, elevation of intracellular H2O2

levels seems to be not sufficient for IRP-1 induction (59).Therefore, the large amounts of ROS released by macrophagesand other cells during chronic inflammation might contributeto the observed cellular iron retention, cellular toxicity by thewell-established Fenton reaction, and likely to anemia.

The Role of Reactive Nitrogen Species (RNS)

Iron Handling

The evidence for the central role of NO in the regulation of ironmetabolism was initially obtained using several NO donormolecules. For example, the early finding that S-nitroso-N-acetyl-d,l-penicillamine (SNAP) was able to activate IRE-binding activity of both IRP1 and IRP2 (55) suggested thatNO produced by activated macrophages during inflammationwas able to affect iron metabolism in vivo. This was later con-firmed by our finding that J774 macrophages activated invitro with physiological stimuli (i.e., IFNg and LPS) showedaltered IRP activities and iron-Tf uptake, these effects beingreversed by preincubation of the cells with the nitric oxidesynthase (NOS) inhibitor, NMMA (60). Thus, cell activationby IFNg and=or LPS inhibits Fe uptake with a concomitantdownregulation of TfR expression. This inhibition of ironuptake was partially reversed by NMMA indicating thatNO, despite its well-known ability to activate IRP1 (55),can actually inhibit iron uptake by macrophages. Iron releaseby IFNg=LPS-activated cells was not greater than from con-trol cells, indicating that the effect is actually caused by inhi-bition of Fe uptake from Tf, rather than accelerated releaseof iron during the incubation period (60). However, othershave demonstrated that NO is able to intercept iron before


incorporation into Ft as well as mobilizing iron from Ft by anactive mechanism which is dependent upon glutathione (61).These authors proposed a model where NO–Fe complex wouldbe then transported out of the cells by ferroportin. Neverthe-less, this model is not supported by the results obtained veryrecently by Haile’s group using a luciferase reporter geneunder the control of the mouse ferroportin promoter and 50

UTR IRE (62). These authors showed that exogenously addedNO inhibits luciferase expression in an IRE-dependent man-ner and this change is accompanied by an increase in IRP1IRE-binding activity. In addition, treatment of RAW264.7macrophages with LPS, to produce endogenous NO, resultsin a similar inhibition of luciferase activity in the IRE-Luctransfected cells, the effect being abrogated by the NOS inhibi-tor L-NAME. Importantly, the Baf=3 cell line, which does notexpress IRP1, fails to respond to exogenous NO. It is unlikely,therefore, that ferroportin is involved in the release of iron byactivated macrophages in view of the downregulated expres-sion of ferroportin mediated by NO-dependent (62) and -inde-pendent mechanisms during inflammation.

Although most work on iron metabolism has been doneusing the Tf uptake model, it is uncertain whether ironuptake from Tf is relevant to iron acquisition by macro-phages in vivo, where phagocytosis of effete erythrocytes islikely to be the main source of iron. Therefore, we haveinvestigated the role of NO in iron release using a systembased on phagocytosis of 59FeTf–antiTf immune complexesas a model for erythrophagocytosis (46). It was found thatcells from normal mice showed an increase in iron releasefollowing activation with IFNg and LPS and a pulse of59FeTf–antiTf immune complexes, which could be reversedby NMMA. In contrast, cells from the iNOS-deficient miceshowed a much smaller increase in iron release followingactivation, this residual effect probably reflecting the factthat the cells from the iNOS-deficient mice can still producesmall amounts of NO. These results further suggest that NO,instead of contributing to the hypoferremia of inflammation,may actually have a counterbalancing effect by promotingiron release from macrophages.


TfR Expression

The TfR expression by macrophage cell lines in the presenceof IFNg and LPS appears to be complex and regulated by anumber of different mechanisms. The IFNg causes a verysensitive NO-independent downregulation of TfR expression(60,63) with a concomitant decrease in IRE-binding activityof IRP2 (57,60,63,64), although high production of NO canrestore TfR expression by activating IRP1 and, to a lesserextent, IRP2 (60). However, costimulation of macrophageswith IFNg and LPS leads to a downregulation of TfR expres-sion, which is initially NO-independent but may be modulatedby NO when the latter accumulates at sufficient level (60).Overall, it is clear that the major effect of IFNg=LPS on bothTfR expression and iron uptake by macrophages is downregu-latory, due partly to downregulation of IRP2 and partly toIRP-independent effects, which may involve NO to someextent. Although NO activates IRP1, it cannot compensatefor the inhibitory effects on TfR expression and Fe uptake,except at high concentrations.

Ft Expression

The regulation of Ft mRNA expression and protein content inIFNg=LPS-activated macrophages is also controversial. Weissand collaborators (54,65) reported that IFNg=LPS treatmentincreases Ft mRNA expression in J774 macrophages butdecreases Ft translation, indicating that IRP activationmediated by NO overcomes the increased mRNA expression.In contrast, increased Ft synthesis and accumulation inIFNg=LPS-stimulated macrophages have more recently beendemonstrated, accompanied by a NO-dependent increase inIRP1 and decrease in IRP2 activity (44,46,57,60,63,66). More-over, IFNg=LPS treatment results in a strong decrease in boththe amount and the proportion of iron incorporated intoFt, together with a corresponding compensatory increaseiron bound to intracellular organelles (60). These findingsprobably indicate an increase in iron bound to mitochondriain the more highly metabolic activated macrophages, andthe overall picture is one of iron being used for metabolic


activity rather than being diverted to an enlarged Ft compart-ment. Nevertheless, an NO-dependent downregulation of Ft-bound iron in activated macrophages, perhaps as a conse-quence of NO-mediated iron release from Ft should not bediscounted, as NMMA partially increased this intracellulariron fraction (44,46,60). In conclusion, it seems unlikely thatIFNg=LPS and=or NO contribute to the hypoferremia ofinflammation through promoting increased acquisition of Tf-bound iron, and that other mechanisms, for example reten-tion of iron acquired through phagocytosis of erythrocytesand cell debris, are probably more important.

Hepcidin: A Putative Iron Regulatory Hormone

The exact means by which crypt cells sense body iron storeshas been a mystery, but the importance of this communicationis highlighted by HH and the reciprocal situation, the anemiaof chronic disease. A clue for solving this mystery may lie inthe recently identified protein hepcidin, a plasma protein pro-duced by hepatocytes in response to inflammation (67). Loss ofhepcidin in USF2 knockout mice is associated with increasedcirculating iron, decreased iron levels in macrophages, andapparently increased intestinal iron absorption (68), whiletransgenic mice overexpressing liver hepcidin have decreasediron levels and develop microcytic hypochromic anemia (69).These authors proposed a model (depicted in Fig. 2) whereTfR2 mediates TBI uptake by hepatocytes, which, in turn,modulates expression of hepcidin, which, in turn, interactswith HFE, b2-microglobulin, and TfR in the duodenal cryptcell to regulate dietary iron absorption.

DISTURBANCE OF IRON TRAFFICKING INMACROPHAGES INFECTED WITHINTRACELLULAR PATHOGENS

Extracellular pathogens have evolved a variety of ways to com-pete for iron. Many produces siderophores, low molecular weightiron chelators that compete with and=or remove Feþ3 from host


iron-binding proteins. Alternatively, some bacteria bind anddirectly remove iron from Tf or lactoferrin, without siderophores(70,71). Such strategies work wellbecause the organisms are ableto gain direct access to host iron storage molecules.

Figure 2 Proposed steps in hepcidin regulation of iron homeosta-sis. (1) Increased hepatocellular uptake of TBI by TfR2 (or exposureto LPS) leads to (2) increased production and secretion of hepcidin,which (3) interacts with the b2 microglobulin (b2M)–HFE–TfR1 com-plex and increases iron uptake or retention by RE macrophage andduodenal crypt cells. (4) Crypt cells differentiate into daughterenterocytes programed to have decreased expression of iron trans-port proteins, leading to (5) decreased dietary iron absorption. [FromRef. 68, Copyright (2001) National Academy of Sciences, U.S.A.]


However, not all pathogens reside in the extracellularenvironment. Several intracellular pathogens, includingMycobacterium tuberculosis and Leishmania sp., have devel-oped complex molecular mechanisms in order to gain accessto iron once inside the host cell (72,73). At the same time,macrophages must have enough available intracellular ironto support bactericidal mechanisms, but too much iron favorsgrowth of bacteria, which no longer can be killed by themacrophage (33).

Recent studies have reported that the intracellularpathogens can exploit and subvert Tf trafficking. Thus, infec-tion of macrophages results in increased IRE-binding activityof IRP1, TfR expression, and total cellular iron contenttogether with targeting of iron-Tf to pathogen-containing pha-gosomes (18,73,74). In addition, intraphagosomal M. tubercu-losis is also able to acquire iron from intracellular iron pool(73), suggesting that iron acquisition by the bacteria canoccur via receptor-mediated endocytosis and fusion of theearly endosome with the M. tuberculosis-containing phago-some as well as from an endogenous site(s) that needs to bedefined. Interestingly, activation of macrophages with IFNg,which decreases iron acquisition from Tf, does not effectivelyimpair the ability of M. tuberculosis to acquire intracellulariron (73). These authors hypothesized that the total iron con-tent of the macrophages may not be the key determinant inthe ability of intraphagosomal M. tuberculosis to acquire iron.Rather it is the iron content of as yet to be determined intra-cellular sites in the macrophage that are critical to thesedynamics. The involvement of Nramp1, Nramp2, and ferro-portin iron transporters in this process is quite plausibleand this should be examined in the near future.

REFERENCES

1. Roy CN, Enns CA. Iron homeostasis: new tales from the crypt.Blood 2000; 96:4020–4027.

2. Trinder D, Fox C, Vautier G, Olynyk JK. Molecular pathogen-esis of iron overload. Gut 2002; 51:290–295.


3. Gunshin H, Mackenzie B, Berger UV, Gunshin Y, Romero MF,Boron WF, Nussberger S, Gollan JL, Hediger MA. Cloning andcharacterization of a mammalian proton-coupled metal-iontransporter. Nature 1997; 388:482–488.

4. Fleming MD, Trenor CC III, Su MA, Foernzler D, Beier DR,Dietrich WF, Andrews NC. Microcytic anaemia mice have amutation in Nramp2, a candidate iron transporter gene. NatGenet 1997; 16:383–386.

5. Fleming MD, Romano MA, Su MA, Garrick LM, Garrick MD,Andrews NC. Nramp2 is mutated in the anemic Belgrade (b)rat: evidence of a role for Nramp2 in endosomal iron transport.Proc Natl Acad Sci USA 1998; 95:1148–1153.

6. McKie AT, Barrow D, Latunde-Dada GO, Rolfs A, Sager G,Mudaly E, Mudaly M, Richardson C, Barlow D, Bomford A,Peters TJ, Raja KB, Shirali S, Hediger MA, Farzaneh F,Simpson RJ. An iron-regulated ferric reductase associatedwith the absorption of dietary iron. Science 2001; 291:1755–1759.

7. Abboud S, Haile DJ. A novel mammalian iron-regulatedprotein involved in intracellular iron metabolism. J Biol Chem2000; 275:19906–19912.

8. Donovan A, Brownlie A, Zhou Y, Shepard J, Pratt SJ,Moynihan J, Paw BH, Drejer A, Barut B, Zapata A, Law TC,Brugnara C, Lux SE, Pinkus GS, Pinkus JL, Kingsley PD,Palis J, Fleming MD, Andrews NC, Zon LI. Positional cloningof zebrafish ferroportin1 identifies a conserved vertebrate ironexporter. Nature 2000; 403:776–781.

9. McKie AT, Marciani P, Rolfs A, Brennan K, Wehr K, BarrowD, Miret S, Bomford A, Peters TJ, Farzaneh F, Hediger MA,Hentze MW, Simpson RJ. A novel duodenal iron-regulatedtransporter, IREG1, implicated in the basolateral transfer ofiron to the circulation. Mol Cell 2000; 5:299–309.


11. Anderson GJ, Powell LW, Halliday JW. Transferrin receptordistribution and regulation in the rat small intestine. Effect


of iron stores and erythropoiesis. Gastroenterology 1990;98:576–585.

12. Kawabata H, Yang R, Hirama T, Vuong PT, Kawano S,Gombart AF, Koeffler HP. Molecular cloning of transferrinreceptor 2. A new member of the transferrin receptor-likefamily. J Biol Chem 1999; 274:20826–20832.

13. Fleming RE, Migas MC, Holden CC, Waheed A, Britton RS,Tomatsu S, Bacon BR, Sly WS. Transferrin receptor 2: contin-ued expression in mouse liver in the face of iron overload andin hereditary hemochromatosis. Proc Natl Acad Sci USA 2000;97:2214–2219.

14. Parkkila S, Waheed A, Britton RS, Bacon BR, Zhou XY,Tomatsu S, Fleming RE, Sly WS. Association of the transferrinreceptor in human placenta with HFE, the protein defective inhereditary hemochromatosis. Proc Natl Acad Sci USA 1997;94:13198–13202.

15. West AP Jr, Bennett MJ, Sellers VM, Andrews NC, Enns CA,Bjorkman PJ. Comparison of the interactions of transferrinreceptor and transferrin receptor 2 with transferrin and thehereditary hemochromatosis protein HFE. J Biol Chem 2000;275:38135–38138.

16. Cullen LM, Anderson GJ, Ramm GA, Jazwinska EC, PowellLW. Genetics of hemochromatosis. Annu Rev Med 1999;50:87–98.

17. Feder JN, Penny DM, Irrinki A, Lee VK, Lebron JA, WatsonN, Tsuchihashi Z, Sigal E, Bjorkman PJ, Schatzman RC. Thehemochromatosis gene product complexes with the transferrinreceptor and lowers its affinity for ligand binding. Proc NatlAcad Sci USA 1998; 95:1472–1477.

18. Roy CN, Penny DM, Feder JN, Enns CA. The hereditaryhemochromatosis protein, HFE, specifically regulates trans-ferrin-mediated iron uptake in HeLa cells. J Biol Chem 1999;274:9022–9028.

19. Salter-Cid L, Brunmark A, Li Y, Leturcq D, Peterson PA,Jackson MR, Yang Y. Transferrin receptor is negatively modu-lated by the hemochromatosis protein HFE: implications forcellular iron homeostasis. Proc Natl Acad Sci USA 1999;96:5434–5439.



21. Zoller H, Pietrangelo A, Vogel W, Weiss G. Duodenal metal-transporter (DMT-1, NRAMP-2) expression in patients withhereditary hemochromatosis. Lancet 1999; 353:2120–2123.


23. Eisenstein RS. Iron regulatory proteins and the molecularcontrol of mammalian iron metabolism. Annu Rev Nutr2000; 20:627–662.

24. Kennedy MC, Mende-Mueller L, Blondin GA, Beinert H.Purification and characterization of cytosolic aconitase frombeef liver and its relationship to the iron-responsive elementbinding protein. Proc Natl Acad Sci USA 1992; 89:11730–11734.

25. Phillips JD, Kinikini DV, Yu Y, Guo B, Leibold EA. Differen-tial regulation of IRP1 and IRP2 by nitric oxide in rat hepa-toma cells. Blood 1996; 87:2983–2992.

26. Iwai K, Drake SK, Wehr NB, Weissman AM, LaVaute T,Minato N, Klausner RD, Levine RL, Rouault TA. Iron-dependent oxidation, ubiquitination, and degradation of ironregulatory protein 2: implications for degradation of oxidizedproteins. Proc Natl Acad Sci USA 1998; 95:4924–4928.

27. Gray NK, Hentze MW. Iron regulatory protein prevents bind-ing of the 43S translation pre-initiation complex to ferritin andeALAS mRNAs. EMBO J 1994; 13:3882–3891.

28. Binder R, Horowitz JA, Basilion JP, Koeller DM, KlausnerRD, Harford JB. Evidence that the pathway of transferrinreceptor mRNA degradation involves an endonucleolytic clea-vage within the 30 UTR and does not involve poly(A) tail short-ening. EMBO J 1994; 13:1969–1980.

29. Wardrop SL, Richardson DR. The effect of intracellular ironconcentration and nitrogen monoxide on Nramp2 expression


and non-transferrin-bound iron uptake. Eur J Biochem 1999;263:41–49.

30. Tchernitchko D, Bourgeois M, Martin ME, Beaumont C.Expression of the two mRNA isoforms of the iron transporterNramp2=DMT1 in mice and function of the iron responsiveelement. Biochem J 2002; 363:449–455.

31. Zoller H, Theurl I, Koch R, Kaser A, Weiss G. Mechanisms ofiron mediated regulation of the duodenal iron transportersdivalent metal transporter 1 and ferroportin 1. Blood CellsMol Dis 2002; 29:488–497.

32. Brittenham GM. New advances in iron metabolism, iron defi-ciency, and iron overload. Curr Opin Hematol 1994; 1:101–106.

33. Alford CE, King TE Jr, Campbell PA. Role of transferrin,transferrin receptors, and iron in macrophage listericidalactivity. J Exp Med 1991; 174:459–466.

34. Dlaska M, Weiss G. Central role of transcription factor NF-IL6for cytokine and iron-mediated regulation of murine induciblenitric oxide synthase expression. J Immunol 1999; 162:6171–6177.

35. Gruenheid S, Canonne-Hergaux F, Gauthier S, Hackam DJ,Grinstein S, Gros P. The iron transport protein NRAMP2 isan integral membrane glycoprotein that colocalizes with trans-ferrin in recycling endosomes. J Exp Med 1999; 189:831–841.

36. Canonne-Hergaux F, Zhang AS, Ponka P, Gros P. Characteri-zation of the iron transporter DMT1 (NRAMP2=DCT1) in redblood cells of normal and anemic mk=mk mice. Blood 2001;98:3823–3830.

37. Gruenheid S, Pinner E, Desjardins M, Gros P. Natural resis-tance to infections with intracellular pathogens: the Nramp1protein is recruited to the membrane of the phagosome.J Exp Med 1997; 185:717–730.

38. Searle S, Bright NA, Roach TI, Atkinson PG, Barton CH,Meloen RH, Blackwell JM. Localisation of Nramp1 in macro-phages: modulation with activation and infection. J Cell Sci1998; 111:2855–2866.

39. Blackwell JM, Goswami T, Evans CA, Sibthorpe D, Papo N,White JK, Searle S, Miller EN, Peacock CS, Mohammed H,


Ibrahim M. SLC11A1 (formerly NRAMP1) and disease resis-tance. Cell Microbiol 2001; 3:773–784.

40. Kuhn DE, Baker BD, Lafuse WP, Zwilling BS. Differentialiron transport into phagosomes isolated from the RAW264.7macrophage cell lines transfected with Nramp1Gly169 orNramp1Asp169. J Leukoc Biol 1999; 66:113–119.

41. Jabado N, Jankowski A, Dougaparsad S, Picard V, GrinsteinS, Gros P. Natural resistance to intracellular infections: nat-ural resistance-associated macrophage protein 1 (Nramp1)functions as a pH-dependent manganese transporter at thephagosomal membrane. J Exp Med 2000; 192:1237–1248.

42. Goswami T, Bhattacharjee A, Babal P, Searle S, Moore E, LiM, Blackwell JM. Natural-resistance-associated macrophageprotein 1 is an Hþ=bivalent cation antiporter. Biochem J2001; 354:511–519.

43. Gutteridge JM, Rowley DA, Halliwell B. Superoxide-dependentformation of hydroxyl radicals in the presence of iron salts. Detec-tion of ‘free’ iron in biological systems by using bleomycin-depen-dent degradation of DNA. Biochem J 1981; 199:263–265.

44. Mulero V, Searle S, Blackwell JM, Brock JH. Solute carrier11a1 (Slc11a1; formerly Nramp1) regulates metabolism andrelease of iron acquired by phagocytic, but not transferrin-receptor-mediated, iron uptake. Biochem J 2002; 363:89–94.

45. Yang F, Liu XB, Quinones M, Melby PC, Ghio A, Haile DJ. Reg-ulation of reticuloendothelial iron transporter MTP1 (Slc11a3)by inflammation. J Biol Chem 2002; 277:39786–39791.

46. Mulero V, Wei XQ, Liew FY, Brock JH. Regulation of phagoso-mal iron release from murine macrophages by nitric oxide.Biochem J 2002; 365:127–132.

47. Biggs TE, Baker ST, Botham MS, Dhital A, Barton CH, PerryVH. Nramp1 modulates iron homoeostasis in vivo and in vitro:evidence for a role in cellular iron release involving de-acidification of intracellular vesicles. Eur J Immunol 2001;31:2060–2070.

48. Zwilling BS, Kuhn DE, Wikoff L, Brown D, Lafuse W. Role ofiron in Nramp1-mediated inhibition of mycobacterial growth.Infect Immun 1999; 67:1386–1892.


49. Van Uffelen BE, Van Steveninck J, Elferink JG. Potentiationand inhibition of fMLP-activated exocytosis in neutrophils byexogenous nitric oxide. Immunopharmacology 1997; 37:257–267.

50. Brock JH, Alvarez-Hernandez X. Modulation of macrophageiron metabolism by tumour necrosis factor and interleukin 1.FEMS Microbiol Immunol 1989; 1:309.

51. Alvarez-Hernandez X, Liceaga J, McKay IC, Brock JH. Induc-tion of hypoferremia and modulation of macrophage iron meta-bolism by tumor necrosis factor. Lab Invest 1989; 61:319–322.

52. Weiss G, Bogdan C, Hentze MW. Pathways for the regulationof macrophage iron metabolism by the anti-inflammatory cyto-kines IL-4 and IL-13. J Immunol 1997; 158:420–425.

53. Tilg H, Ulmer H, Kaser A, Weiss G. Role of IL-10 for inductionof anemia during inflammation. J Immunol 2002; 169:2204–2209.

54. Ludwiczek S, Aigner E, Theurl I, Weiss G. Cytokine mediatedregulation of iron transport in human monocytic cells. Blood2003; 101:4148–4154.

55. Pantopoulos K, Weiss G, Hentze MW. Nitric oxide and oxida-tive stress (H2O2) control mammalian iron metabolism bydifferent pathways. Mol Cell Biol 1996; 16:3781–3788.

56. Bouton C, Hirling H, Drapier JC. Redox modulation of ironregulatory proteins by peroxynitrite. J Biol Chem 1997;272:19969–19975.

57. Recalcati S, Taramelli D, Conte D, Cairo G. Nitric oxide-mediated induction of ferritin synthesis in J774 macrophagesby inflammatory cytokines: role of selective iron regulatoryprotein-2 downregulation. Blood 1998; 91:1059–1066.

58. Caltagirone A, Weiss G, Pantopoulos K. Modulation of cellulariron metabolism by hydrogen peroxide. Effects of H2O2 on theexpression and function of iron-responsive element-containingmRNAs in B6 fibroblasts. J Biol Chem 2001; 276:19738–19745.

59. Pantopoulos K, Mueller S, Atzberger A, Ansorge W, StremmelW, Hentze MW. Differences in the regulation of iron regula-tory protein-1 (IRP-1) by extra- and intracellular oxidativestress. J Biol Chem 1997; 272:9802–9808.


60. Mulero V, Brock JH. Regulation of iron metabolism in murineJ774 macrophages: role of nitric oxide-dependent and indepen-dent pathways following activation with gamma interferonand lipopolysaccharide. Blood 1999; 94:2383–2389.

61. Watts RN, Richardson DR. The mechanism of nitrogen monox-ide (NO)-mediated iron mobilization from cells. NO interceptsiron before incorporation into ferritin and indirectly mobilizesiron from ferritin in a glutathione-dependent manner. Eur JBiochem 2002; 269:3383–3392.

62. Liu X, Hill P, Haile DJ. Role of the ferroportin iron-responsiveelement in iron and nitric oxide dependent gene regulation.Blood Cells Mol Dis 2002; 29:315–326.

63. Kim S, Ponka P. Effects of interferon-gamma and lipopolysac-charide on macrophage iron metabolism are mediated by nitricoxide-induced degradation of iron regulatory protein 2. J BiolChem 2000; 275:6220–6226.

64. Bouton C, Oliveira L, Drapier JC. Converse modulation ofIRP1 and IRP2 by immunological stimuli in murine RAW264.7 macrophages. J Biol Chem 1998; 273:9403–9408.

65. Weiss G, Goossen B, Doppler W, Fuchs D, Pantopoulos K,Werner-Felmayer G, Wachter H, Hentze MW. Translationalregulation via iron-responsive elements by the nitric oxide=NO-synthase pathway. EMBO J 1993; 12:3651–3657.

66. Kim S, Ponka P. Nitrogen monoxide-mediated control of ferri-tin synthesis: implications for macrophage iron homeostasis.Proc Natl Acad Sci USA 2002; 99:12214–12219.

67. Nicolas G, Chauvet C, Viatte L, Danan JL, Bigard X, Devaux I,Beaumont C, Kahn A, Vaulont S. The gene encoding the ironregulatory peptide hepcidin is regulated by anemia, hypoxia,and inflammation. J Clin Invest 2002; 110:1037–1044.

68. Nicolas G, Bennoun M, Devaux I, Beaumont C, GrandchampB, Kahn A, Vaulont S. Lack of hepcidin gene expression andsevere tissue iron overload in upstream stimulatory factor 2(USF2) knockout mice. Proc Natl Acad Sci USA 2001;98:8780–8785.

69. Nicolas G, Bennoun M, Porteu A, Mativet S, Beaumont C,Grandchamp B, Sirito M, Sawadogo M, Kahn A, Vaulont S.


Severe iron deficiency anemia in transgenic mice expressingliver hepcidin. Proc Natl Acad Sci USA 2002; 99:4596–4601.

70. Cornelissen CN, Sparling PF. Iron piracy: acquisition of trans-ferrin-bound iron by bacterial pathogens. Mol Microbiol 1994;14:843–850.

71. Brown JS, Holden DW. Iron acquisition by gram-positivebacterial pathogens. Microbes Infect 2002; 4:1149–1156.

72. Borges VM, Vannier-Santos MA, de Souza W. Subvertedtransferrin trafficking in Leishmania-infected macrophages.Parasitol Res 1998; 84:811–822.

73. Olakanmi O, Schlesinger LS, Ahmed A, Britigan BE. Intra-phagosomal Mycobacterium tuberculosis acquires iron fromboth extracellular transferrin and intracellular iron pools.Impact of interferon-gamma and hemochromatosis. J BiolChem 2002; 277:49727–49734.

74. Howe D, Mallavia LP. Coxiella burnetii infection increasestransferrin receptors on J774A.1 cells. Infect Immun 1999;67:3236–3241.


5

Inhibition of Erythropoiesis byInflammatory Cytokines

ROBERT T. MEANS, JR.

Hematology=Oncology Division, Department ofMedicine, Ralph H. Johnson VA Medical Center,

and the Medical University of South Carolina,Charleston, South Carolina, U.S.A.

INTRODUCTION

Inhibition of Erythroid Progenitors in thePathogenesis of the Anemia of Chronic Disease

The anemia of chronic disease is one of the most commonhematologic syndromes encountered in clinical medicine. Itis probably the most frequent type of anemia other than irondeficiency resulting from blood loss (1). When all the anemicpatients admitted to the medical center of an urban hospitalin Texas were evaluated during 4 months in 1985–1986(excluding those who were actively bleeding, undergoing

127

hemolysis, or had a diagnosis of a clonal hematologic disor-der), 52% met laboratory diagnostic criteria for the anemiaof chronic disease (low serum iron concentration and normalor elevated serum ferritin concentration) (2). The anemia ofchronic disease is traditionally associated with chronic infec-tious diseases like tuberculosis, empyema and lung abscess,osteomyelitis, subacute bacterial endocarditis, cellulitis,chronic fungal infections, and the human immunodeficiencyvirus (3,4); with chronic inflammatory diseases such as rheu-matoid arthritis or systemic lupus erythematosus (5–7); andwith cancer (8–10). While the majority of diseases associatedwith the anemia of chronic disease fall into one of these cate-gories, in one series 40% of patients with this form of anemialacked a traditional ‘‘chronic disease’’ (2).

For nearly 40 years, it has been known that there arethree primary pathogenetic processes involved in the anemiaof chronic disease—a modest shortening of red cell survivalthat creates an increased demand for red cell production, animpaired erythropoietic response to this demand, and theabnormalities of iron metabolism which are the diagnostichallmark of this condition. In both rheumatoid arthritis andcancer patients, a blunted erythropoietin response to anemiaappears to contribute to the second pathogenetic process(8,11). However, while patients with the anemia of chronicdisease do not achieve the increments in erythropoietin pro-duction observed in similarly anemic patients with iron defi-ciency, they do have greater circulating erythropoietinconcentrations than healthy individuals who are not anemic(11). This observation suggests that erythroid progenitorsalso exhibit some degree of ‘‘erythropoietin resistance’’ atthe cellular level. Recent studies correlating the inhibitederythropoietic response in children with cancer with theserum erythropoietin concentration support this concept (12).

Inflammatory Cytokines in the Anemia ofChronic Disease

One of the challenges in the investigation of the pathogenesisof the anemia of chronic disease has been to identify a

128 Means

mechanism that links the diverse associated syndromes to aunifying common pathogenesis. The observation that the pre-sence of this anemia correlates with the activity of the asso-ciated disease (5,13) led investigators to consider mediatorsof the immune and inflammatory responses such as tumornecrosis factor (TNF) (14), interleukin-1 (IL-1) (15), and theinterferons (IFNs) (16–18) as agents responsible for this syn-drome. These cytokines have been implicated in all of thepathophysiologic mechanisms associated with the anemia ofchronic disease, as described in Fig. 1 and in other chaptersof this book (19). The association between anemia etiology(the anemia of chronic disease vs. iron deficiency or anemiaof other etiologies) and serum concentrations of one of these

Figure 1 Contributions of cytokines to the pathogenesis of theanemia of chronic disease. EPO, erythropoietin; IFN, interferon;IL-1, interleukin-1; RBC, red blood cell; TGFb, transforming growthfactor b; TNF, tumor necrosis factor. (Reproduced with permissionfrom Ref. 19.)

Inhibition of Erythropoiesis 129

agents (TNF) is shown in Fig. 2. Concentrations of TNF, IL-1,and=or the IFNs are increased in patients with disordersassociated with the anemia of chronic disease (18,20–23)and in animal models of the anemia of chronic disease (24).Therapeutic administration of these cytokines to patientsmay result in anemia as well (25,26).

INHIBITION OF ERYTHROID PROGENITORS BYSPECIFIC CYTOKINES

Tumor necrosis factor, IL-1, and the IFNs have all beenreported to inhibit erythropoiesis in vivo and in vitro(20,26–35). In addition, the IFNs, TNF, and IL-1 have all beenreported to act in synergy to inhibit in vitro erythropoiesis(33,36). As a result of the well-described feedback loops inter-relating these cytokines, they may act sequentially as well.For example, TNF may stimulate IL-1 release from macro-phages, IL-1 in turn may stimulate gIFN release fromT-lymphocytes, and gIFN may then enhance the effects andthe production of IL-1 and=or TNF.

Figure 2 Serum TNF concentrations observed in anemiasyndromes. ACD, anemia of chronic disease; AOE, anemia ofother etiologies; FeD, iron deficiency anemia.

130 Means

Tumor Necrosis Factor

As observed above, exposure to TNF inhibits erythropoiesis,either in vivo (28,29) or in vitro (27). The specific character-istics of this inhibitory effect differ depending upon themodel in which it is studied. A technique originally devel-oped by Sawada et al. (37), which allowed the generationof large numbers of highly purified human erythroid col-ony-forming units (CFU-E) from peripheral blood cellsenriched for erythroid burst-forming units (BFU-E), wasused to investigate the effects of TNF on CFU-E colony for-mation in vitro. Recombinant human (rh) TNF inhibitedCFU-E colony formation by normal marrow light densitymononuclear (LDMN) cells (containing approximately 2CFU-E=1000 cells) in a dose-dependent fashion but did notinhibit colony formation by highly purified CFU-E generatedfrom peripheral blood cells (approximately 300 CFU-E=1000cells), indicating that the inhibitory effect of rhTNF is indir-ect and mediated by a marrow accessory cell (38). Selectivedepletion of T- and B-lymphocytes and of marrow adherentcells did not alter the inhibitory effect of rhTNF on marrowcells; however, depletion of marrow stromal elements by soy-bean agglutinin (SBA) (39) abolished the inhibitory effect.Coculture of cells from SBA-depleted marrow with marrowcells precipitated by SBA restored the inhibitory effect ofrhTNF, but coculture with marrow lymphocytes or mononuc-lear adherent cells did not. Cell-free conditioned mediumprepared from marrow cells exposed to rhTNF inhibitedcolony formation by highly purified CFU-E, indicating thatthe inhibitory effect of TNF is mediated by a soluble factorreleased from marrow accessory cells in response to rhTNF,and that these accessory cells are neither lymphocytes normacrophages (38).

In contrast, when a slightly less differentiated erythroidprogenitor was studied, the inhibitory effect of TNF on ery-throid colony formation appeared to be the result of directaction. Colony formation by erythroid progenitors generatedby the same technique used above, but incubated in semisolidmedium for 2 days less, was directly inhibited by TNF. In


these cells, TNF exerted its inhibitory effects through inter-ference with cell cycle regulation (40).

Interleukin-1

In anemic patients with rheumatoid arthritis, plasma IL-1 con-centrations are directly related to the severity of anemia (41).Interleukin-1 also inhibits erythroid progenitor colony forma-tion in vitro. In studies very similar to those previouslydescribed for TNF, the inhibitory effect of rhIL-1 on CFU-Ecolony formation is indirect and dependent on marrow acces-sory cells (42). In this case, however, depletion of T-lymphocytesablates the inhibitory effect of rhIL-1 on marrow CFU-Ecolonies. Coculture of highly purified CFU-E colonies (whichare not inhibited by rhIL-1) with autologous T-lymphocytes,but not with autologous marrow adherent cells, results in inhi-bition of colony formation in the presence of rhIL-1. The inhibi-tory effect of rhIL-1 is mediated by a soluble factor releasedfrom T-lymphocytes, and is abrogated by neutralizing antibodyto human gIFN.

Interferons

a-Interferon

Recombinant human aIFN, like rhTNF and rhIL-1, exerts anindirect effect on colony formation by humanCFU-E, inhibitingcolony formation by CFU-E derived from unpurified marrowcells but not by highly purified CFU-E. This effect is ablatedby T-lymphocyte depletion, but not by neutralizing antibodyto gIFN. Although aIFN does not inhibit colony formation byhighly purified CFU-E, it does enhance the inhibitory effect ofrhgIFN in a synergistic fashion (43). The inhibitory effect ofaIFN on erythropoiesis is mediated through apoptosis (44).

b-Interferon

In studies extending the previously described findings forTNF, the inhibitory effects of rhTNF-conditioned mediumon colony formation by highly purified CFU-E and of rhTNF

132 Means

itself on colony formation by unpurified marrow LDMN cellscould be ablated by neutralizing antibody to human bIFN,and bIFN could inhibit colony formation by highly purifiedCFU-E. Limiting dilution analysis demonstrated that thiswas a direct effect (45). However, the amount of bIFNdetected in rhTNF-conditioned medium by ELISA was toolow to account for the degree of inhibition observed, indicatingthat another factor (or factors), acting synergistically or in co-operation with bIFN, is also required for inhibition of CFU-Ecolony formation by TNF (45). Unpublished studies with ourcollaborators have demonstrated that highly purified CFU-Eexpress mRNA for the Type I IFN receptor (the shared recep-tor for a- and bIFN).

g-Interferon

The inhibitory effect of rhIL-1 is mediated by a soluble factorreleased from T-lymphocytes, and is abrogated by neutraliz-ing antibody to human gIFN. rhgIFN, in turn, directly inhibitscolony formation by highly purified CFU-E (42). This findingis confirmed by limiting dilution analysis. Furthermore, theinhibitory effect of rhgIFN on CFU-E can be ablated by a gIFNceptor immunoadhesion, consisting of a soluble human gIFNreceptor coupled to an IgG1 heavy chain (46).

gIFN appears to exert its effects on erythroid colonyformation by a caspase-mediated apoptotic mechanism (47).Ceramide is a product of cytokine-induced enzymatic hydroly-sis of cell membrane sphingomyelin, and appears to be anobligate messenger in the inhibitory effects of gIFN on ery-throid colony formation (48). Endogenous ceramide producedby exposure to bacterial sphingomyelinase (0.2–2.0U=mL) orto exogenous cell-permeable ceramide (C2-ceramide at con-centrations < 10 mM) significantly inhibits bone marrowCFU-E colony formation. This effect is reversed by the cera-mide antagonist, sphingosine-1-phosphate. Inhibition ofCFU-E by rhgIFN, but not rhbIFN, is also reversed by sphin-gosine-1-phosphate. In addition, recombinant human erythro-poietin, 10U=mL, reverses CFU-E inhibition by C2-ceramide10 mM. Exposure of marrow cells to rhgIFN produces a


significant increase in ceramide content. These findingsstrongly suggest that ceramide is involved in the inhibitionof human CFU-E colony formation by gIFN (48). The clinicalrelevance of these observations stems from a growing bodyof data indicating the importance of ceramide in the biologyof cancer (and hence, of cancer-associated anemia of chronicdisease), and as a mediator of apoptosis (49). Ceramide is fre-quently implicated in Fas-mediated processes; therefore, therole proposed for ceramide is consistent with reports indicat-ing that inhibition of erythroid colony formation by gIFNinvolves Fas (50).

Another novel approach to understanding the mechan-isms by which gIFN exerts its inhibitory effects involves eval-uating changes in growth factor receptors during erythroiddevelopment. Exposure to very high (2500U=mL) concentra-tions of rhgIFN in vitro results in a decrease in erythropoietinand stem cell factor (SCF), but not insulin-like growth factor-I receptors. This decrease is observed at both the protein andmRNA levels (51). This finding is only demonstrable atrhgIFN concentration significantly greater than those whichinhibit CFU-E colony formation (100–1000 U=mL); however,it is a mechanistically appealing explanation for the effectof gIFN.

Effects of Other Cytokines on ErythroidProgenitors

Transforming Growth Factor b

Transforming growth factor b1 (TGFb1) inhibits CFU-Ecolony formation by marrow mononuclear cells (52).

Interleukin-10

Interleukin-10 (IL-10) is a cytokine which has differing poten-tial contributions to the anemia of chronic disease, dependingon the context. In aplastic anemia, IL-10 appears to reversesuppression of BFU-E colony formation by suppression ofaccessory cell production of TNF and gIFN (53). Similarly,

134 Means

increased IL-10 production is associated with a decreased riskof severe anemia in children with malaria (54). Theseantianemia effects must then be contrasted with reports thatIL-10 administration is associated with anemia in patientswith chronic active Crohn’s disease, and that this anemiaexhibits abnormalities of iron metabolism consistent withthe anemia of chronic disease (55). As indicated in Fig. 1, mostof the cytokines implicated in the anemia of chronic diseaseexert proanemia effects on several of the pathogeneticprocesses involved. Interleukin-10 is unique, in that itappears to exert proanemic effects on iron metabolism, butantianemic effects on erythropoiesis.

EFFECTS OF ERYTHROPOIETIN AND OTHERCOLONY STIMULATING FACTORS ONCYTOKINE INHIBITION OF ERYTHROPOIESIS

Clinical administration of recombinant human erythropoietinor the long-acting analogue, darbepoietin [also called novelerythropoiesis-stimulating protein (NESP)], can correct theanemia of patients with rheumatoid arthritis or cancer(10,56,57). Erythropoietin can also reverse anemia in animalmodels of the anemia of chronic disease associated withincreased serum IL-1, TNF, and gIFN concentrations (24).The inhibitory effects of rhgIFN on colony formation byCFU-E in vitro can be reversed by high concentrations ofrecombinant human erythropoietin in vitro (58). However,the inhibitory effects of rhaIFN, rhbIFN (43), and TGFb1 invitro are not corrected by recombinant human erythropoietin.

Stem cell factor has also been reported to reverse the invitro inhibitory and proapoptotic effects of rhgIFN on BFU-Ecolony formation (59). Studies in our laboratory have shownthat rhSCF also partially reverses the inhibitory effects ofrhbIFN on CFU-E colony formation in vitro. Exposure ofhighly purified CFU-E to rhbIFN increases cellular expres-sion of the proapoptotic protein Bad. Stem cell factordecreases cellular Bad expression in both control and IFN-exposed cells (Fig. 3).


IMPLICATIONS FOR THERAPY

The principal clinical implication of the cytokine-mediatedmodel of the anemia of chronic disease is that it provides arational pathophysiologic basis for diagnosis in patients lack-ing a traditional ‘‘chronic disease.’’ It even explains why theanemia of chronic disease sometimes appears to developacutely.

The ability of recombinant human erythropoietin or ofNESP to reverse cytokine-mediated inhibition of erythropoi-esis in both in vitro and in vivo model systems is consistentwith the observed response of the anemia of chronic diseaseto therapy with erythropoietin products. However, the cyto-kine-specific differences observed experimentally have impli-cations for growth factor treatment in the anemia of chronicdisease. The significant differences in the response of inhibi-tion of CFU-E colony formation by g- and bIFNs to erythro-poietin, for example, may partially explain the variation inthe response of specific patients to recombinant human ery-thropoietin therapy. The dose–response pattern of inhibitionlikely underlies the clinical observation by Kaltwasser et al.(60) that disease activity is one of the two major predictors

Figure 3 Effects of stem cell factor (SCF) and bIFN on expressionof Bad by highly purified human CFU-E. Bars reflect densitometricquantification of Western blot data.

136 Means

of the response of anemic rheumatoid arthritis patients torecombinant human erythropoietin. Finally, studies suggest-ing a role for other hematopoietic growth factors such asSCF in reversing the anemia of chronic disease open thepossibility of combined cytokine therapy for this cytokine-induced syndrome.

ACKNOWLEDGMENTS

This study was supported in part by the U.S. Department ofVeterans Affairs Veterans Health Administration Researchfunds and grant HL HL69418 from the U.S. National Heart,Lung, and Blood Institute.

REFERENCES


2. Cash JM, Sears DA. The spectrum of diseases associated withthe anemia of chronic disease: a study of 90 cases. Am J Med1990; 87:638–644.

3. Cartwright GE, Wintrobe MM. The anemia of infection. XVII.A review. Adv Intern Med 1952; 5:165–226.

4. Mocroft A, Kirk O, Barton SE, Dietrich M, Proenca R,Colebunders R, Pradier C, d’Armino Montforte A, LedergerberB, Lundgren JD. Anaemia is an independent predictivemarker for clinical prognosis in HIV-infected patients fromacross Europe. AIDS 1999; 13:943–950.

5. Baer AN, Dessypris EN, Krantz SB. The pathogenesis ofanemia in rheumatoid arthritis. A clinical and laboratoryanalysis. Semin Arthritis Rheum 1990; 14:209–223.

6. Daniak N, Hardin J, Floyd V, Callahan M, Hoffman R.Humoral suppression of erythropoiesis in systemic lupuserythematosus (SLE) and rheumatoid arthritis. Am J Med1980; 69:537–544.


7. Voulgarelis M, Kokori SI, Ioannidis JP, Tzioufas AG, KyriakiD, Moutsopoulos HM. Anaemia in systemic lupus erythemato-sus: aetiological profile and the role of erythropoietin. AnnRheum Dis 2000; 59:217–222.

8. Miller CB, Jones RJ, Piantadosi S, Abeloff MD, Spivak JL.Decreased erythropoietin response in patients with the anemiaof cancer. N Engl J Med 1990; 322:1689–1692.

9. Skillings JR, Sridhar FG, Wong C, Paddock L. The frequencyof red cell transfusion for anemia in patients receivingchemotherapy. A retrospective cohort studys. Am J Clin Oncol1993; 16:22–25.

10. Smith RE, Jaiyesima IA, Meza LA, Tchekniediyan NS, ChanD, Griffith H, Brosman S, Bukowski R, Murdoch M, RarickM, Saven A, Colowick AB, Fleischman A, Gayko U, Glaspy J.Novel erythropoiesis stimulating protein (NESP) for the treat-ment of the anemia of chronic disease associated with cancer.Br J Cancer 2001; 84(suppl 1):24–30.

11. Baer AN, Dessypris EN, Goldwasser E, Krantz SB. Bluntederythropoietin response to anaemia in rheumatoid arthritis.Br J Haematol 1987; 66:559–564.

12. Carazza F, Beguin Y, Bergmann P, Andre M, Ferstner A,Devalck C, Fonu P, Buyse M, Sariban E. Anemia in childrenis associated with decreased erythropoietic activity and notwith inadequate erythropoietin production. Blood 1998;92:1793–1798.

13. Jefferey MR. Some observations on anemia in rheumatoidarthritis. Blood 1953; 8:502–518.

14. Beutler B, Cerami A. Cachectin: more than a tumor necrosisfactor. N Engl J Med 1987; 316:379–385.

15. Dinarello CA. Interleukin-1 and its biologically relatedcytokines. Adv Immunol 1989; 44:153–205.

16. Vilcek J, Gray PW, Rinderknecht E, Sevastopoulos CG. Inter-feron g: a lymphokine for all seasons. Lymphokines 1985; 11:1–32.

17. Murray HW. Interferon-gamma, the activated macrophage,and host defense against microbial challenge. Ann InternMed 1988; 108:595–608.

138 Means

18. Hooks JJ, Moutsopopoulos HM, Geis SA, Stahl NI, Decker JL,Notkins AC. Immune interferon in the circulation of patientswith autoimmune disease. N Engl J Med 1979; 301:5–8.

19. Means RT. Recent developments in the anemia of chronicdisease. Curr Hematol Rep 2003; 2:116–121.

20. Maury CPJ, Andersson LC, Teppo A-M, Partanen S, JavonenE. Mechanism of anaemia in rheumatoid arthritis: demonstra-tion of raised interleukin 1b concentration in anaemic patientsand of interleukin 1 mediated suppression of normal erythro-poiesis and proliferation of human erythroleukemia (HEL)cells in vitro. Ann Rheum Dis 1988; 47:972–978.

21. Balkwill F, Burke F, Talbot D, Tavemier J, Osborne R, NaylorS, Durbin H, Fiers W. Evidence for tumor necrosis fac-tor=cachectin production in cancer. Lancet 1987; 2:1229–1232.

22. Teppo A-M, Maury CPJ. Radioimmunoassay of tumor necrosisfactor in serum. Clin Chem 1987; 33:2024–2027.

23. Mauritz NJ, Holmdahl R, Jonsson R, Van der Meide PH,Scheynius A, Klarskug L. Treatment with gamma-interferontriggers the onset of collagen arthritis in mice. ArthritisRheum 1988; 31:1297–1303.

24. Coccia MA, Cooke K, Stoney G, Pistillo J, Del Castillo J,Duryea D, Tarpley GA, Molineux G. Novel erythropoiesisstimulating protein (darbepoietin alfa) alleviates anemia asso-ciated with chronic inflammatory disease in a rodent model.Exp Hematol 2001; 10:1201–1209.

25. Blick M, Sherwin SA, Rosenblum M, Gutterman J. Phase Istudy of recombinant tumor necrosis factor in cancer patients.Cancer Res 1987; 47:2986–2989.

26. Vadhan-Raj S, Al-Katib A, Bhulla R, Pelus L, Nathan CF,Sherwin SA, Oettgen HF, Krown SE. Phase I trial of recombi-nant interferon gamma in cancer patients. J Clin Oncol 1986;4:137–146.

27. Roodman GD, Bird A, Hutzler D, Montgomery W. Tumornecrosis factor-alpha and hematopoietic progenitors: effectsof tumor necrosis factor on the growth of erythroid progenitorsCFU-E and BFU-E and the hematopoietic cell lines K562,HL60, HEL cells. Exp Hematol 1987; 15:928–935.


28. Johnson CS, Cook CA, Furmanski P. In vivo suppression oferythropoiesis by tumor necrosis factor-a (TNF-a): reversalwith exogenous erythropoietin (EPO). Exp Hematol 1990;18:109–113.

29. Johnson RA, Waddelow TA, Caro J, Oliff A, Roodman GD.Chronic exposure to tumor necrosis factor in vivo preferentiallyinhibits erythropoiesis in nude mice. Blood 1989; 74:130–138.

30. Maury CPJ. Anaemia in rheumatoid arthritis: role of cyto-kines. Scand J Rheumatol 1987; 18:3–5.

31. Johnson CS, Keckler DJ, Topper MI, Braunschweigger PG,Furmanski P. In vivo hematopoietic effect of recombinantinterleukin-1a in mice: stimulation of granulocytic, monocytic,megakaryocytic, and early erythroid progenitors; suppressionof late stage erythropoiesis, and reversal of erythroid suppres-sion with erythropoietin. Blood 1989; 73:678–683.

32. Mamus SW, Beck-Schroeder SK, Zanjani ED. Suppression ofnormal human erythropoiesis by gamma interferon in vitro:role of monocytes and T-lymphocytes. J Clin Invest 1985;75:1496–1503.

33. Raefsky EL, Platanias LC, Zoumbos NC, Young NS. Studies ofinterferon as a regulator of hematopoietic cell proliferation.J Immunol 1985; 135:2507–2512.

34. Zanjani ED, McGlave PB, Davies SF, Bonisadre M, KaplanME, Sarosi GA. In vitro suppression of erythropoiesis by bonemarrow adherent cells from some patients with fungal infec-tion. Br J Haematol 1982; 50:479–490.

35. Sugimoto M, Wakabayashi Y, Hirose S-I, Takaku F. Immuno-logical aspects of the anemia of rheumatoid arthritis. Am JHematol 1987; 25:1–11.

36. Broxmeyer HE, Williams DE, Lu L, Cooper S, Anderson SL,Beyer GS, Hoffman R, Rubin BY. The suppressive influencesof tumor necrosis factors on bone marrow hematopoietic pro-genitor cells from normal donors and patients with leukemia:synergism of tumor necrosis factor and interferon-g. J Immu-nol 1986; 136:4487–4495.

37. Sawada K, Krantz SB, Kans JS, Dessypris EN, Sawyer S, GlickAD, Civin CI. Purification of human erythroid colony-forming

140 Means

units and demonstration of specific binding of erythropoietin. JClin Invest 1987; 80:357–366.

38. Means RT, Dessypris EN, Krantz SB. Inhibition of humancolony-forming units erythroid by tumor necrosis factorrequires accessory cells. J Clin Invest 1990; 86:538–541.

39. Ebell W, Castro-Malaspina H, Moore MAS, O’Reilly RJ. Deple-tion of stromal elements in human marrow grafts separated bysoybean agglutinin. Blood 1985; 65:1105–1111.

40. Dai C, Chung I, Jiang S, Krantz S. Reduction of cell cycleprogression in human erythroid progenitor cells treated withtumor necrosis factor alpha (TNF) occurs with reducedCDK6 and is reversed by CDK6 transduction [abstr]. ExpHematol 2002; 30(suppl 1):111a.

41. Eastgate JA, Wood NC, DiGiovine FS, Symons JA, GrinlintonFA,DuffGW.Correlation of plasma interleukin1 levelswith dis-ease activity in rheumatoid arthritis. Lancet 1988; 2:706–709.

42. Means RT, Dessypris EN, Krantz SB. Inhibition of humanerythroid colony-forming units by interleukin-1 is mediatedby gamma interferon. J Cell Physiol 1992; 150:59–64.

43. Means RT, Krantz SB. Inhibition of human erythroid colony-forming units by interferons a and b: differing mechanismsdespite shared receptor. Exp Hematol 1996; 24:204–208.

44. Tarumi T, Sawada K, Sato N, Kobayashi S, Takano H,Yasukouchi T, Takashashi T, Sekiguchi S, Koike T.Interferon-alpha-induced apoptosis in human erythroid pro-genitors. Exp Hematol 1995; 23:1310–1318.

45. Means RT, Krantz SB. Inhibition of human erythroid colony-forming units by tumor necrosis factor requires beta inter-feron. J Clin Invest 1993; 91:416–419.

46. Means RT, Krantz SB, Luna J, Marsters SA, Ashkenazi A.Inhibition of murine erythroid colony formation in vitro bygamma interferon and correction by interferon inhibitor. Blood1994; 83:911–915.

47. Dai CH, Krantz SB. Interferon gamma induces upregulationand activation of caspases 1,3, and 8 to produce apoptosis inhuman erythroid progenitor cells. Blood 1999; 93:3309–3316.


48. Dallalio G, North M, Worden BD, Means RT. Inhibition ofhuman erythroid colony formation by ceramide. Exp Hematol1999; 27:1133–1138.

49. Hannun YA. Apoptosis and the dilemma of cancer chemotherapy. Blood 1997; 89:1845–1853.

50. Dai C-H, Price JO, Brunner T, Krantz SB. Fas ligand is pre-sent in human erythroid colony-forming cells interacts withFas induced by interferon g to produce erythroid cell apoptosis.Blood 1998; 91:1235–1242.

51. Taniguchi S, Dai C-H, Price JO, Krantz SB. Interferon gammadownregulates stem cell factor and erythropoietin receptorsbut not insulin-like growth factor-I receptors in humanerythroid colony-forming cells. Blood 1997; 90:2244–2252.

52. Hino M, Tojo A, Miyazono K, Urabe A, Takaku F. Effects of btransforming growth factors on hematopoietic progenitor cells.Br J Haematol 1988; 70:143–147.

53. Geissler K, Kabrna E, Kollars M, Ohler L, Berer A, BurgmannH, Winkler S, Willheim M, Hinterberger W, Lechner K.Interleukin-10 inhibits in vitro hematopoietic suppressionand production of interferon-gamma and tumor necrosis fac-tor-alpha by peripheral blood mononuclear cells from patientswith aplastic anemia. Hematol J 2002; 3:206–213.

54. Biemba G, Gordeuk VR, Thuma P, Weiss G. Markers ofinflammation in children with severe malarial anaemia. TropMed Intl Health 2000; 5:256–262.

55. Tilg H, Ulmer H, Kaser A, Weiss G. Role of IL-10 for inductionof anemia during inflammation. J Immunol 2002; 169:2204–2209.

56. Means RT, Olsen NJ, Krantz SB, Dessypris EN, Graber SE,Stone WJ, CNeil VL, Pincus T. Treatment of the anemia ofrheumatoid arthritis with recombinant human erythropoietin:clinical and in vitro studies. Arthritis Rheum 1989; 32:638–642.

57. Pincus T, Olsen NJ, Russell U, Wolfe F, Harris R, Schnitzer T,Abels R, Boccagno J, Krantz SB. Multicenter study of recombi-nant human erythropoietin in correction of anemia in rheuma-toid arthritis. Am J Med 1990; 89:161–168.

142 Means

58. Means RT, Krantz SB. Inhibition of human erythroid colonyforming units by gamma interferon can be corrected by recom-binant human erythropoietin. Blood 1991; 78:2564–2567.

59. Dai CH, Krantz SB, Kollar K, Price JO. Stem cell factor canovercome inhibition of highly purified human burst-formingunits-erythroid by interferon gamma. J Cell Physiol 1998;165:323–332.

60. Kaltwasser JP, Kessler U, Gottschalk R, Stucki G, Moller B.Effect of recombinant human erythropoietin and intravenousiron on anemia and disease activity in rheumatoid arthritis.J Rheumatol 2001; 28:2430–2436.


6

Endogenous Erythropoietin in theAnemia of Chronic Disorders

YVES BEGUIN

National Fund for Scientific Research,Division of Hematology,

Department of Medicine; and Laboratory of Celland Gene Therapy, CHU Sart-Tilman, Center for

Cellular and Molecular Therapy,University of Liege, Liege, Belgium

INTRODUCTION

Anemia of chronic disease (ACD) is defined as the anemiaassociated with infection, inflammation, cancer, or traumathat has the characteristic picture of hypoferremia, hyperfer-ritinemia, decreased transferrin concentration, and increasediron stores (1). The pathogenesis of ACD involves thecombination of a shortened erythrocyte survival in circulationwith failure of the bone marrow to increase red cell production

145

in compensation (2–7). Inappropriate red cell production is itselfrelated to a combination of factors, including impaired availabil-ity of storage iron, inadequate erythropoietin (Epo) response toanemia, and overproduction of cytokines, which are capable ofinhibiting erythropoiesis (2–5). These cytokines are involvedin the retention of iron in the reticuloendothelial system,gastrointestinal tract, and hepatocytes. Theymay interfere withEpo production by the kidney, and may exert direct inhibitoryeffects on erythroid precursors (3,4,8–12). Indeed, theireffect is much wider, involving the whole hematopoieticsystem (13).

Cancer is one of the leading causes of ACD. However,the anemia observed in cancer patients may have multiplemechanisms (2,14,15). Hemodilution may artificially dilutethe red cell mass. Bleeding, autoimmune or microangio-pathic hemolysis, hypersplenism, and hemophagocytosismay all reduce the red cell life span. Nutritional deficiencies,including iron, folate, vitamin B12, and global malnutrition,may impair red cell production. The bone marrow may beinvolved by metastases, necrosis, myelodysplasia, and auto-immune red cell aplasia. These various causes, not includingthe ‘‘anemia of chronic disorders,’’ have been reviewed indetail elsewhere (16). Surprisingly, there are no reports onthe relative proportion of cancer patients in general or ofpatients with any form of cancer, in particular, that presentthe typical features of ACD. In other words, the true inci-dence of ACD in cancer patients is completely unknown.Hence, the relevance of the biologic features of ACD to theoverall erythropoietic activity of cancer patients remainselusive.

Furthermore, chemotherapy and radiotherapy have amajor impact on the incidence and severity of anemia incancer patients. Compared to untreated cancer alone,chemotherapy may double the incidence of anemia (17). Theincidence and severity of anemia largely depends on the formof cancer as well as the type and dose intensity of chemother-apy administered to patients (18). This is also true in childrenwhere the incidence of chemotherapy-induced anemia mayeven be greater because of the nature of the cancer being

146 Beguin

treated (many leukemias) and of the relative intensity oftherapies applied (19). Various models have mostly identifiedolder age, lower baseline Hb, and rapid drop of Hb afterthe first cycle as additional factors that are predictive oftransfusion requirements in patients receiving chemotherapy(20–23). Chemotherapy may directly affect erythropoiesisinthe bone marrow and also impact on endogenous Epoproduction.

In this review, we will examine the evidence for defectiveEpo production in patients with ACD. We conducted a wideliterature survey on the topic and critically analyzed thepapers identified in this search. Solid experimental data indi-cate that several cytokines interfere with Epo production.However, it is unclear how these data can be directly appliedin vivo. Many clinical papers reporting serum Epo levels invarious disorders associated with ACD, in particular in can-cer patients, have methodological problems. Two major suchproblems can be identified. The first problem relates to theheterogeneity of the patients studied in terms of diseaseand stage of the disease, as well as the simultaneous inclusionof patients at diagnosis, during treatment and after comple-tion of therapy. The second problem involves the interpreta-tion of serum Epo levels in individual patients or in groupsof subjects, with lack of appropriate controls and inadequateinterpretation of Epo data.

Therefore, we will first present the experimental dataon the effect of various cytokines on Epo production. Second,we will comment on appropriate methods allowing interpre-tation of serum Epo levels in patients. We will then reviewthe evidence for defective Epo production in patients withACD, focusing in particular on HIV (as a model of chronicinfection), rheumatoid arthritis (as a model of chronicinflammatory disorder), and cancer. In the case of cancer,we will attempt to examine various diseases independentlywhenever possible, and we will try to delineate the respec-tive roles of cancer itself and of chemotherapy. Finally, wewill illustrate how baseline serum Epo levels can helppredict response to recombinant human erythropoietin(rHuEpo) therapy.

Endogenous Erythropoietin 147

EFFECTS OF CYTOKINES ON ERYTHROPOIETINPRODUCTION (Table 1)

Peripheral blood mononuclear cells from patients withchronic renal failure released soluble factors that suppressedEpo production by HepG2 cells, but these factors did notappear to be TNF-a or IL-1 (24). Neopterin also induces a sup-pression of hypoxia-induced Epo synthesis in HepG2 cells in adose-dependent manner (25). It has been reported that IL-1a,IL-1b, TNF-a, IFN-g, and TGF-b inhibited, whereas IL-6 sti-mulated, cobalt-induced or hypoxia-induced Epo productionat the mRNA level by the hepatoma cell line Hep3B (26,27).The inhibitory effect of IFN-g was found to be additive to thatof IL-1 and even synergistic with that of TNF-a, and wascapable of preventing any response to IL-6 (27). The sameinhibition of Epo gene expression and protein productionwas observed with the HepG2 line for IL-1 and TNF but notfor TGF-b, IFN-g, or IL-6 (28–30). Contrary to phorbol-ester-induced inhibition of Epo production, inhibition by IL-1b or TNF-a was independent of protein kinase C (31). Inhibi-tion of hepatic Epo production by TNF-a appears to bemediated by the 55kDa (TNF-RI) rather than the 75kDa(TNF-RII) receptor (32). IL-1, TNF-a, and IL-6 also blockedhypoxia-induced Epo formation by the isolated rat kidney (29).

Cytokine-induced inhibition of Epo production by HepG2cells is not mediated by impairment of hypoxia-inducedfactor-1 (HIF-1) whose activity is rather enhanced by IL-1b

Table 1 Effect of Various Cytokines on Epo Production by theHepatoma Cell Lines Hep 3B and G2, by the Isolated Rat Kidneyand in Vivo

Hep 3B Hep G2 Rat kidney In vivo

IL-1 # # # #TNF-a # # #TGF-b # ¼ ¼ #IL-6 " ¼ # # "IFN-g # ¼

148 Beguin

or TNF-a, while VEGF expression remains unaffected (33).Several cytokines stimulate inducible nitric oxide (NO)synthase gene expression in several tissues. It is thereforenot surprising that NO donors dose dependently reducedEpo production in the HepG2 cell line, either by directlyinfluencing the cellular redox state or by increasing reactiveoxygen species in the cell (34). Indeed, reactive oxygen spe-cies, including H2O2, have been shown to suppress the invitro synthesis of Epo (35). H2O2, whose production is reducedin hypoxic conditions, has been proposed as a potential signal-ing molecule between the oxygen sensor and the transcrip-tional machinery (35). Desferrioxamine and cobalt chlorideantagonize the inhibition of Epo production by reactiveoxygen species, by reducing the action of H2O2, and by inter-fering with its production and=or scavenging, respectively(36). Similarly, the antioxidant vitamins A, E, and C signifi-cantly increased Epo production by the hypoxic isolated ratkidney (37). While vitamin A also dose dependently increasedEpo synthesis in Epo-producing hepatoma cell cultures, vita-mins E and C had no such effects (37). In another experimentin which Epo synthesis by HepG2 cells was reduced by mono-cyte-conditioned medium as well as IL-1b, TNF-a, and IL-6,dexamethasone decreased cytokine secretion by monocytesbut did not affect Epo production on its own (38).

Injection of bacterial lipopolysaccharide (LPS) or IL-1b tonormoxic or hypoxic rats resulted in increased TNF-a mRNAand reduced Epo mRNA in the kidney, as well as decreasedserum Epo levels (39). In vivo administration of TGF-b wasassociated with depressed serum Epo levels in one study(40) but not in another (41). Administration of IL-6 to cancerpatients resulted in elevated serum Epo levels that paralleledthe development of anemia (42). Treatment of patients withchronic active hepatitis B with interferon-a resulted in a tran-sient increase in plasma Epo levels (43). The exogenousadministration of rHuEpo to mice treated with IL-1 was ableto correct the suppression of CFU-E as well as of other ery-throid parameters (44–46). Erythropoietin could also reversethe anemia of mice treated with single injections of TNF(47) but not always when mice were continuously exposed


to TNF (47–49). Exogenous Epo was nevertheless capable ofpreventing the anemia induced by TGF-b (40).

INTERPRETATION OF SERUM Epo LEVELS

What Is a Normal Epo Value?

Erythropoietin production is regulated through a feedbacksystem between the bone marrow and the kidney, whichdepends on a renal oxygen sensor (50,51). The capacity ofthe kidney to respond acutely to hypoxia by increasingEpo production may be modulated by prior sensitization.Post-transfusion polycythemic mice exposed to hypoxia (52)or cobalt chloride (53) did not show the increased rate ofEpo production observed in normal animals (52). Mice madepolycythemic by exposure to intermittent hypoxia showed anapparent sensitization of Epo-producing cells to hypoxic sti-muli, explaining their greater Epo response to acute hypoxia,dexamethasone, testosterone, or isoproterenol, compared tohypertransfused mice (54–56). This was true for renal butnot for extrarenal Epo production (57).

Serum Epo levels may vary considerably (51,58). Levelsare usually between 10 and 20mU=mL in normal subjects,may decrease somewhat in primary polycythemia, butincrease exponentially when an anemia develops below anHct of 30–35% (59). Therefore, a serum Epo value mustalways be evaluated in relation to the degree of anemia(Figs.1 and 2) (51). In addition, it should be compared toappropriate reference subjects who should display a normalEpo response to anemia, including patients with iron defi-ciency or hemolytic anemia (see below). Erythropoietin levelsinappropriately low for the degree of anemia are encounterednot only in renal failure (60), but also in a number of otherconditions, including the anemia of chronic disorders (2,3).Inappropriately high serum Epo levels are often observed insecondary polycythemia, a feature permitting its diagnosticseparation from primary polycythemia (61).

Serum Epo levels increase exponentially in proportion tothe degree of anemia. We thus constructed reference regres-sions representing the normal relationships between Hct on

150 Beguin

Figure 2 Interpretation of endogenous serum Epo levels. Thehatched area represents the 95% confidence limits of the regressionof Epo vs. Hct in an appropriate group of reference subjects, e.g.,patients with IDA (2). However, serum Epo also depends on ery-thropoietic activity, with elevated and reduced levels in patientswith low [aplastic anemia (1)] or high [thalassemia intermedia(3)] erythropoietic activity, respectively. A group of patients withACD (4) shows a blunted Epo response to anemia.

Figure 1 Interpretation of endogenous serum Epo levels. Anindividual serum Epo value of 100mU=mL (dotted line) can be inter-preted in relation with the degree of anemia through the O=P ratio.For an Hct of 30%, this Epo value is adequate (O=P ratio ¼ 1.00),but forHct of 23% or 37%, the same absolute Epo value would be defec-tive (O=P ratio¼ 0.70) or excessive (O=P ratio¼ 1.30), respectively.


one hand and Epo on the other, based on normal subjects andpatients with hemolytic anemia (Fig. 2) (62). Two differentregression equations were described for Hct > or < 40%. Thiscutoff Hct was chosen because it allowed for the best correla-tion for Epo data and because of literature data indicating thatbeyond such an Hct there is little modification of Epo levels.For Hct below 40%, the following regression (R¼ –0.83,P¼ 0.0000) was obtained between Epo (mU=mL) and Hct(%): log(Epo)¼ 3.420–(0.056 Hct). For Hct over 40%, theregression equation (R¼ –0.12, NS) was: log(Epo)¼ 1.311–(0.003 Hct). Based on these formulas, predicted log(Epo)values were derived for each Hct, O=P ratios of observed=predicted log(Epo) were derived, and 95% confidence limitswere obtained in order to define a range of reference valuesfor individual O=P ratios (Fig. 1). These limits are 0.80–1.20for O=P Epo (62).

The adequacy of Epo production can thus be evaluated bytwo methods. When investigating a group of patients, this canbe achieved by comparing patients and appropriate referencesubjects by regression analysis (Fig. 2) (63). In this case, oneshould ensure that the study group encompasses a range ofHct values similar to that of the reference group; otherwise,the slopes of the regressions may be flawed. When studyingan individual patient, the adequacy of Epo production can beevaluated by the O=P ratio (Fig. 1) (62). An O=P ratio below0.80 indicates inadequate Epo production for the degree ofanemia even if the absolute Epo value is high. It should beemphasized that the specific regression equations obtainedin our study, on which O=P Epo ratios are based, cannot beautomatically transposed to any other study. One must firsteither ensure that the Epo assay used yields Epo values simi-lar to those measured in our Epo assay or construct one’s ownreference regressions with appropriate reference subjects.

Serum Epo Levels and Erythropoietic Activity

Many studies have reported higher serum Epo levels inpatients with low compared to high erythropoietic activity

152 Beguin

(Fig. 2). In an early study, urinary Epo secretion was similarin patients with marrow failure or hemolysis (64). The slope ofthe correlation between Epo and Hb was steeper for patientswith iron deficiency anemia (IDA) compared to those withaplastic anemia or transient erythroblastopenia, becauseEpo values in moderately anemic subjects were higher inthe latter group (65). For similar degrees of anemia, patientswith aplastic anemia had higher serum Epo levels thanpatients with iron deficiency or hemolytic anemia (66). Atany hemoglobin value, serum Epo levels in patients with purered cell aplasia were fourfold higher than in those with IDA,and tenfold higher than in patients with megaloblastic orsickle cell anemia (67). In 34 patients with aplastic anemia,serum Epo levels were much higher than in patients with irondeficiency at similar degrees of anemia (68). The same conclu-sions were obtained in another group of 42 patients with idio-pathic aplastic or Fanconi’s anemia (69). One log higherserum Epo values were encountered in patients with ery-throid hypoplasia or aplasia (erythropoietic activity <0.6times normal) compared to subjects with thalassemia inter-media (erythroid activity >2 times normal) (70). To accountfor this effect of erythroid activity, serum Epo levels can becorrected by the ratio of the sTfR (a quantitative marker oferythropoietic activity) value in the patient relative to a nor-mal sTfR value (70). High serum Epo levels are also observedtransiently after intensive chemotherapy, whether followedby bone marrow transplantation or not, without concomitantchange in hemoglobin or hematocrit (70–76). The peak Epovalues are observed 7 days after transplant, i.e., about 14days after the start of the conditioning regimen, at the timeof the nadir of erythropoietic activity. Within 24–72 hr afterstarting IV iron therapy in patients with IDA, markeddecreases in serum Epo were found before any change in Hb(70). Similar observations were obtained with rHuEpo ther-apy in pure red cell aplasia (70) with vitamin B12, or folatetherapy in megaloblastic anemia (70,77–79).

These findings thus point to an inverse relationshipbetween marrow erythropoietic activity and serum Epo levels:the higher the number of erythroid precursors, the lower the


serum Epo value. As Epo exerts its action on target cells afterbinding to a specific Epo receptor (80), it is tempting to spec-ulate that serum Epo levels may partly depend on the rateof Epo utilization by Epo receptor-bearing cells, primarilyerythroid precursors (70,81). Similarly, marrow recoveryafter autologous stem cell transplantation (ASCT) wouldrestore Epo utilization by erythroid cells, thus progressivelyreturning Epo levels to a range appropriate for the degreeof anemia (76). In patients with particularly fast engraftment,the duration of this correction phase is much shorter and mayeven finally lead to decreased Epo levels (76).

The idea that marrow utilization influences serum Epolevels was initially based on the observation that radiation-induced marrow hypoplasia was associated with a slowerdecline of serum Epo levels induced by hypoxia (82). However,the rate of Epo disappearance from the plasma of dogs withnormal, hypoplastic, or hyperplastic marrow, was later shownto be similar, regardless of the experiment was performedin nephrectomized (83) or unmanipulated (84) animals.Nephrectomy or hepatectomy does not influence the pharma-cokinetics of a large dose of native Epo (85) or a tracer dose ofrHuEpo (86). Organ accumulation in the kidney and bonemarrow of rats was minimal after intravenous injection of atracer dose of rHuEpo (87,88). Furthermore, erythropoietinlife span was similar in normal rats and in rats with bonemarrow suppressed by cyclophosphamide or hypertransfusionor stimulated by hemolysis or bleeding (89). Similar conclu-sions were reached in mice 48hr after initiation of hemolysis,bleeding or marrow suppression by 5-FU, or 2–24hr afterstarting rHuEpo therapy, although the delay between induc-tion of the desired experimental condition and measurementof Epo life span appears to be rather short (90). However, innormal human subjects (91,92) as well as in rats (93), theinitial clearance of rHuEpo is decreased when the dosesinjected are increased, approaching a plateau at high doses.Furthermore, a surge in serum Epo levels after intense phle-botomy translates into decreased clearance of a tracer dose ofrHuEpo (94). On the other hand, the pharmacokinetics ofrHuEpo in hemodialysis patients was not different before

154 Beguin

and after 6weeks of treatment with rHuEpo (95). In other stu-dies, rHuEpo appeared to be eliminated from the plasma morerapidly after multiple doses than after a single dose in normalvolunteers (96), whereas the elimination half-life of rHuEpowas increased on day 8 after two injections of rHuEpo to nor-mal volunteers (97). The clearance of radiolabeled rHuEporemained unchanged in rats injected with or without previousinjections of unlabeled rHuEpo for 19 days (98) but wasincreased in sheep 8days after experimental bleeding to Hblevels of 3–4 g=dL, before returning to baseline 4weeks later(94). It was also progressively decreased in sheep following5-FU- or busulfan-induced marrow ablation (99). Therefore,variations observed in serum Epo levels after intensivechemotherapy cannot simply be explained by changes inEpo consumption by the bone marrow.

The abnormal persistence of elevated plasma Epo levelsin rats after cessation of intensive rHuEpo treatment givenfor 20 days could relate to suppression of erythroid activity(100). However, this was contradicted by our experiment withhypertransfused rats, in which polycythemia resulted inappropriate reduction rather than elevation of serum Epolevels, with subsequent depression of erythropoietic activity(100). Therefore, it is unlikely that persisting elevated Epolevels were due to nonutilization by a severely depressederythroid marrow. Alternatively, Bozzini et al. (101) havesuggested the existence of a yet unidentified feedbackmechan-ism between Epo-responsive erythroid cells and Epo-produ-cing cells. Cobalt- or hypoxia-induced Epo production innormocythemic mice is increased when erythropoiesis isacutely depressed and reduced when erythropoiesis is recentlystimulated (101–103). Plasma Epo levels during hypoxia inmice with 5-FU- or irradiation-induced aplasia were higherthan in normal mice (104). On the other hand, hypoxia-induced Epo response in transfused polycythemic mice ismuch higher when erythropoiesis has been previously stimu-lated for prolonged periods of time (101–103). These appar-ently contradictory observations in normal and polycythemicmice may be reconciled if it is a retracting erythron that caninduce this Epo-hypersecretory state (101). However,


although the erythron must shrink more after rHuEpo-thantransfusion-induced polycythemia, it is unclear howhypoxia-induced Epo production would be relevant to ourobserved discrepancy in serum Epo levels between the twoconditions.

In conclusion, serum Epo levels are the result of a bal-ance between the rate of Epo production and its utilizationby the erythroid marrow (Fig. 3). This should also be takeninto account when interpreting the adequacy of a serumEpo value in various situations. Whereas it is indisputablethat marrow erythropoietic activity independently influencesserum Epo levels, it remains to be determined whether theerythroid precursor mass acts directly by utilizing circulatingEpo or indirectly by influencing the rate of Epo production.Some other factors linking the erythron to Epo productionmay also exist. For instance, products resulting from red cellhemolysis may indirectly stimulate marrow erythropoieticactivity as well as renal Epo production (105,106).

SERUM ERYTHROPOIETIN IN ANEMIA OFCHRONIC DISORDERS

Serum Epo levels have been examined in a variety of diseasesassociated with the anemia of chronic disorders. Rather than

Figure 3 Serum Epo levels are the result of a balance betweenEpo production in the kidney and Epo utilization by the erythro-poietic marrow. It remains to be determined whether the erythroidprecursor mass acts directly by utilizing circulating Epo or indir-ectly by influencing the rate of Epo production.

156 Beguin

producing an exhaustive list of papers encompassing thewhole spectrum of diseases that have been investigated forthe adequacy of Epo production, we will focus on specificexamples that have been particularly well documented. HIVinfection will be taken as a model of chronic infection andrheumatoid arthritis as a paradigm for chronic inflammatorydiseases. We will then turn to the more complex analysis ofthe data in the field of cancer.

Serum Epo in HIV Infection

Anemia is a common problem in human immunodeficiency(HIV) infection, being present in 70–95% of patients withAIDS, and frequently exacerbated by therapeutic agents suchas zidovudive (107,108). Severe in vitro inhibition of erythro-poiesis and transient stimulation of granulopoiesis areobserved after bone marrow infection with various HIVisolates (109). Several papers have examined the adequacyof endogenous Epo response to anemia in AIDS patients.Serum Epo levels were elevated in HIV-seronegative andHIV-seropositive asymptomatic homosexuals and in patientswith lymphadenopathy, AIDS-related complex (ARC) andAIDS, but were normal in asymptomatic HIV-seronegativeor HIV-seropositive intravenous drug users (110). However,no attempt was made to correlate these Epo values to Hb orHct values. Serum Epo levels were higher in HIV-infectedsubjects compared to normal individuals but again no controlanemic group was available for proper evaluation (111). Theregression line of serum Epo vs. Hb was quite similar inasymptomatic HIV-infected and uninfected 12-month oldinfants (112). HIV-infected subjects with AIDS or ARC notreceiving zidovudine therapy exhibited a strong inverse rela-tionship between serum Epo and Hb, but there was nocomparison with a control group (113). In a group of 82HIV-positive subjects, 41% of whom were receiving azidothy-midine antiviral therapy, the slope of the regression of serumEpo vs. Hb was less steep than in a control group of patientswith iron deficiency or aplastic anemia (114). However, fewHIV-infected subjects were anemic and no details are


available on the range of Hb values in the controls comparedto the study subjects. There are only two papers for which thedata can be fully interpreted and both indicate a blunted Epoproduction in patients with AIDS. Among 152 patientsinfected with HIV, anemia was present in 18% of asympto-matic, 50% of ARC and 75% of AIDS patients (115). The rela-tionship between serum Epo and Hb disclosed a markedlyblunted Epo response to anemia in AIDS patients comparedto patients with IDA. The serum Epo–Hb relationship in agroup of 42 patients with either ARC or AIDS, including 13patients on zidovudine, closely resembled that of patientswith the anemia of chronic disorders due to chronic infection,and both were considerably blunted compared to the relation-ship in subjects with iron deficiency (116). In addition, ironmetabolism reflected a pattern of ACD with low transferrinsaturation and elevated serum ferritin concentration. Forany given degree of anemia, patients treated with zidovudinehad significantly higher serum Epo concentration thanzidovudine-naive patients (111,113,115). Indeed, the anemiaassociated with zidovudine therapy appeared to be due tored cell hypoplasia or aplasia (117). This occurred in thepresence of elevated serum Epo values that again were notevaluated in relation to the degree of anemia in one study(117) but in another investigation even surpassed the Eporesponse of subjects with IDA (115). In conclusion, althoughthe number of studies is limited, endogenous Epo responseappears to be somewhat blunted in AIDS patients, but serumEpo levels are increased by zidovudine therapy.

Serum Epo in Rheumatoid Arthritis

The pathogenesis of anemia in systemic autoimmune diseases,including a possible defect in endogenous Epo production, hasbeen reviewed elsewhere (118). In addition to the effect of cyto-kines on Epo-producing cells, vascular interstitial damage inthe kidney peritubular cell area has been suggested as a causeof Epo deficiency in at least some of these systemic autoim-mune disorders (118). There is some evidence for impaired ery-thropoietin response to anemia in rheumatoid disease (119).

158 Beguin

Increased Epo levels were observed in RA patients thatremained anemic over the years compared to nonanemicpatients, but no appropriate control group was included inthe study (120). In another study of 50 RA patients, serumEpo levels were slightly increased over normal values but‘‘unrelated to low Hb concentration,’’ but the data were notcompared to an appropriate anemic control group (121). In agroup of 14 anemic RA patients, serum Epo levels weresignificantly higher in those classified as IDA on the basis ofabsent iron stores in the bone marrow than in those classifiedas having ACD, but their Hb was a little lower as well (122).Among 58 patients with rheumatoid arthritis, 40 were anemicand 26 were classified as ACD and 14 as IDA (123). With simi-lar averageHb values in the two groups, serumEpo concentra-tion was slightly but not significantly higher in the IDA group.Within a group of 67 RA patients, 20 patients judged to haveIDA based on reduced serum ferritin concentration had higherserum Epo levels than 24 other patients with normal or ele-vated ferritin concentration, while Hb values covered a similarrange in the two groups (124). Among 136 patients with rheu-matoid arthritis, 75 cases were anemic and a definitive causewas apparent in 65 of them (125). The majority (n¼ 43) hadACD and 15 had iron deficiency. Their Hb values were similarand correlated inversely with serum Epo, but Epo was signifi-cantly lower in those with ACD. Yet in another study, anevaluation of stainable bone marrow iron allowed the classifi-cation of 35 RA patients into ACD or IDA categories (126). Asignificant problem with all these studies is the absence of acontrol group with pure IDA instead of RA patients withIDA. The first of a few studies to compare RA patients witha control group with IDA came up with a relatively bluntedEpo response to anemia in RA patients, but the control groupdid not have the same range of Hb values as the study groupand the comparison is therefore not entirely valid (127). Inanother such study, serum Epo in both iron replete and irondeficient RA patients remained within the 95% confidence lim-its of the regression obtained in patients with iron deficiencyor hemolytic anemia, but there was no clear inverse correla-tion with the Hb values in either group (128). However, the


range of Hb values was obviously different in the control andRA groups, respectively. In a third study, among 97 anemicRA patients, serum Epo levels were lower in those with serumferritin concentrations greater than 20 mg=L despite similarHb values (129). In addition, at comparable Hb levels, serumEpo levels in RA patients with IDA were significantly lowerthan in IDA controls without RA. In another report, the aver-age serum Epo value was lower in RA patients than in IDAcontrols at similar average Hb (130). In a final study, theEpo response to anemia was clearly diminished in patientswith RA, both iron replete and iron deficient, compared to sub-jects with pure IDA (131).

On the other hand, in childrenwith systemic-onset juvenilechronic arthritis (JCA), defective iron supply for erythropoiesisrather than inadequate endogenous erythropoietin productionappears to be involved in the pathogenesis of anemia (132).Neither O=P Epo ratios nor regression analysis evidenced anydefect in endogenous Epo production in this group of children.Indeed, in childrenwith systemic, oligoarticular or polyarticularJCA, serum Epo levels were similar to those of patients withiron deficiency and similar degrees of anemia, while transferrinsaturation was low and serum ferritin ranged from irondeficiency to considerably elevated values (133).Whereas severeanemia associated with active systemic-onset juvenile rheuma-toid arthritis can be successfully treatedwith rHuEpo (134), thiscan also be achieved with IV iron alone (135). Some response toiron has been observed in RA as well (136). In addition, treat-ment of chronic disease in rheumatoid arthritis with TNF-ablockade resulted in dose-dependent Hb increments accompa-nied by a reduction of serum Epo concentration that suggestthat TNF-a directly affected bone marrow precursors ratherthan suppressed Epo production (137).

In conclusion, rheumatoid arthritis patients often haveblunted Epo response to anemia. This is much more promi-nent in those patients with other biological features of ACDthan in those predominantly with IDA. However, thesefindings are not necessarily transposable to other systemicautoimmune disorders, as, for instance, children with juvenilearthritis have normal Epo response to anemia.

160 Beguin

SERUM ERYTHROPOIETIN IN CANCER

Initial Studies

Earlier studies suggested that the anemia of inflammationproduced in rats (138) and the anemia of cancer in mice(139) were accompanied by an inappropriate erythropoietinresponse for the degree of anemia. Similar conclusions werederived from studies measuring serum Epo levels by bio-assay in patients with anemia secondary to chronic infec-tion or malignancy, including Hodgkin’s and non-Hodgkin’slymphoma, multiple myeloma, and solid tumors (140–142).However, this was not observed in other studies of tumor-bearing rats (143) and other human investigations found thatserum Epo levels were diminished relative to expected levelsonly in patients with infection or inflammation but not inthose with malignancies (144). Similarly, normal results werederived from studies of patients with cancer of the uterinecervix (145), renal cell carcinoma (146), and disseminatedlung carcinoma (147).

When radioimmunoassays became available, a studyshowed that, compared to controls suffering from bloodlosses, iron deficiency, hemolysis or pernicious anemia,patients with hematologic malignancies under treatmentwith chemotherapy displayed a normal relationship betweenhematocrit and serum Epo levels (148). Similarly, in amouse model of experimental melanoma, serum Epo concen-trations remained adequate for the degree of anemia untilterminal stages of the disease when the animals becameseverely cachectic (149). However, in a small group ofpatients with miscellaneous solid tumors, the averageserum Epo value was less than in IDA controls with similaraverage Hb (130). An important study was conducted in 81anemic patients with solid tumors in which it was foundthat for any given degree of anemia serum Epo levels werelower as compared to a group of control patients with IDA(150). In addition, the expected inverse relationship betweenserum Epo and hemoglobin was absent, but this was due toa small group of about 10 patients with inappropriate Eporesponse while all others were within the normal range.


In addition only 22 patients were untreated and it wasshown that the Epo response was further decreased by che-motherapy, often including cisplatin. Adequate Epo produc-tion was restored in the presence of hypoxia but the possiblerole of infections in some patients was not addressed. Withall these limitations, this study was taken as a landmarkfrom which it is now widely believed that Epo productionis defective in patients with cancer and that this is themajor cause of anemia in them. However, the picture ismuch less clear than that.

Studies in Patients Scheduled for rHuEpo Therapy

Several investigations have been carried out in patientsstarting rHuEpo therapy. However, inclusion of manypatients receiving chemotherapy may yield inaccurate con-clusions about the adequacy of Epo production in cancerpatients (see below). For instance, the majority of 12patients with solid tumors selected for rHuEpo therapy, sev-eral of them receiving chemotherapy, had inappropriatelylow serum Epo levels (151). In a large study of anemic can-cer patients selected to be treated with rHuEpo, Epo levelsfor any Hb value were significantly lower in patients receiv-ing cisplatin-based compared to noncisplatin chemotherapy(152,153). In another trial of rHuEpo for cisplatin-associatedanemia, serum Epo levels were said to be inappropriatelylow for the degree of anemia and not to correlate withhemoglobin levels, but no detailed data were available tosubstantiate this statement (154). A large study of transfu-sion-dependent chemotherapy-treated patients with multiplemyeloma or low-grade non-Hodgkin’s lymphoma showed thathalf of them had inappropriate Epo levels before startingrHuEpo therapy (155). This was also the case in anotherstudy of similar patients not requiring transfusions, inwhich the majority of the patients were found to have inade-quately low serum Epo levels before the start of rHuEpo (156).In a multicenter study of patients selected for rHuEpo therapyfor nonplatinum chemotherapy-induced anemia (157), serumEpo levels correlated inversely with baseline hemoglobin and

162 Beguin

appeared to be inappropriate for the degree of anemia only in asmall minority of the patients. Another small study foundinadequate Epo production as evidenced by low O=P Epo ratioin eight patients with lymphoma or multiple myeloma receiv-ing chemotherapy, two of whom had mild degrees of renal fail-ure (158). However, another investigation of six similarpatients has found no evidence of defective Epo secretion inthese disorders (159).

Solid Tumors

There are only few studies examining untreated patientswith solid tumors. Among 84 such patients, only 13 weremoderately anemic, and their serum Epo levels were slightlyelevated but no control group was provided for comparison(160). In a group of 20 moderately anemic or nonanemic chil-dren with various solid tumors, serum Epo did not correlatewith the degree of anemia but no control group was provided(161). Among 20 women with uterine or ovarian cancer,seven were anemic and their serum Epo relationship withthe hematocrit appeared somewhat blunted (162). In35 untreated patients with lung cancer, anemia was mainlydue to impaired erythroid marrow response to erythropoietinstimulation, and a defect in Epo production was operative inonly few of them (163). In a large cohort of 232 cancerpatients, pretreatment O=P Epo ratios were decreased,apparently indicating defective endogenous Epo production(164). However, these O=E Epo ratios are not valid becausethe range of Hb values in the group of patients with IDAwho served to derive the expected relationship betweenEpo and Hb was quite different from the one observed incancer patients, many of them having quite normal Hbvalues. In a large study of 56 children with miscellaneous solidtumors examined before any treatment, careful comparisonwith an appropriate pediatric control group showed thatserum Epo levels were adequate for the degree of anemia evenif erythropoiesis (as assessed by sTfR levels) was significantlyreduced, although to a lesser extent than in leukemic subjects(165). Among 92 patients with cirrhosis and hepatocellular


carcinoma, 55 had anemia and 37 a normal Hb value (166).Virtually, all anemic subjects had serum Epo values in therange expected from the 95% confidence limits of iron defi-ciency controls, whereas only two of nonanemic subjects hadinappropriately high serum Epo and polycythemia. Anotherinvestigation of 30 patients with hepatocellular carcinomafound no evidence of Epo deficiency (167).

Chronic Myeloid Disorders

The regulation of Epo production in patients with myelodys-plastic syndromes (MDS) appears to be extremely variable.In a study of 14 patients, serum Epo levels were markedly ele-vated, and the slope of the correlation between Epo and hema-tocrit was similar to that reported for simple IDA (168). Inanother group of 46 patients, the slope of the regression wascloser to that of controls with IDA than to that of controls withaplastic anemia (68). In a larger study of 75 MDS patients,there was also an overall inverse relationship between Epolevels and the degree of anemia (169). However, a wide rangeof Epo responses was encountered among patients with simi-lar hemoglobin concentrations, and there were many patientswith inappropriately low Epo levels as well as many otherswith inappropriately high values. A similar observation wasmade in another group of 46 patients with MDS who alsohad an overall inverse relationship between Epo and hemo-globin levels (170). A wide range of Epo responses betweenpatients with similar hemoglobin concentrations wasobserved, with the highest values measured in those with lessthan 10% erythroblasts in the bone marrow. However,another investigation of 20 patients with MDS by the samegroup found no correlation between serum Epo concentrationand total erythroid production, thereby negating any effect ofthe level of the erythropoietic activity on serum Epo concen-tration (171). The erythroid abnormality of patients withMDS was further analyzed in 19 nontransfusion-dependentpatients (172). Serum Epo concentration was appropriate forthe degree of anemia in 15=19 patients and was positivelyrelated to the percentage of highly fluorescent reticulocytes

164 Beguin

but not to the absolute reticulocyte count. Contrary to normalcontrols who exhibit a maximum concentration in the after-noon, a circadian rhythm of serum Epo concentration is notobserved in MDS patients (173). Interestingly, higher serumEpo levels were associated with poorer survival but hemoglo-bin values were not provided, so it cannot be excluded thatthis is simply an effect of more severe anemia (174). Andro-gen therapy in MDS (and a few aplastic anemia) patientshas been associated with a significant increase in serumEpo compared to untreated patients and even more so to irondeficiencycontrols, although the slope of the Hct vs. Eporegression was not different (175). Finally, patients with par-oxysmal nocturnal hemoglobinuria have serum Epo levelsthat, for any given degree of anemia, are elevated comparedto IDA patients but similar to those with aplastic anemia(176,177).

Among 61 anemic patients with myelofibrosis with mye-loid metaplasia, inappropriately low levels of serum Epo wereonly found in eight patients (178). An inverse correlation wasobserved between serum Epo concentration and hemoglobinas well as between the O=P Epo ratio and ferrokinetic mea-surements of erythropoiesis. In four separate reports of 174(179), 65 (180), 49 (181), and 40 (182) subjects with essentialthrombocythemia, serum Epo concentrations were signifi-cantly below normal levels in many patients. However, thesepatients were generally not anemic, and their pattern wassimilar to that of 343 patients with polycythemia vera (179).

Leukemia and Lymphoma

Compared to patients with iron-deficiency anemia, serum Epotiters displayed similar inverse relationships with hemoglo-bin concentration in separate analyses of 47 patients withacute leukemia, 54 with non-Hodgkin’s lymphoma, 34 withmultiple myeloma, 16 with myelofibrosis, but curiously notin 19 with chronic myelogenous leukemia (68). However, theslope of the regression was blunted in lymphoma and mye-loma patients, and several patients with multiple myelomaclearly had inappropriately low serum Epo levels. The O=P


Epo ratio was similar in patients with leukemia compared tohealthy controls or patients with iron-deficiency anemia, indi-cating that serum Epo production was appropriate for thedegree of anemia.

Other reports have focused on leukemias. Twelvepatients with hairy cell leukemia were found to have a normalfeedback mechanism for Epo production in response to ane-mia, but no formal control group was presented (183). Therole of Epo in chronic lymphocytic leukemia (CLL) has beenreviewed (184). Among 47 patients with CLL, Epo productionwas found to be adequate for the degree of anemia, and thisconclusion was not altered in advanced stages of the disease(185). Inappropriate Epo levels were only found in threepatients, two of whom had active infections. When patientswith acute leukemia were compared with patients withulcerative colitis, serum Epo levels were found to be higherfor similar degrees of anemia and somewhat less well corre-lated with hemoglobin (186–188). Although ulcerative colitisrepresents a form of chronic disorder and therefore does notappear to be an ideal control group, this result at least indi-cated that there was no evident Epo deficiency in patientswith acute leukemia. There are some studies of children withacute leukemia, in which it was also found that serum Epowas considerably increased and inversely related to hemoglo-bin concentration (189,190). In a large study of 55 childrenwith acute leukemia examined at diagnosis, careful compari-son with an appropriate pediatric control group revealed thaterythropoiesis (as assessed by sTfR levels) was severelydepressed, but serum Epo levels were appropriate for thedegree of anemia in virtually all of them (165).

Finally, several papers analyzed Epo levels in patientswith lymphoid malignancies. Erythropoietin production inresponse to anemia was considered normal in 12 childrenwith lymphoma, but no formal control group was presented(161). Others examined the Epo–Hb relationship in 63untreated patients with Hodgkin’s disease and found no evi-dence for depressed serum Epo levels, as the minority ofpatients who had anemia responded with adequate Epoproduction (191). Erythropoietin production has been more

166 Beguin

precisely evaluated in multiple myeloma (192). A negativecorrelation between erythropoiesis and the degree of renalimpairment has been observed (193–195). Using biologicalor radioimmunological assays, serum Epo levels were foundto be appropriate for the degree of anemia when renal func-tion was normal but inadequate when renal function wasimpaired (195–198). However, in another study, it was shownthat serum Epo levels were inadequate not only in patientswith renal impairment but also in a number of patients withnormal renal function (63). Approximately 25% of all patientshad defective Epo production and this increased to 30% ofanemic patients, 50% of stage 3 patients, and 60% of thosewith renal impairment. Plasma viscosity may contribute tothis phenomenon by blunting anemia-induced Epo productionby the kidney (199).

Conclusions

In conclusion, few studies have been conducted in a way thatdefinitive conclusions can be obtained, i.e., studies in untreatedanemic cancer patients with a suitable control group to provideeither comparison of regressions of serum Epo vs. Hct or Hb inthe patient and control groups or O=P Epo ratios in individualpatients. Most studies indicate that patients with leukemia orchronic myeloid disorders have appropriate Epo responses toanemia. A significant proportion of patients with multiplemyeloma and possibly lymphoma have impaired Epo respon-siveness. There is little evidence for defective endogenousEpo production in patients with solid tumors. However, thereis no report specifically addressing metastatic vs. localizeddisease. Therefore, the overall incidence of Epo deficiency insolid tumor patients remains poorly defined.

SERUM ERYTHROPOIETIN ANDCHEMOTHERAPY

Experimental Data

Experiments were conducted in various animal species toexplore the effect of chemotherapy and total body irradiation


on the capacity to increase Epo production in response tohypoxia. In rats exposed to hypoxia, neither cyclophospha-mide nor sublethal irradiation modified Epo production sig-nificantly in the following days (200). Lethal irradiation ledto anemia-driven Epo peaks that were not encountered inmice rescued by bone marrow transplantation (201). Adminis-tration of nitrogen mustard to sheep suffering from phenylhy-drazine-induced hemolytic anemia produced considerablyhigher titers of serum Epo (202). Administration of vanadiumto mice-bearing lymphoma was followed by prolongedenhanced Epo activity (203). Serum Epo levels during contin-uous exposure to hypoxia in mice with marrow aplasiainduced by whole body irradiation or 5-fluorouracil injectionwere higher than in control mice similarly exposed (104).

These in vivo data apparently suggest an enhancingeffect of chemotherapy on Epo production. As there are nopreformed stores of Epo, this cannot be due to a suddenrelease of Epo by the kidney, mediated by cytostatic drugs.Some other speculations have been offered as explanationfor this phenomenon (73). Cytotoxic therapy could cause adirect injury to the Epo-producing cells mimicking hypoxia.The blood flow to the kidney and=or liver could be altered soas to expose Epo-producing cells to hypoxia. As protein synth-esis and gene transcription are necessary for the normaldegradation of Epo mRNA, it is also possible that cytotoxictherapy could enhance Epo mRNA stability. However, someexperimental data contradict these assumptions. The kidneysof dogs were isolated in situ and perfused with bloodcontaining or not containing chlorambucil or thiotepa (204).Cobalt-induced Epo production was markedly suppressed18–36hr after the infusion of alkylating agents. In vitro stu-dies were conducted to examine the effect of chemotherapeu-tic agents on Epo synthesis in cultures of the hepatoma cellline, HepG2. The RNA synthesis-inhibiting drugs daunorubi-cin, cyclophosphamide, ifosfamide, and CDDP, as well as thetubulin-binding agent, vincristine, dose dependentlydecreased production of erythropoietin. The DNA synthesis-inhibiting drugs methotrexate and cytosine-arabinoside didnot have inhibitory properties (205,206). Together, these

168 Beguin

results indicate that chemotherapeutic agents may inhibitEpo production locally but that this effect is offset by othermechanisms, possibly nonutilization by a myelosuppressedbone marrow, leading to increased serum Epo levels.

Cisplatin (CDDP) is associated with a number of seriousside effects, including nephrotoxicity and myelosuppression,in particular anemia of long duration (207). As cisplatin isassociated with frequent and occasionally severe renal impair-ment, it has been speculated that Epo deficiency could be amajor factor in the development of CDDP-induced anemia.Experimental data support this concept. RNA synthesis-inhibiting drugs, including CDDP, produced a dose-dependentdecrease of Epo production by the human hepatoma cell line,HepG2, which partly correlated with cytotoxicity (205). Inanother study, CDDP also had a strong inhibitory effect onhypoxia- or cobalt-induced Epo mRNA expression and proteinproduction in the Hep3B cell line, with no apparent celldamage (208). Five days after injection of CDDP to mice orrats, hypoxia-induced Epo production was not adverselyaffected in spite of severe tubular necrosis (209). However,another study reported a significant drop of serum Epo concen-tration and kidney Epo mRNA content in rats 4–14days afterreceiving a bolus injection of cisplatin (210). Rats injected witha single high dose of CDDP developed acute renal failure andanemia that could be prevented or corrected by daily injectionsof 100U=kg rHuEpo (211,212). In addition, there was a signifi-cantly greater recovery of renal function with increased tubularregeneration.

The most informative study was conducted by Matsu-moto who compared the effect of 5-FU and CDDP on erythro-poiesis in rats and the role of rHuEpo in this setting (213).5-FU-induced anemia developed rapidly with a nadir at day10, whereas the anemia caused by CDDP was less prominentand developed later with a nadir at day 21. In 5-FU-inducedanemia, marked serum Epo elevation was observed at days7–14. Although serum Epo levels correlated negatively withthe hemoglobin, they fell rapidly afterwards, indicating thatthe early rise could be an effect of chemotherapy itself ratherthan anemia. This was followed by an increase of spleen but


not marrow CFU-E and a rise in reticulocytes, followed byrapid correction of the anemia. In contrast, CDDP-inducedanemia was not associated with changes in serum Epo orCFU-E values. As no animal decreased its hemoglobin below13 g=dL, it is not surprising that serum Epo levels were notelevated around day 20. On the other hand, CDDP did notproduce the early release of Epo into the circulation asobserved with 5-FU. These results with CDDP were confir-med in another study (214). After injection of 5-FU, treatmentwith rHuEpo did not prevent the fall of hemoglobin but some-what accelerated recovery in a dose-dependent fashion (213).Anemia could be completely prevented if rHuEpo was startedone week before administration of 5-FU. After CDDP treat-ment, rHuEpo was very effective in correcting the anemiain a dose-dependent manner, even when started only 2weeksafter CDDP had been given (213,214).

Nonplatinum Chemotherapy in Patients

Several studies have been conducted in cancer patients. In sixpatients receiving intensive chemotherapy for acute leuke-mia, serum Epo levels increased substantially after treatmentand gradually returned to baseline, often at the time of mar-row recovery (71). Intensive chemotherapy given for induc-tion of acute leukemia resulted in marked elevation ofserum Epo concentration starting one or two days later andpeaking after about 7 days, before normalizing later on (72).High serum Epo levels are also observed transiently afterintensive conditioning before bone marrow transplantationwithout concomitant change in hemoglobin or hematocrit(70–76). Another small study observed a large increment ofserum Epo soon after the initiation of chemotherapy for leu-kemia, which reached values of aplastic anemia patients atsimilar Hb levels (215). The same group reported the repeatedpostchemotherapy elevation of serum Epo levels in leukemicpatients, pinpointing a nice reciprocal relationship withserum iron (216), and obtained similar findings in patientswith lung cancer (217). After treatment with high-dose meth-otrexate, serum Epo increased in some children despite

170 Beguin

unchanged or even increased hemoglobin values, whereasafter treatment with high-dose arabinoside cytosine, serumEpo increased markedly in all in response to decreasinghemoglobin (189). Similar observations were made in adultsadministered a 5-day course of 5-fluorouracil and leucovorinafter the removal of colon cancer (218). Serum Epo levels pro-gressively increased for 15 days in the presence of constanthemoglobin levels. Urinary Epo excretion also increases afterchemotherapy (219). Patients with essential thrombocythe-mia on hydroxyurea, a-interferon or radioactive phosphorustherapy had increased Epo levels compared to untreatedpatients (180). All-trans retinoid acid treatment in patientswith acute promyelocytic leukemia was also associated witha transient increase in serum Epo values that correlatedinversely with reticulocyte counts, similar to the relationshipbetween TPO and platelets (220). A single dose of cyclopho-sphamide also increases serum Epo levels in patients withvasculitis-associated hypertension, implying that the effectof chemotherapeutic agents is not limited to cancer patients(221). Whole body hyperthermia does not affect the serumEpo response to chemotherapy (222). Taken together, theseresults powerfully demonstrate a transient surge in serumEpo values during 1–2weeks after chemotherapy.

Serum Epo levels after six cycles of non-nephrotoxicchemotherapy for stage 2 breast cancer increased slightly inrelation to a small decrease in Hct and correlated negativelywith insulin-like growth factor-1 (223,224). A large studyreported the evolution of serum Epo and O=P ratios in 232patients with miscellaneous tumors receiving a variable num-ber of chemotherapy cycles, including cisplatin in 65% of thecases (164). While serum Epo increased progressively as ananemia developed in the majority of them, the O=P Epo ratiodecreased until the fourth cycle and recovered at cycle 6.However, the relationship between serum Epo and Hb wasfully maintained in 55 children with acute leukemia followedup at the end of induction and during the course of mainte-nance chemotherapy, whereas erythropoietic activity (sTfRlevels) was further reduced compared to pretreatment levels(165). Identical conclusions were derived from the follow-up


of 56 children with solid tumors (165). Pediatric patientsinvestigated at various stages of induction, consolidation,and maintenance chemotherapy for acute leukemia main-tained a significant inverse correlation between serum Epoand Hb that was particularly close in those with Hb less than10 g=dL (225). These data suggest that nonplatinum che-motherapy in general does not induce Epo deficiency in themid- or long-term.

Cisplatin Chemotherapy in Patients

In a study of 24patients with gynecologic malignancies, therewas a significant decrease of serum Epo levels between 2 and6hr after chemotherapy with cisplatin and cyclophosphamide,followed by a return to baseline values after 12hr (226).Combination chemotherapy regimens based on cisplatin(100mg=m2) or carboplatin (300mg=m2) were associated withthe usual peak of serum Epo levels observed 1–2weeks afterchemotherapy (227,228). Plasma Epo concentration increasedsimilarly in advanced cancer patients 15 days after chemother-apy did or did not contain cisplatin (229). In seven patientswith ovarian carcinoma undergoing cisplatin chemotherapy,serum Epo was increased 24hr and 7days later independentof concomitant anemia (230). In another small study, serumEpo in solid tumor patients receiving cisplatinwas higher thanin similarly anemic patients treated without cisplatin (231).Therefore, apart from a possible very early inhibition of Eposecretion, cisplatin is no exception to the development of aserum Epo peak 1–2weeks after chemotherapy.

In patients with gynecologic cancer receiving multiplecourses of combination chemotherapy including 50mg=m2 cis-platin, prechemotherapy serum Epo values were progres-sively elevated in relation with the degree of anemiaachieved, although a comparison with only eight anemic con-trols is of little value (232). Few among head and neck as wellas other cancer patients receiving cisplatin (100mg=m2)-based chemotherapy developed inappropriately low Epolevels, and there was no correlation with the amount of cispla-tin administered or the degree of renal impairment (233). A

172 Beguin

linear relationship between log(Epo) and Hb was retainedafter treatment of 12 children with various solid tumors,but no comparison with pretreatment values or normal con-trols was provided (234). A longitudinal study of patients withovarian or bladder cancer treated with nine courses of CDDP(60mg=m2) and doxorubicin (60mg=m2) showed progressiveanemia correlating with renal tubular dysfunction (214).O=P Epo ratios declined progressively in proportion to thedegree of renal dysfunction and recovered after cessation ofCDDP therapy along with restoration of tubular functiondespite persistently depressed creatinine clearance. Overall,there is some evidence for Epo deficiency after completion ofplatinum-based chemotherapy, although this is certainly nota universal finding.

SERUM ERYTHROPOIETIN AS PREDICTOR OFRESPONSE TO rHuEpo

Based on our knowledge of the pathophysiology of theACDandcancer, it is clear that the most useful approach is to treat theunderlying disorder (2,14,15). However, red cell transfusionsare regularly needed in patients with ACD. In this context,rHuEpo may be of particular value in stimulating endogenouserythropoiesis, and has now been widely tested in the treat-ment of ACD patients with a variety of diseases (235), includ-ing HIV infection (236), rheumatoid arthritis (237), andcancer with or without concomitant chemotherapy (155,156).

Theoretically, patients with a defect in the capacity toproduce Epo would be more likely to respond to rHuEpo thanthose with adequate serum Epo levels for their degree of ane-mia. As Epo levels must be interpreted in relation to thedegree of anemia, the ratio of observed-to-predicted Epo levels(O=P ratio) represents a better assessment of the adequacy ofEpo production (62). In patients with hematologic malignan-cies, it has been observed that low baseline serum Epo levels(238) or decreased O=P ratios (158) were associated with asignificantly higher probability of response. This has beenconfirmed in large multicenter trials in patients with multiple


myeloma or non-Hodgkin’s lymphoma (155,156). An O=Pratio < 0.9 was found to be associated with high responserates, whereas patients with an O=P ratio >0.9 rarelybenefited from therapy (239). Studies in patients with solidtumors have failed to confirm such a consistent predictivevalue of baseline Epo even when Epo deficiency was demon-strated in part of the patients (151,153,154,157,240). How-ever, a study aiming at preventing anemia in patients withovarian carcinoma undergoing platinum-based chemotherapyshowed a trend for lower transfusion needs in those with anO=P ratio < 0.8 (241). In addition, a small study in patientswith a variety of solid tumors suggested that the ratio ofbaseline Epo=corrected reticulocyte count could provide somepredictive information (242).

A combination of baseline parameters and early changesobserved after 2weeks of rHuEpo may provide another usefulapproach. Among evaluable patients treated in a large multi-center study (156), the failure rate was almost 90% whenbaseline serum O=P Epo was higher than 0.9 or when serumO=P Epo was less than 0.9 but the hemoglobin increment byweek 2 was <0.3 g=dL. On the other hand, the success ratewas around 90% when baseline serum O=P Epo was less than0.9 and hemoglobin increased by�0.3 g=dL. Similar findingswere obtained in a smaller study in children with solidtumors: an O=P ratio < 1.0 at baseline and a hemoglobinincrement >0.5 g=dL after 2weeks were associated withhigher response rates (243). In another large single centerstudy (239), the combined use of baseline serum Epo and the2-week increment of sTfR proved to be very powerful. Only18% of patients with a baseline serum Epo greater than100mU=mL responded to treatment, and only 29% respondedwhen the baseline serum Epo was < 100mU=mL but the2-week sTfR increment was less than 25%. On the other hand,the response rate was 96% among patients with a lowbaseline serum Epo and a substantial sTfR elevation.

In conclusion, baseline serum Epo should be measured atbaseline in patients with hematologic malignancies: treat-ment should not be initiated if endogenous serum Epo is above100mU=mL (or 200mU=ml in severely anemic patients) or the

174 Beguin

O=P ratio is >0.9. In addition, the best algorithms appear tobe those combining an assessment of the adequacy of endogen-ous Epo production (at least in hematologic malignancies)together with some early indicators of erythropoietic marrowresponse (changes in hemoglobin or sTfR). Of importance, inpatients treated with chemotherapy, serum Epo should beevaluated just prior to chemotherapy for its interpretation tobe valid. Indeed, without any change in hematocrit, serumEpo may be inappropriately elevated in the 2 weeks afterchemotherapy compared to prechemotherapy values, mostprobably because myelosuppression then decreases Epo utili-zation by target cells (see above). Therefore, it cannot beexcluded that the failure to predict response in solid tumorpatients may just be related to an inadequate timing of serumEpo sampling. While evaluation of endogenous Epo produc-tion may be relevant in various forms of anemia, it is of nointerest in subjects in whom the aim of rHuEpo therapy isto prevent an anemia that is not yet present, in those in whombetter tumor oxygenation before radiotherapy or induction offetal hemoglobin is sought, or in disorders characterized byuniversal Epo deficiency.

ACKNOWLEDGMENTS

This work was supported in part by grants from the NationalFund for Scientific Research (Fonds National de la RechercheScientifique, FNRS), Belgium.

REFERENCES

1. Cartwright GE. The anemia of chronic disorders. SeminHematol 1966; 3:351–375.

2. Beguin Y. Erythropoietin and the anemia of cancer. Acta ClinBelg 1996; 51:36–52.

3. Means RT Jr, Krantz SB. Progress in understanding thepathogenesis of the anemia of chronic disease. Blood 1992;80:1639–1647.


4. Means RT Jr. Advances in the anemia of chronic disease. IntJ Hematol 1999; 70:7–12.



7. Sears D. Anemia of chronic disease. Med Clin North Am 1992;76:567–579.

8. Torti FM, Torti SV. Cytokines, iron homeostasis, and cancer.Adv Exp Med Biol 1994; 354:161–170.

9. Konijn AM. Iron metabolism in inflammation. Clin Haematol1994; 7:829–849.

10. Weiss G, Wachter H, Fuchs D. Linkage of cell-mediatedimmunity to iron metabolism. Immunol Today 1995; 16:495–500.

11. Weiss G. Iron and anemia of chronic disease. Kidney IntSuppl 1999; 69:S12–S17.

12. Means RT Jr. Pathogenesis of the anemia of chronic disease:a cytokine-mediated anemia. Stem Cells 1995; 13:32–37.

13. Trey JE, Kushner I. The acute phase response and thehematopoietic system: the role of cytokines. Crit Rev OncolHematol 1995; 21:1–18.

14. Spivak JL. Cancer-related anemia: its causes and character-istics. Semin Oncol 1994; 21(suppl 3):3–8.

15. Moliterno AR, Spivak JL. Anemia of cancer. Hematol OncolClin North Am 1996; 10:345–363.

16. Doll DC, Weiss RB. Neoplasia and the erythron. J Clin Oncol1985; 3:429–446.

17. Tchekmedyian NS. Anemia in cancer patients: significance,epidemiology, and current therapy. Oncology (Huntingt)2002; 16:17–24.

18. Groopman JE, Itri LM. Chemotherapy-induced anemia inadults: incidence and treatment. J Natl Cancer Inst 1999;91:1616–1634.

176 Beguin

19. Michon J. Incidence of anemia in pediatric cancer patients inEurope: results of a large, international survey. Med PediatrOncol 2002; 39:448–450.

20. Heddens D, Alberts DS, Hannigan EV, Williams SD, GarciaD, Roe DJ, Bell J, Alvarez RD. Prediction of the need for redcell transfusion in newly diagnosed ovarian cancer patientsundergoing platinum-based treatment. Gynecol Oncol 2002;86:239–243.

21. Skillings JR, Sridhar FG, Wong C, Paddock L. The frequencyof red cell transfusion for anemia in patients receivingchemotherapy. Am J Clin Oncol 1993; 16:22–25.

22. Benoist S, Panis Y, Pannegeon V, Alves A, Valleur P. Predic-tive factors for perioperative blood transfusions in rectalresection for cancer: a multivariate analysis of a group of212 patients. Surgery 2001; 129:433–439.

23. Hensley ML, Lebeau D, Leon LF, Venkatraman E, WaltzmanR, Sabbatini P, Almadrones L, Chi D, Spriggs D. Identifica-tion of risk factors for requiring transfusion during front-linechemotherapy for ovarian cancer. Gynecol Oncol 2001; 81:485–489.

24. Cassidy MJD, De Jager C, Ebrahim O, Camachio P,Robson R. Peripheral blood mononuclear cells from patientswith chronic renal failure release factors which suppress ery-thropoietin secretion in vitro. Nephrol Dial Transplant 1994;9:775–779.

25. Mara M, Zivny J, Eretova V, Kvasnicka J, Kuzel D, Umlau-fova A, Marova E. Changes in markers of anemia and ironmetabolism and how they are influenced by antianemics inpostpartum period. Acta Obstet Gynecol Scand 2001; 80:142–148.

26. Faquin WC, Schneider TJ, Goldberg MA. Effect of inflamma-tory cytokines on hypoxia-induced erythropoietin production.Blood 1992; 79:1987–1994.

27. Vannucchi AM, Grossi A, Rafanelli D, Statello M, Cinotti S,Rossi-Ferrini P. Inhibition of erythropoietin production invitro by human interferon gamma. Br J Haematol 1994;87:18–23.


28. Fandrey J, Jelkmann WE. Interleukin-1 and tumor necrosisfactor-alpha inhibit erythropoietin production in vitro. AnnN Y Acad Sci 1991; 628:250–255.

29. Jelkmann W, Pagel H, Wolff M, Fandrey J. Monokinesinhibiting erythropoietin production in human hepatomacultures and in isolated perfused rat kidneys. Life Sci1992; 50:301–308.

30. Jelkmann W, Wolff M, Fandrey J. Modulation of the produc-tion of erythropoietin by cytokines: in vitro studies and theirclinical implications. Contrib Nephrol 1990; 87:68–77.

31. Fandrey J, Huwiler A, Frede S, Pfeilschifter J, Jelkmann W.Distinct signaling pathways mediate phorbol-ester-inducedand cytokine-induced inhibition of erythropoietin geneexpression. Eur J Biochem 1994; 226:335–340.

32. Jelkmann W, Hellwig-Buergel T. Tumor necrosis factor p55receptor (TNF-RI) mediates the in vitro inhibition of hepaticerythropoietin production. Exp Hematol 1999; 27:224–228.

33. Hellwig-Burgel T, Rutkowski K, Metzen E, Fandrey J,Jelkmann W. Interleukin-1 beta and tumor necrosis factor-alpha stimulate DNA binding of hypoxia-inducible factor-1.Blood 1999; 94:1561–1567.

34. Schobersberger W, Hoffmann G, Fandrey J. Nitric oxidedonors suppress erythropoietin production in vitro. PflugersArch 1996; 432:980–985.

35. Fandrey J, Frede S, JelkmannW. Role of hydrogen peroxide inhypoxia-induced erythropoietin production. Biochem J 1994;303:507–510.

36. Fandrey J, Frede S, Ehleben W, Porwol T, Acker H,JelkmannW. Cobalt chloride and desferrioxamine antagonizethe inhibition of erythropoietin production by reactive oxygenspecies. Kidney Int 1997; 51:492–496.

37. Jelkmann W, Pagel H, Hellwig T, Fandrey J. Effects of anti-oxidant vitamins on renal and hepatic erythropoietin produc-tion. Kidney Int 1997; 51:497–501.

38. Leng HM, Kidson SH, Keraan MM, Randall GW, Folb PI.Cytokine-mediated inhibition of erythropoietin synthesis bydexamethasone. J Pharm Pharmacol 1996; 48:971–974.

178 Beguin

39. Frede S, Fandrey J, Pagel H, Hellwig T, Jelkmann W.Erythropoietin gene expression is suppressed after lipopoly-saccharide or interleukin-1 beta injections in rats. Am J Phy-siol 1997; 273:R1067–R1071.

40. Chuncharunee S, Carter CD, Studtmann KE, Caro J,Coffey RJ, Dessypris EN. Chronic administration of trans-forming growth factor-beta suppresses erythropoietin-dependent erythropoiesis and induces tumour necrosis factorin vivo. Br J Haematol 1993; 84:374–380.

41. Miller KL, Carlino JA, Ogawa Y, Avis PD, Carroll KG. Altera-tions in erythropoiesis in TGF-beta 1-treated mice. ExpHematol 1992; 20:951–956.

42. Nieken J, Mulder NH, Buter J, Vellenga E, Limburg PC,Piers DA, de Vries EG. Recombinant human interleukin-6induces a rapid and reversible anemia in cancer patients.Blood 1995; 86:900–905.

43. Oczko-Grzesik B, Wiecek A, Kokot F. Influence of IFN-alphaon plasma erythropoietin levels in patients with hepatitis Bvirus-associated chronic active hepatitis. J Interferon Cyto-kine Res 2001; 21:669–676.

44. Johnson CS, Keckler DJ, Topper MI, Braunschweiger PG,Furmanski P. In vivo hematopoietic effects of recombinantinterleukin-1 alpha in mice: stimulation of granulocytic,monocytic, megakaryocytic, and early erythroid progenitors,suppression of late-stage erythropoiesis, and reversal oferythroid suppression with erythropoietin. Blood 1989; 73:678–683.

45. Furmanski P, Johnson CS. Macrophage control of normal andleukemic erythropoiesis: identification of the macrophage-derived erythroid suppressing activity as interleukin-1 andthe mediator of its in vivo action as tumor necrosis factor.Blood 1990; 75:2328–2334.

46. de Haan G, Dontje B, Loeffler M, Nijhof W. Microenvironmen-tally dependent effects on murine haematopoiesis by aprolonged interleukin-1 treatment. Br J Haematol 1993; 85:15–19.

47. Johnson CS, Cook CA, Furmanski P. In vivo suppression oferythropoiesis by tumor necrosis factor-alpha (TNF-alpha):


reversal with exogenous erythropoietin (EPO). Exp Hematol1990; 18:109–113.

48. Clibon U, Bonewald L, Caro J, Roodman GD. Erythropoietinfails to reverse the anemia in mice continuously exposed totumor necrosis factor-alpha in vivo. Exp Hematol 1990;18:438–441.

49. Roodman GD, Johnson RA, Clibon U. Tumor necrosis factoralpha and the anemia of chronic disease: effects of chronicexposure to TNF on erythropoiesis in vivo. Adv Exp Med Biol1989; 271:185–196.

50. Jelkmann W. Erythropoietin: structure, control of produc-tion, and function. Physiol Rev 1992; 72:449–489.

51. Barosi G. Inadequate erythropoietin response to anemia: defi-nition and clinical relevance. Ann Hematol 1994; 68:215–223.

52. Bozzini CE, Alippi RM, Barcelo AC. Enhanced effect ofincreased erythrocyte production rate on plasma erythropoie-tin levels of mice during subsequent exposure to hypobaria.Adv Exp Med Biol 1989; 271:23–27.

53. Alippi RM, Boyer P, Leal T, Barcelo AC, Martinez MP, Boz-zini CE. Higher erythropoietin secretion in response to cobal-tous chloride in post-hypoxic than in hypertransfusedpolycythemic mice. Haematologica 1992; 77:446–449.

54. Alippi RM, Barcelo AC, Bozzini CE. Enhanced erythropoiesisinduced by hypoxia in hypertransfused, post-hypoxic mice.Exp Hematol 1983; 11:878–883.

55. Alippi RM, Barcelo AC, Bozzini CE. Erythropoietic responseto hypoxia in mice with polycythemia induced by hypoxia ortransfusion. Exp Hematol 1983; 11:122–128.

56. Alippi RM, Barcelo AC, Bozzini CE. Comparison of erythro-poietic response to erythropoietin-secreting stimuli in micefollowing polycythemia induced by transfusion or hypoxia.Exp Hematol 1985; 13:159–162.

57. Alippi RM, Barcelo AC, Kofoed JA, Bozzini CE. Hypoxia-induced renal and extrarenal erythropoietin production inposthypoxic or hypertransfused polycythemic rats. ExpHematol 1986; 14:329–332.

180 Beguin

58. Erslev AJ. Erythropoietin. N Engl J Med 1991; 324:1339–1344.

59. Garcia JF, Ebbe SN, Hollander L, Cutting HO, Miller MO,Cronkite EP. Radioimmunoassay of erythropoietin: circulat-ing levels in normal and polycythemic human beings. J LabClin Med 1982; 99:624–635.

60. Caro J, Brown S, Miller O, Murray T, Erslev AJ. Erythropoie-tin levels in uremic nephric and anephric patients. J Lab ClinMed 1979; 93:449–458.

61. Cotes PM, Dore CJ, Liu Yin JA, Lewis SM, Messinezy M,Pearson TC, Reid C. Determination of serum immunoreactiveerythropoietin in the investigation of erythrocytosis. N Engl JMed 1986; 315:283–287.

62. Beguin Y, Clemons G, Pootrakul P, Fillet G. Quantitativeassessment of erythropoiesis and functional classification ofanemia based on measurements of serum transferrin receptorand erythropoietin. Blood 1993; 81:1067–1076.

63. Beguin Y, Yerna M, Loo M, Weber M, Fillet G. Erythropoiesisin multiple myeloma: defective red cell production due toinappropriate erythropoietin production. Br J Haematol1992; 82:648–653.

64. Alexanian R. Erythropoietin excretion in bone marrow failureand hemolytic anemia. J Lab Clin Med 1973; 82:438–445.

65. Bray GL, Taylor B, O’Donnell R. Comparison of the erythro-poietin response in children with aplastic anemia, transienterythroblastopenia, and iron deficiency. J Pediatr 1992;120:528–532.

66. Schrezenmeier H, Noe G, Raghavachar A, Rich IN,Heimpel H, Kubanek B. Serum erythropoietin and serumtransferrin receptor levels in aplastic anaemia. Br J Haema-tol 1994; 88:286–294.

67. de Klerk G, Rosengarten PCJ, Vet RJWM, Goudsmit R.Serum erythropoietin (ESF) titers in anemia. Blood 1981;58:1164–1170.

68. Urabe A, Mitani K, Yoshinaga K, Iki S, Yagisawa M, Ohbaya-shi Y, Takaku F. Serum erythropoietin titers in hematologi-


cal malignancies and related diseases. Int J Cell Cloning1992; 10:333–337.

69. Gaines Das RE, Milne A, Rowley M, Smith EC, Cotes PM.Serum immunoreactive erythropoietin in patients with idio-pathic aplastic and Fanconi’s anaemias. Br J Haematol1992; 82:601–607.

70. Cazzola M, Guarnone R, Cerani P, Centenara E, Rovati A,Beguin Y. Red blood cell precursor mass as an independentdeterminant of serum erythropoietin level. Blood 1998; 91:2139–2145.

71. PirosoE,ErslevAJ,CaroJ. Inappropriate increase inerythropoie-tin titers during chemotherapy. AmJHematol 1989; 32:248–254.

72. Birgegard G, Wide L, Simonsson B. Marked erythropoietinincrease before fall in Hb after treatment with cytostaticdrugs suggests mechanism other than anaemia for stimula-tion. Br J Haematol 1989; 72:462–466.

73. Schapira L, Antin JH, Ransil BJ, Antman KH, Eder JP,McGarigle CJ, Goldberg MA. Serum erythropoietin levels inpatients receiving intensive chemotherapy and radiotherapy.Blood 1990; 76:2354–2359.

74. Beguin Y, Clemons GK, Oris R, Fillet G. Circulating erythro-poietin levels after bone marrow transplantation: inappropri-ate response to anemia in allogeneic transplants. Blood 1991;77:868–873.

75. Beguin Y, Oris R, Fillet G. Dynamics of erythropoietic recov-ery after bone marrow transplantation: role of marrow prolif-erative capacity and erythropoietin production in autologousversus allogeneic transplants. Bone Marrow Transplant1993; 11:285–292.

76. Beguin Y, Baron F, Fillet G. Influence of marrow erythropoie-tic activity on serum erythropoietin levels after autologoushematopoietic stem cell transplantation. Haematologica1998; 83:1076–1081.

77. Kendall RG, Cavill I, Norfolk DR. Erythropoietin consump-tion during stimulated erythropoiesis. Ann N Y Acad Sci1994; 718:350–352.

182 Beguin

78. Finne PH, Skoglund R, Wetterhus S. Urinary erythropoietinexcretion during initial treatment of pernicious anaemia.Scand J Haematol 1973; 10:62–68.

79. Kendall RG, Cavill I, Norfolk DR. Serum erythropoietinlevels during haematinic therapy. Br J Haematol 1992; 81:630–631.

80. Youssoufian H, Longmore G, Neumann D, Yoshimura A, Lod-ish HF. Structure, function, and activation of the erythropoie-tin receptor. Blood 1993; 81:2223–2236.

81. Bowen DT, Janowska-Wieczorek A. Serum erythropoietin fol-lowing cytostatic therapy [letter]. Br J Haematol 1990; 74:372–373.

82. Stohlman F Jr, Brecher G. Humoral regulation of erythropoi-esis. Relationship of plasma erythropoietin level to bone mar-row activity. Proc Soc Exp Biol Med 1959; 100:40–43.

83. Naets JP, Wittek M. Effect of erythroid hypoplasia on utiliza-tion of erythropoietin. Nature 1965; 206:726–727.

84. Bozzini CE. Influence of erythroid activity of the bone mar-row on the plasma disappearance of injected erythropoietinin dogs. Nature 1966; 209:1140–1141.

85. Steinberg SE, Garcia JF, Matzke GR, Mladenovic J. Erythro-poietin kinetics in rats: generation and clearance. Blood 1986;67:646–649.

86. Widness JA, Veng-Pedersen P, Schmidt RL, Lowe LS,Kisthard JA, Peters C. In vivo 125I-erythropoietin pharmaco-kinetics are unchanged after anesthesia, nephrectomy andhepatectomy in sheep. J Pharmacol Exp Ther 1996; 279:1205–1210.

87. Spivak JL, Hogans BB. The in vivo metabolism of recombi-nant human erythropoietin in the rat. Blood 1989; 73:90–99.

88. Spivak JL. Metabolism of recombinant human erythropoietinin the rat. Adv Exp Med Biol 1989; 271:29–38.

89. Piroso E, Erslev AJ, Flaharty KK, Caro J. Erythropoietin lifespan in rats with hypoplastic and hyperplastic bone marrows.Am J Hematol 1991; 36:105–110.


90. Lezon CE, Martinez MP, Conti MI, Bozzini CE. Plasma disap-pearance of exogenous erythropoietin in mice under differentexperimental conditions. Endocrine 1998; 8:331–333.

91. Flaharty KK, Caro J, Erslev A, Whalen JJ, Morris EM,Bjornsson TD, Vlasses PH. Pharmacokinetics and erythro-poietic response to human recombinant erythropoietin inhealthy men. Clin Pharmacol Ther 1990; 47:557–564.

92. Cheung WK, Goon BL, Guilfoyle MC, Wacholtz MC. Pharma-cokinetics and pharmacodynamics of recombinant humanerythropoietin after single and multiple subcutaneous dosesto healthy subjects. Clin Pharmacol Ther 1998; 64:412–423.

93. Kato M, Kamiyama H, Okazaki A, Kumaki K, Kato Y,Sugiyama Y. Mechanism for the nonlinear pharmacokineticsof erythropoietin in rats. J Pharmacol Exp Ther 1997;283:520–527.

94. Chapel SH, Veng-Pedersen P, Schmidt RL, Widness JA.Receptor-based model accounts for phlebotomy-inducedchanges in erythropoietin pharmacokinetics. Exp Hematol2001; 29:425–431.

95. Kampf D, Eckardt KU, Fischer HC, Schmalisch C, Ehmer B,Schostak M. Pharmacokinetics of recombinant humanerythropoietin in dialysis patients after single and multiplesubcutaneous administrations. Nephron 1992; 61:393–398.

96. McMahon FG, Vargas R, Ryan M, Jain AK, Abels RI, Perry B,Smith IL. Pharmacokinetics and effects of recombinanthuman erythropoietin after intravenous and subcutaneousinjections in healthy volunteers. Blood 1990; 76:1718–1722.

97. Sans T, Joven J, Vilella E, Masdeu G, Farre M. Pharma-cokinetics of several subcutaneous doses of erythropoietin:potential implications for blood transfusion. Clin Exp Phar-macol Physiol 2000; 27:179–184.

98. Kinoshita H, Ohishi N, Kato M, Tokura S, Okazaki A. Distri-bution of recombinant human erythropoietin following multi-ple intravenous administration and effects of age on thedistribution in rats. Arzneimittelforschung 1992; 42:579–584.

99. Chapel S, Veng-Pedersen P, Hohl RJ, Schmidt RL,McGuire EM, Widness JA. Changes in erythropoietin

184 Beguin

pharmacokinetics following busulfan-induced bone marrowablation in sheep: evidence for bone marrow as a majorerythropoietin elimination pathway. J Pharmacol ExpTher 2001; 298:820–824.

100. Piron M, Loo M, Gothot A, Tassin F, Fillet G, Beguin Y.Cessation of intensive treatment with recombinant humanerythropoietin is followed by secondary anemia. Blood2001; 97:442–448.

101. Bozzini CE, Alippi RM, Barcelo AC, Conti MI, Bozzini C,Lezon CE, Olivera MI. The biology of stress erythropoiesisand erythropoietin production. Ann N Y Acad Sci 1994;718:83–92.

102. Lezon C, Alippi RM, Barcelo AC, Martinez MP, Conti MI,Bozzini CE. Depression of stimulated erythropoietin produc-tion in mice with enhanced erythropoiesis. Haematologica1995; 80:491–494.

103. Bozzini CE, Barcelo AC, Conti MI, Martinez MP, Lezon CE,Bozzini C, Alippi RM. Unexpected hypoxia-dependent ery-thropoietin secretion during experimental conditions notaffecting tissue oxygen supply=demand ratio. Kidney Int1997; 51:413–415.

104. Barcelo AC, Bozzini CE. Erythropoietin formation duringhypoxia in mice with impaired responsiveness to erythropoie-tin induced by irradiation or 5-fluorouracil injection. Experi-entia 1982; 38:504–505.

105. Bergamaschi G, Recalde HH, Ponchio L, Rosti V, Cazzola M.Erythrophagocytosis increases the expression of erythroidpotentiating activity mRNA in human monocyte-macrophages.Exp Hematol 1993; 21:70–73.

106. Erslev AJ. The effect of hemolysates on red cell productionand erythropoietin release. J Lab Clin Med 1971; 78:1–7.

107. Means RT Jr. Cytokines and anaemia in human immunode-ficiency virus infection. Cytokines Cell Mol Ther 1997; 3:179–186.

108. Moyle G. Anaemia in persons with HIV infection: prognosticmarker and contributor to morbidity. AIDS Rev 2002; 4:13–20.


109. Calenda V, Chermann JC. Severe in vitro inhibition oferythropoiesis and transient stimulation of granulopoiesisafter bone-marrow infection with eight different HIV-2isolates. AIDS 1992; 6:943–948.

110. Reddy MM, Grieco MH. Erythropoietin and granulocyte-macrophage colony-stimulating factor (GM-CSF) levels insera of patients with HIV infection. Int J STD AIDS 1991;2:128–132.

111. Allen UD, King SM, Gomez MP, Lapointe N, Forbes JC,Thorne A, Kirby MA, Bowker J, Raboud J, Singer J, Muk-waya G, Tobin J, Read SE. Serum immunoreactive erythro-poietin levels and associated factors amongst HIV-infectedchildren. AIDS 1998; 12:1785–1791.

112. Semba RD, Broadhead R, Taha TE, Totin D, Ricks MO,Kumwenda N. Erythropoietin response to anemia amonghuman immunodeficiency virus-infected infants in Malawi.Haematologica 2001; 86:1221–1222.

113. Rarick MU, Loureiro C, Groshen S, Sullivan Halley J,Gill PS, Bernstein Singer M, Levine AM. Serum erythropoie-tin titers in patients with human immunodeficiency virus(HIV) infection and anemia. J Acquir Immune Defic Syndr1991; 4:593–597.

114. Kreuzer KA, Rockstroh JK, Jelkmann W, Theisen A,Spengler U, Sauerbruch T. Inadequate erythropoietinresponse to anaemia in HIV patients: relationship to serumlevels of tumour necrosis factor-alpha, interleukin-6 andtheir soluble receptors. Br J Haematol 1997; 96:235–239.

115. Spivak JL, Barnes DC, Fuchs E, Quinn TC. Serum immunor-eactive erythropoietin in HIV-infected patients. JAMA 1989;261:3104–3107.

116. Camacho J, Poveda F, Zamorano AF, Valencia ME,Vazquez JJ, Arnalich F. Serum erythropoietin levels in anae-mic patients with advanced human immunodeficiency virusinfection. Br J Haematol 1992; 82:608–614.

117. Walker RE, Parker RI, Kovacs JA, Masur H, Lane HC, Car-leton S, Kirk LE, Gralnik HR, Fauci AS. Anemia and erythro-poiesis in patients with the acquired immunodeficiency

186 Beguin

syndrome (AIDS) and Kaposi sarcoma treated with zidovu-dine. Ann Intern Med 1988; 108:372–376.

118. Bertero MT, Caligaris-Cappio F. Anemia of chronic disordersin systemic autoimmune diseases. Haematologica 1997;82:375–381.

119. Vreugdenhil G, Swaak AJ. The role of erythropoietin in theanaemia of chronic disease in rheumatoid arthritis. ClinRheumatol 1990; 9:22–27.

120. Graudal N, Nielsen OJ, Galloe AM, Graudal HN. The courseof erythropoietin in patients with rheumatoid arthritis withnormal and low blood-hemoglobin. A longitudinal study.Scand J Rheumatol 1993; 22:86–89.

121. Chijiwa T, Nishiya K, Hashimoto K. Serum transferrinreceptor levels in patients with rheumatoid arthritis arecorrelated with indicators for anaemia. Clin Rheumatol2001; 20:307–313.

122. Vreugdenhil G, Wognum AW, van Eijk HG, Swaak AJ. Anae-mia in rheumatoid arthritis: the role of iron, vitamin B12,and folic acid deficiency, and erythropoietin responsiveness.Ann Rheum Dis 1990; 49:93–98.

123. Giordano N, Cecconami L, Marcucci P, Battisti E, Magaro L,Morozzi G, Mariano A, Marcolongo R. The role of iron,vitamin B12, folic acid and erythropoietin in the anemia ofrheumatoid arthritis. Clin Exp Rheumatol 1992; 10:201–202.

124. Takashina N, Kondo H, Kashiwazaki S. Suppressed serumerythropoietin response to anemia and the efficacy of recom-binant erythropoietin in the anemia of rheumatoid arthritis.J Rheumatol 1990; 17:885–887.

125. Remacha AF, Rodriguez de la Serna A, Garcia Die F, Geli C,Diaz C, Gimferrer E. Erythroid abnormalities in rheumatoidarthritis: the role of erythropoietin. J Rheumatol 1992; 19:1687–1691.

126. Nielsen OJ, Andersen LS, Ludwigsen E, Bouchelouche P,Hansen TM, Birgens H, Hansen NE. Anaemia of rheumatoidarthritis: serum erythropoietin concentrations and red cell


distribution width in relation to iron status. Ann Rheum Dis1990; 49:349–353.

127. Hochberg MC, Arnold CM, Hogans BB, Spivak JL. Serumimmunoreactive erythropoietin in rheumatoid arthritis:impaired response to anemia. Arthritis Rheum 1988; 31:1318–1321.

128. Noe G, Augustin J, Hausdorf S, Rich IN, Kubanek B. Serumerythropoietin and transferrin receptor levels in patientswith rheumatoid arthritis. Clin Exp Rheumatol 1995; 13:445–451.

129. Boyd HK, Lappin TR, Bell AL. Evidence for impaired ery-thropoietin response to anaemia in rheumatoid disease. BrJ Rheumatol 1991; 30:255–259.

130. Boyd HK, Lappin TR. Erythropoietin deficiency in the anae-mia of chronic disorders. Eur J Haematol 1991; 46:198–201.

131. Baer AN, Dessypris EN, Goldwasser E, Krantz SB. Bluntederythropoietin response to anaemia in rheumatoid arthritis.Br J Haematol 1987; 66:559–564.

132. Cazzola M, Ponchio L, de Benedetti F, Ravelli A, Rosti V,Beguin Y, Invernizzi R, Barosi G, Martini A. Defective ironsupply for erythropoiesis and adequate endogenous erythro-poietin production in the anemia associated with systemic-onset juvenile chronic arthritis. Blood 1996; 87:4824–4830.

133. Kirel B, Yetgin S, Saatci U, Ozen S, Bakkaloglu A, Besbas N.Anaemia in juvenile chronic arthritis. Clin Rheumatol 1996;15:236–241.

134. Fantini F, Gattinara M, Gerloni V, Bergomi P, Cirla E. Severeanemia associated with active systemic-onset juvenile rheuma-toid arthritis successfully treated with recombinant human ery-thropoietin: a pilot study. Arthritis Rheum 1992; 35:724–726.

135. Martini A, Ravelli A, Di Fuccia G, Rosti V, CazzolaM, Barosi G.Intravenous iron therapy for severe anaemia in systemic-onsetjuvenile chronic arthritis. Lancet 1994; 344:1052–1054.

136. Cimmino MA, Parisi M, Querci G, Ghio R, Accardo S. Intra-venous iron is effective in treating the anaemia of rheumatoidarthritis and is not associated with flares of synovitis [letter].Clin Rheumatol 1997; 16:215–216.

188 Beguin

137. Davis D, Charles PJ, Potter A, Feldmann M, Maini RN,Elliott MJ. Anaemia of chronic disease in rheumatoid arthri-tis: in vivo effects of tumour necrosis factor alpha blockade.Br J Rheumatol 1997; 36:950–956.

138. Zucker S, Lysik R. Bone marrow erythropoiesis in anemia ofinflammation. J Lab Clin Med 1974; 84:620–631.

139. DeGowin RL, Gibson DP. Erythropoietin and the anemia ofmice bearing extramedullary tumor. J Lab Clin Med 1979;94:303–311.

140. Dainiak N, Kulkarni V, Howard D, Kalmanti M, Dewey MC,Hoffman R. Mechanisms of abnormal erythropoiesis in malig-nancy. Cancer 1983; 51:1101–1106.

141. Ward HP, Kurnick JE, Pisarczyk MJ. Serum level of erythro-poietin in anemias associated with chronic infection, malig-nancy, and primary hematopoietic disease. J Clin Invest 1971;50:332–335.

142. Cox R, Musial T, Gyde OH. Reduced erythropoietin levelsas a cause of anaemia in patients with lung cancer. Eur JCancer Clin Oncol 1986; 22:511–514.

143. Zarrabi MH, Lysik R, DiStefano J, Zucker S. The anaemia ofchronic disorders: studies of iron reutilization in the anaemiaof experimental malignancy and chronic inflammation. Br JHaematol 1977; 35:647–658.

144. Zucker S, Friedman S, Lysik RM. Bone marrow erythropoi-esis in the anemia of infection, inflammation, and malig-nancy. J Clin Invest 1974; 53:1132–1138.

145. Lockner D. Iron absorption and serum erythropoietin concen-tration in patients with cancer of the uterine cervix. ActaMed Scand 1966; 180:651–655.

146. Nseyo UO, Williams PD, Murphy GP. Clinical significance oferythropoietin levels in renal carcinoma. Urology 1986;28:301–306.

147. Schreuder WO, Ting WC, Smith S, Jacobs A. Testosterone,erythropoietin and anaemia in patients with disseminatedbronchial cancer. Br J Haematol 1984; 57:521–526.


148. Erslev AJ, Wilson J, Caro J. Erythropoietin titers in anemic,nonuremic patients. J Lab Clin Med 1987; 109:429–433.

149. Leng HM, Albrecht CF, Kidson SH, Folb PI. Erythropoietinproduction in anemia associated with experimental cancer.Exp Hematol 1999; 27:806–810.

150. Miller CB, Jones RJ, Piantadosi S, Abeloff MD, Spivak JL.Decreased erythropoietin response in patients with theanemia of cancer. N Engl J Med 1990; 322:1689–1692.

151. Ponchio L, Beguin Y, Farina G, Pedrazzoli P, Pedrotti C,Poggi G, Rosti V, Bergamaschi G, Battistel V, Cazzola M.Evaluation of erythroid marrow response to recombinanthuman erythropoietin in patients with cancer anaemia.Haematologica 1992; 77:494–501.

152. Abels RI. Recombinant human erythropoietin in the treatmentof the anaemia of cancer. ActaHaematol 1992; 87(suppl 1):4–11.

153. Abels RI. Use of recombinant human erythropoietin in thetreatment of anemia in patients who have cancer. SeminOncol 1992; 19:29–35.

154. Cascinu S, Fedeli A, Del Ferro E, Luzi Fedeli S, Catalano G.Recombinant human erythropoietin treatment in cisplatin-associated anemia: a randomized, double-blind trial withplacebo. J Clin Oncol 1994; 12:1058–1062.

155. Osterborg A, Boogaerts MA, Cimino R, Essers U, HolowieckiJ, Juliusson G, Jager G, Najman A, Peest D. Recombinanthuman erythropoietin in transfusion-dependent anemicpatients with multiple myeloma and non-Hodgkin’s lym-phoma. A randomized multicenter study. Blood 1996; 87:2675–2682.

156. Cazzola M, Messinger D, Battistel V, Bron D, Cimino R,Enller-Ziegler L, Essers U, Greil R, Grossi A, Jager G,LeMevel A, Najman A, Silingardi V, Spriano M, van HoofA, Ehmer B. Recombinant human erythropoietin in theanemia associated with multiple myeloma or non-Hodgkin’slymphoma: dose finding and identification of predictorsof response. Blood 1995; 86:4446–4453.

157. Platanias LC, Miller CB, Mick R, Hart RD, Ozer H, McEvillyJM, Jones RJ, Ratain MJ. Treatment of chemotherapy-

190 Beguin

induced anemia with recombinant human erythropoietin incancer patients. J Clin Oncol 1991; 9:2021–2026.

158. Cazzola M, Ponchio L, Beguin Y, Rosti V, Bergamaschi G,Liberato NL, Fregoni V, Nalli G, Barosi G, Ascari E. Subcu-taneous erythropoietin for treatment of refractory anemiain hematologic disorders. Results of a phase I=II clinical trial.Blood 1992; 79:29–37.

159. Oster W, Herrmann F, Gamm H, Zeile G, Lindemann A, Mul-ler G, Brune T, Kraemer HP, Mertelsmann R. Erythropoietinfor the treatment of anemia of malignancy associated withneoplastic bone marrow infiltration. J Clin Oncol 1990; 8:956–962.

160. Kettelhack C, Schoter D, Matthias D, Schlag PM. Serum ery-thropoietin levels in patients with solid tumours. Eur J Can-cer 1994; 30A:1289–1291.

161. Kalmanti M, Kalmantis T. Committed erythroid progenitorsand erythropoietin levels in anemic children with lymphomasand tumors. Pediatr Hematol Oncol 1989; 6:85–93.

162. Chudek J, Wiecek A, Kokot F, Grochal A, Switala J, Kanwis-zer J. Serum erythropoietin concentration in women withuterine or ovarian tumors. Przegl Lek 1998; 55:47–50.

163. Dowlati A, R’Zik S, Fillet G, Beguin Y. Anaemia of lungcancer is due to impaired erythroid marrow response to ery-thropoietin stimulation as well as relative inadequacy oferythropoietin production. Br J Haematol 1997; 97:297–299.

164. Lee SJ, Kwon JH, Jung CW. Erythropoietin response isinadequate in cancer patients receiving chemotherapy. IntJ Hematol 2001; 74:416–420.

165. Corazza F, Beguin Y, Bergmann P, Andre M, Ferster A,Devalck C, Fondu P, Buyse M, Sariban E. Anemia in childrenwith cancer is associated with decreased erythropoietic activ-ity and not with inadequate erythropoietin production. Blood1998; 92:1793–1798.

166. Sawabe Y, Iida S, Tabata Y, Yonemitsu H. Serum erythro-poietin measurements by a one-step sandwich enzyme linkedimmunosorbent assay in patients with hepatocellular carci-noma and liver cirrhosis. Jpn J Clin Oncol 1993; 23:273–277.


167. MalaguarneraM, Bentivegna P, Di Fazio I, Laurino A, RomanoM, Trovato BA. Erythropoietin in hepatocellular carcinoma.Bull Cancer 1996; 83:977–980.

168. Vadhan-Raj S, HittelmanWN, Lepe-Zuniga JL, Gutterman JU,Broxmeyer HE. Regulation of endogenous erythropoietin levelsin anemia associated with myelodysplastic syndromes [letter].Blood 1990; 75:1749–1750.

169. Verhoef GE, De Schouwer P, Ceuppens JL, Van Damme J,Goossens W, Boogaerts MA. Measurement of serum cytokinelevels in patients with myelodysplastic syndromes. Leukemia1992; 6:1268–1272.

170. Jacobs A, Janowska-Wieczorek A, Caro J, Bowen DT, LewisT. Circulating erythropoietin in patients with myelodysplas-tic syndromes. Br J Haematol 1989; 73:36–39.

171. Bowen DT, Jacobs A, Cotes PM, Lewis TC. Serum erythro-poietin and erythropoiesis in patients with myelodysplasticsyndromes. Eur J Haematol 1990; 44:30–32.

172. Bowen DT, Culligan D, Beguin Y, Kendall R, Willis N. Esti-mation of effective and total erythropoiesis in myelodysplasiausing serum transferrin receptor and erythropoietin concen-trations, with automated reticulocyte parameters. Leukemia1994; 8:151–155.

173. Pasqualetti P, Collacciani A, Casale R. Circadian rhythm ofserum erythropoietin in myelodysplastic syndromes. EurRev Med Pharmacol Sci 2000; 4:111–115.

174. Wallvik J, Stenke L, Bernell P, Nordahl G, Hippe E, Hast R.Serum erythropoietin (EPO) levels correlate with survivaland independently predict response to EPO treatment inpatients with myelodysplastic syndromes. Eur J Haematol2002; 68:180–185.

175. Piedras J, Hernandez G, Lopez-Karpovitch X. Effect of andro-gen therapy and anemia on serum erythropoietin levels inpatients with aplastic anemia and myelodysplasticsyndromes. Am J Hematol 1998; 57:113–118.

176. McMullin MF, Hillmen P, Elder GE, Lappin TR, Luzzatto L.Serum erythropoietin levels in paroxysmal nocturnal haemo-

192 Beguin

globinuria: implications for therapy [see comments]. Br JHaematol 1996; 92:815–817.

177. Nakakuma H, Nagakura S, Kawaguchi T, Horikawa K,Iwamoto N, Kagimoto T, Takatsuki K. Markedly high plasmaerythropoietin and granulocyte-colony stimulating factorlevels in patients with paroxysmal nocturnal hemoglobi-nuria. Int J Hematol 1997; 66:451–457.

178. Barosi G, Liberato LN, Guarnone R. Serum erythropoietin inpatients with myelofibrosis with myeloid metaplasia. Br JHaematol 1993; 83:365–369.

179. Najean Y, Schlageter MH, Toubert ME, Rain JD. Erythro-poietin concentration in the serum from patients withprimary thrombocythaemia [letter]. Eur J Haematol 1995;55:272–273.

180. Andreasson B, Lindstedt G, Kutti J. Plasma erythropoietinin essential thrombocythaemia: at diagnosis and in responseto myelosuppressive treatment. Leuk Lymphoma 2000; 38:113–120.

181. Messinezy M, Westwood NB, El Hemaidi I, Marsden JT,Sherwood RS, Pearson TC. Serum erythropoietin valuesin erythrocytoses and in primary thrombocythaemia. Br JHaematol 2002; 117:47–53.

182. Griesshammer M, Kubanek B, Beneke H, Heimpel H,Bangerter M, Bergmann L, Schrezenmeier H. Serum ery-thropoietin and thrombopoietin levels in patients with essen-tial thrombocythaemia. Leuk Lymphoma 2000; 36:533–538.

183. Jansen JH, Wientjens GJ, Fibbe WE, Ralph P, Van Damme J,den Ottolander GJ, Willemze R, Kluin Nelemans JC. Serummonocyte colony-stimulating factor, erythropoietin and inter-leukin-6 in relation to pancytopenia in hairy cell leukemia.Leukemia 1992; 6:735–737.

184. Mauro FR, Gentile M, Foa R. Erythropoietin and chronic lym-phocytic leukemia. Rev Clin Exp Hematol Suppl 2002; 1:21–31.

185. Beguin Y, Lampertz S, Bron D, Fillet G. Serum erythropoie-tin in chronic lymphocytic leukemia. Br J Haematol 1996;93:154–156.


186. JohannsenH, JelkmannW,WiedemannG, OtteM,Wagner T.Erythropoietin=haemoglobin relationship in leukaemia andulcerative colitis. Eur J Haematol 1989; 43:201–206.

187. JelkmannW, JohannsenH,WiedemannG, OtteM,Wagner T.Dependence of serum erythropoietin level on erythropoiesis inleukemia. Hamatol Bluttransfus 1990; 33:83–86.

188. Jelkmann W, Wiedemann G. High serum immunoreactiveerythropoietin in leukaemic patients with bone marrowinsufficiency of erythropoiesis [letter]. Eur J Haematol1990; 45:271–272.

189. Hellebostad M, Marstrander J, Slrdahl SH, Cotes PM,Refsum HE. Serum immunoreactive erythropoietin in chil-dren with acute leukaemia at various stages of disease—and the effects of treatment. Eur J Haematol 1990; 44:159–164.

190. Kivivuori SM, Viinikka L, Teppo AM, Siimes MA. Serumtransferrin receptor and erythropoiesis in children withnewly diagnosed acute lymphoblastic leukemia. Leuk Res1994; 18:823–828.

191. Pohl C, Schobert I, Moter A, Woll EM, Schwonzen M,Hiersche A, Diehl V. Serum erythropoietin levels in patientswith Hodgkin’s lymphoma at the time of diagnosis. AnnOncol 1992; 3:172–173.

192. Beguin Y. Erythropoiesis and erythropoietin in multiple mye-loma. Leuk Lymphoma 1995; 18:413–421.

193. Paaske Hansen O, Drivsholm A. Interrelationships betweenblood volume, venous hematocrit and renal failure in myelo-matosis. Scand J Haematol 1978; 20:461–466.

194. Birgens HS, Paaske Hansen O, Henriksen JH, Wantzin P.Quantitation of erythropoiesis in myelomatosis. Scand JHaematol 1979; 22:357–363.

195. Nielsen OJ, Brandt M, Drivsholm A. The secretory erythro-poietin response in patients with multiple myeloma andWaldenstrom’s macroglobulinaemia. Scand J Clin Lab Invest1990; 50:697–703.

196. Majumdar G, Westwood NB, Bell-Witter C, Muggleston D,Phillips J, Pearson TC. Serum erythropoietin and circulating

194 Beguin

BFU-E in patients with multiple myeloma and anaemia butwithout renal failure. Leuk Lymphoma 1993; 9:173–176.

197. Ariad S, Clifford D, Penfold G, MacPhail AP, Bezwoda WR.Erythropoietin response in anaemic patients with multiplemyeloma and other lymphoid malignancies infiltrating thebone marrow. Eur J Haematol 1992; 49:59–62.

198. Paaske Hansen O, Thorling EB, Drivsholm A. Serumerythropoietin in myelomatosis. Scand J Haematol 1977; 19:106–110.

199. Singh A, Eckardt KU, Zimmermann A, Gotz KH, Hamann M,Ratcliffe PJ, Kurtz A, Reinhart WH. Increased plasma viscos-ity as a reason for inappropriate erythropoietin formation.J Clin Invest 1993; 91:251–256.

200. Host H, Skjaelaaen P. Comparative effects of cyclophospha-mide and total body irradiation on erythropoietin productionin rats. Scand J Haematol 1966; 3:154–157.

201. McDonald TP, Lange RD, Congdon CC, Toya RE. Effect ofhypoxia, irradiation, and bone marrow transplantation onerythropoietin levels in mice. Radiat Res 1970; 42:151–163.

202. Magid E, Hansen P. The erythropoietin content of plasmafrom sheep treated with nitrogen mustard. Scand J ClinLab Invest 1966; 18:347–352.

203. Chakraborty A, Chatterjee M. Enhanced erythropoietin andsuppression of gamma-glutamyl transpeptidase (ggt) activityin murine lymphoma following administration of vanadium.Neoplasma 1994; 41:291–296.

204. Fisher JW, Roh BL. Influence of alkylating agents on kidneyerythropoietin production. Cancer Res 1964; 24:983–988.

205. Wolff M, Jelkmann W. Effects of chemotherapeutic andimmunosuppressive drugs on the production of erythropoie-tin in human hepatoma cultures. Ann Hematol 1993; 66:27–31.

206. Jelkmann W, Wolff M, Fandrey J. Inhibition of erythropoie-tin production by cytokines and chemotherapy may contri-bute to the anemia in malignant diseases. Adv Exp MedBiol 1994; 345:525–530.


207. Von Hoff DD, Schilsky R, Reichert CM, Reddick RL,Rozencweig M, Young RC, Muggia FM. Toxic effects of cis-dichlorodiammineplatinum(II) in man. Cancer Treat Rep1979; 63:1527–1531.

208. HoriguchiH,KayamaF,OgumaE,WillmoreWG,HradeckyP,BunnHF. Cadmium and platinum suppression of erythropoie-tin production in cell culture: clinical implications. Blood 2000;96:3743–3747.

209. Rothman SA, Paul P, Weick JK, McIntyre WR, Fantelli F.Effect of cis-diamminedichloroplatinum on erythropoietinproduction and hematopoietic progenitor cells. Int J CellCloning 1985; 3:415–423.

210. Unami A, Nishina N, Terai T, Sato S, Tamura T, Noda K,Mine Y. Effects of cisplatin on erythropoietin production inrats. J Toxicol Sci 1996; 21:157–165.

211. Vaziri ND, Zhou XJ, Liao SY. Erythropoietin enhances recov-ery from cisplatin-induced acute renal failure. Am J Physiol1994; 266:F360–F366.

212. Baldwin MD, Zhou XJ, Ing TS, Vaziri ND. Erythropoietinameliorates anemia of cisplatin induced acute renal failure.ASAIO J 1998; 44:44–47.

213. Matsumoto T, Endoh K, Kamisango K, Akamatsu K, KoizumiK, Higuchi M, Imai N, Mitsui H, Kawaguchi T. Effect ofrecombinant human erythropoietin on anticancer drug-induced anaemia. Br J Haematol 1990; 75:463–468.

214. Wood PA, Hrushesky WJ. Cisplatin-associated anemia: anerythropoietin deficiency syndrome. J Clin Invest 1995; 95:1650–1659.

215. Sawabe Y, Kikuno K, Iseki T, Lida S, Tabata Y, Yonemitsu H.Changes in serum erythropoietin and the reticulocyte countduring chemotherapy for leukemias. Eur J Haematol 1996;57:384–388.

216. Sawabe Y, Kikuno K, Iseki T, Iida S, Yonemitsu H. Serumerythropoietin values and serum iron status duringchemotherapy for leukemia. Eur J Haematol 1998; 60:315–316.

196 Beguin

217. Sawabe Y, Takiguchi Y, Kikuno K, Iseki T, Ito J, Iida S, Kur-iyama T, Yonemitsu H. Changes in levels of serum erythro-poietin, serum iron and unsaturated iron binding capacityduring chemotherapy for lung cancer. Jpn J Clin Oncol1998; 28:182–186.

218. Cerruti A, Castello G, Balleari E, Bogliolo G, Lerza R, Pan-nacciuli I. Serum erythropoietin increase in patients receiv-ing adjuvant therapy with 5-fluorouracil and leucovorin.Exp Hematol 1994; 22:1261–1263.

219. Miranda CE, Scaro JL, Buys MC, Torrejon I, Martin B,Guerra L. Urinary elimination of erythropoietin in patientsunder treatment with cytostatics drugs. Acta PhysiolPharmacol Ther Latinoam 1998; 48:207–210.

220. Kinjo K, Kizaki M, Takayama N, Michikawa N, Oda A,Okamoto S, Tahara T, Kato T, Miyazaki H, Ikeda Y. Serumthrombopoietin and erythropoietin levels in patients withacute promyelocytic leukaemia during all-trans retinoic acidtreatment. Br J Haematol 1999; 105:382–387.

221. Franek E, Marcinkowski W, Kokot F, Wiecek A, Nowicki M.A single dose of cyclophosphamide (Cph) does increase ery-thropoietin concentration in patients with hypertension dueto vasculitis [letter]. Clin Nephrol 1994; 42:139–140.

222. Katschinski DM, Jelkmann W, Wiedemann GJ, Mentzel M,Mulkerin DL, Touhidi R, Robins HI. Dynamic changes inserum erythropoietin levels in solid tumour patients under-going 41.8�C whole body hyperthermia and=or chemother-apy. Int J Hyperthermia 1997; 13:563–569.

223. Shamseddine A, Medawar W, Seoud M, Ibrahim K, Habbal Z,Kahwaji S, Khalil A. The relationship between serum levelsof erythropoietin (EPO) and insulin-like growth factor-1(ILGF-1) and hematocrit (HCT) in breast cancer patientsreceiving non-nephrotoxic chemotherapy. Eur J GynaecolOncol 1998; 19:591–593.

224. Shamseddine A, Khalil A, Seoud M, Kahwaji S, Taher A,Bizri AR, Medawar W. The relation between erythropoietin,hematocrit and hemoglobin in breast cancer patients onnon-nephrotoxic chemotherapy. Eur J Gynaecol Oncol 1998;19:577–579.


225. Dowd MD, Morgan ER, Langman CB, Murphy S. Serum ery-thropoietin levels in children with leukemia. Med PediatrOncol 1997; 28:259–267.

226. Lechner W, Artner Dworzak E, Solder E, Sachsenmaier M,Kolle D, Moncayo H, Reitsamer R. Influence on erythropoie-tin levels of treatment with cisplatinum-endoxan. Arch Gyne-col Obstet 1992; 252:49–53.

227. Saijo Y, Nakai Y, Saito J, Sugawara S, Suzuki S, Numata Y,Motomiya M. Changes in serum erythropoietin levels duringchemotherapy for lung cancer. Chemotherapy 1992; 38:281–285.

228. Onat H, Inanc SE, Dalay N, Karaloglu D, Erturk N,Yasasever V. Effect of cisplatin on erythropoietin and ironchanges [letter]. Eur J Cancer 1993; 29A:777.

229. Canaparo R, Casale F, Muntoni E, Zara GP, Della PC, Berno E,Pons N, Fornari G, Eandi M. Plasma erythropoietin concentra-tions in patients receiving intensive platinum or nonplatinumchemotherapy. Br J Clin Pharmacol 2000; 50:146–153.

230. Pedain C, Herrero J, Kunzel W. Serum erythropoietin levelsin ovarian cancer patients receiving chemotherapy. Eur JObstet Gynecol Reprod Biol 2001; 98:224–230.

231. Orhan B, Yalcin S, Evrensel T, Kurt E,Manavoglu O, Erbas T.Does cisplatin stimulate erythropoietin secretion from theperitubular cells of the kidney? [letter]. Clin Nephrol 1998;50:202–203.

232. Hasegawa I, Tanaka K. Serum erythropoietin levels in gyne-cologic cancer patients during cisplatin combination che-motherapy. Gynecol Oncol 1992; 46:65–68.

233. Smith DH, Goldwasser E, Vokes EE. Serum immunoerythro-poietin levels in patients with cancer receiving cisplatin-based chemotherapy. Cancer 1991; 68:1101–1105.

234. Bray GL, Reaman GH. Erythropoietin deficiency: a complica-tion of cisplatin therapy and its treatment with recombinanthuman erythropoietin. Am J Pediatr Hematol Oncol 1991;13:426–430.

198 Beguin

235. Cazzola M, Mercuriali F, Brugnara C. Use of recombinanthuman erythropoietin outside the setting of uremia. Blood1997; 89:4248–4267.

236. Fischl M, Galpin JE, Levine JD, Groopman JE, Henry DH,Kennedy P, Miles S, Robbins W, Starrett B, Zalusky R, AbelsRI, Tsai HC, Rudnick SA. Recombinant human erythropoie-tin for patients with AIDS treated with zidovudine. N EnglJ Med 1990; 322:1488–1493.

237. Pincus T, Olsen NJ, Russell IJ, Wolfe F, Harris ER, SchnitzerTJ, Boccagno JA, Krantz SB. Multicenter study of recombi-nant human erythropoietin in correction of anemia inrheumatoid arthritis. Am J Med 1990; 89:161–168.

238. Ludwig H, Fritz E, Leitgeb C, Pecherstorfer M, Samonigg L,Schuster J. Prediction of response to erythropoietintreatment in chronic anemia of cancer. Blood 1994; 84:1056–1063.

239. Cazzola M, Ponchio L, Pedrotti C, Farina G, Cerani P,Lucotti C, Novella A, Rovati A, Bergamaschi G, Beguin Y. Pre-diction of response to recombinant human erythropoietin(rHuEpo) in anemia of malignancy. Haematologica 1996; 81:434–441.

240. Oberhoff C, Neri B, Amadori D, Petry KU, Gamucci T,Rebmann U, Nowrousian MR, Voigtmann R, Monfardini S,Armand JP, Herrmann R, Netter-Pinon J, Tubiana-Mathieu N, Zwierzina H. Recombinant human erythropoietinin the treatment of chemotherapy-induced anemia and pre-vention of transfusion requirement associated with solidtumors: a randomized, controlled study [see comments].Ann Oncol 1998; 9:255–260.

241. ten Bokkel Huinink WW, de Swart CA, van Toorn DW,Morack G, Breed WP, Hillen HF, van der Hoeven JJ,Reed NS, Fairlamb DJ, Chan SY, Godfrey KA, KristensenGB, van Tinteren H, Ehmer B. Controlled multicentre studyof the influence of subcutaneous recombinant human erythro-poietin on anaemia and transfusion dependency in patientswith ovarian carcinoma treated with platinum-basedchemotherapy [see comments]. Med Oncol 1998; 15:174–182.


242. Charuruks N, Voravud N, Limpanasithikul W. Ratio ofbaseline erythropoietin (EPO) level and corrected reticulo-cyte count as an indicator for a favourable response to recom-binant human erythropoietin (rhEPO) therapy in anaemiccancer patients. J Clin Lab Anal 2001; 15:260–266.

243. Leon MP, Jimenez MM, Barona ZP, Riol DM, Castro PL, Sier-rasesumaga AL. Recombinant human erythropoietin in anemiaassociated with pediatric cancer: study of the identification ofpredictors of response. An Esp Pediatr 1998; 49:17–22.

200 Beguin

7

Erythrophagocytosis and DecreasedErythrocyte Survival

J. J. M. MARX

Eijkman–Winkler Centre for Microbiology,Infectious Diseases and Inflammation,

University Medical Centre Utrecht,Utrecht, The Netherlands

INTRODUCTION

Most body iron is utilized for production of erythrocytes:about 20 mg per day. This is much more than the amount ofapproximately 1 mg iron that needs to be absorbed from thegut to compensate for normal daily iron losses. After releasefrom the bone marrow erythrocytes circulate for 100–120days. During this period, they show physiological aging andare finally selected by macrophages of the mononuclear pha-gocyte system (MPS) for destruction. Iron is removed fromhemoglobin (Hb) and either released to the plasma for

201

redistribution and production of heme-, and other iron pro-teins, or stored in ferritin and hemosiderin. In conditions ofhemolysis, erythrocyte life span is shortened and erythropha-gocytosis is accelerated, leading to increased iron processingand iron overload. Erythrophagocytosis and MPS iron releaseare also modified in inflammation and hemochromatosis.Our understanding of iron handling by macrophages hasincreased considerably in recent years due to the discoveryof several new genes involved in iron transport, and quantita-tive analysis of macrophage iron transport. The impact ofacute and chronic changes in erythrophagocytosis and macro-phage iron release on iron kinetics and internal iron distribu-tion are much more important than changes in ironabsorption, due to the vast amount of iron that is handledby the MPS each day.

THE PHYSIOLOGY OF ERYTHROCYTE AGING

After transition from erythroblasts to enucleated cells reti-culocytes remain for 3–4 days in the bone marrow. Cytoske-leton structures including microtubules and microfilamentsplay key roles in the genesis of the anucleate reticulocytefrom its nucleated precursor cell, as well as in the earlystages of reticulocyte development (1). In the circulationreticulocytes are transformed into erythrocytes within twodays. During maturation of reticulocyte into red cellchanges in cell shape and extensive remodelling of themembrane skeleton take place, resulting in the matureerythrocyte with its highly deformable yet remarkablystable membrane.

During their life span erythrocytes undergo continuouschanges. Reliable data on the changes of red blood cells (RBCs)during aging can be obtained by fractionation of RBC popula-tions with a combination of counterflow centrifugation andpercoll separation (2). Approximately 20% of hemoglobin islost from the circulating RBC during its life span (3). It wascalculated that there is a net loss of 385 amol of hemoglobinper cell. Hemoglobin is lost from RBCs of all ages, but

202 Marx

predominantly from older cells. Hemoglobin-containing vesi-cles circulate in plasma (4), and erythrocytes contain vesicles,which are filled with hemoglobin. In a recent study, suchvesicles were isolated from plasma and studied by electronmicroscopy and flow cytometry. In addition, their hemoglobincomposition was determined and compared with that of RBCfractions of different ages (5). It appeared that in old RBCsof asplenic individuals, the decrease of hemoglobin contentper cell such as normally seen in old RBCs is absent, and thathemoglobin-containing vesicles within old RBCs are ‘‘pitted’’by the spleen. The spleen, therefore, plays an important rolein this initial step of cell aging, during which erythrocytesdecrease cell surface area and volume. MCV and MCHdecrease during RBC aging.

Erythrocytes, containing no mitochondria, depend fortheir energy production totally on ATP generated duringglycolysis. Defects of glycolytic enzymes may cause dysfunctionof energy-dependent membrane transport, and impaired resis-tance against oxidative stress, due to a decrease of reduced glu-tathione, generated via the hexose monophosphate pathway,an important scavenger of hydrogen peroxide. During agingof erythrocytes concentration of glycolytic enzymes is markedlydecreasing. However, the decrease of glycolytic and otherenzymes, and even some enzyme deficiencies, is well toleratedby the RBC, with no impairment in function. The capacity of theerythrocyte to tolerate deficiencies in most enzymes indicateseither that the metabolic pathways, which the enzyme servesare not required by the red cell or that redundancies in metabo-lism exist which allow the erythrocyte to compensate for theenzyme deficiency (6).

There is no full explanation yet for the physiological age-dependent clearance of RBCs from circulation. One importantfactor is certainly formation of denatured=oxidized hemoglo-bin (hemichromes) arising late during an RBC’s life spaninducing clustering of the integral membrane protein, band3. In turn, band 3 clustering generates an epitope on thesenescent cell surface leading to autologous IgG binding andconsequent phagocytosis. Interestingly RBCs that wereallowed to senesce for 115 days in vivo also suffered from

Erythrophagocytosis and Decreased Erythrocyte Survival 203

compromised intracellular reducing power, containing only30% of the reduced glutathione found in unfractionated cells,making RBCs more vulnerable for oxygen radical damage (7).

Surface exposure of phosphatidylserine (PS), normallylocalized to the inner leaflet of cell membranes, may contri-bute to disappearance of erythrocytes from the circulationbecause of normal aging, signalling macrophages to ingestthem. Exposure of PS only occurs in old erythrocytes. A modelwas proposed of red cell senescence that assumes both anage-dependent destruction of senescent red cells precededby several hours of PS exposure and a random destructionof red cells without PS exposure. It was demonstrated thatthe exposure of PS parallels the rate at which biotinylatedred cells are removed from circulation. On the other hand,exposed PS does not cause the reduced red cell life span ofpatients with hemolytic anemia, being too young to exposePS, with the possible exception of those with unstable hemo-globins or sickle cell anemia (8). The many investigations onRBC aging indicate that the trigger for physiological seques-tration from the circulation is multifactorial.

METHODS FOR ESTIMATIONOF ERYTHROCYTELIFE SPAN

The RBC survival and life span can be measured with radio-isotopes and nonradioactive procedures. Red cells can beradio-labeled for survival studies by two different methods:(a) in vivo cohort labelling of newly formed cells by injectionof transferrin-bound 59Fe, and (b) in vitro at random labellingof circulating red cells with 51Cr (9). The standard method isestimation of disappearance from the circulation of 51Cr-labelled erythrocytes over a period of several weeks. In vivoelution of 51Cr influences the measurements of RBC survivaland life span of RBCs (10). It remained the method of choice,however, because the 59Fe method never became popular dueto the need for complicated mathematical calculations (11).

Because the standard 51Cr method exposes the patient toradiation, alternative methods for measuring RBC survival

204 Marx

were developed. One method is based on determining thenumber of biotin-labelled red cells that persist in the circula-tion by using fluorescein-labeled avidin and flow cytometry.Biotinylated cells persist in the circulation with life spansapproaching normal. A problem is that about one-half of thebiotin label leaves the red cells and the circulation over thefirst few weeks, causing early curvilinear disappearance.The other half, however, remains permanently attached,which produced linear disappearance and approximatelynormal life span estimates for the linear survival curveappearing after the first few weeks. In that period red cellsurvival can be measured accurately in humans usingenumeration of biotinylated red cells (12).

A simple and rapid nonradioactive method was describedfor determining RBC life span based on measurement ofalveolar carbon monoxide (CO) concentration corrected foratmospheric CO as determined with a device that simulatesthe body’s equilibration with CO. Using this technique, itwas found that the RBC life span of 40 healthy volunteersaveraged 122� 23 days, a value comparable to that obtainedwith complex cohort labelling methods. The ability of thissimple technique to detect increased RBC turnover wasdemonstrated in four subjects being treated with ribavirinand interferon for hepatitis C, a treatment reported toshorten RBC life span. Measurement of CO in samplescollected by these four subjects in their home environmentsindicated that each had a shortened RBC life span (range30–69 days) (13).

CAUSES OF DECREASED ERYTHROCYTESURVIVAL AND LIFE SPAN

The life span of RBC can be considerably decreased in hemo-lytic anemias. The Hb in mild hemolysis may remain normalif it is sufficiently compensated by increased erythropoiesis.Early destruction of RBCs is possible in the circulation (intra-vascular hemolysis) or by accelerated erythrophagocytosis,mainly in the spleen (extravascular hemolysis). In some


anemias destruction of erythrocyte precursors can alreadytake place in the bone marrow (dyserythropoiesis). Any com-bination of these forms of hemolysis is possible, depending onthe cause of early RBC destruction. The cause of hemolyticanemia can be inherited or acquired. Defects concern ingeneral components of the erythrocyte that are graduallymodulated during physiological aging: the RBC membrane,hemoglobin, glycolytic enzymes, and oxygen radical scaven-gers. External attacks on RBC function and integrity arepossible by physical, chemical, microbial, and immunologicalcauses. Also hypersplenism can result in acceleratederythrocyte breakdown.

Erythrocytes undergo apoptosis upon increase of cytosolicCa2þ activity. Erythrocytes from healthy individuals, frompatients with sickle cell anemia, thalassemia, or glucose-6-phosphate dehydrogenase deficiency all respond to osmoticshock, to oxidative stress and to energy depletion withenhanced annexin binding. However, the sensitivity of sicklecells and of glucose-6-phosphate dehydrogenase deficient cellsto osmotic shock and of sickle cells, thalassemic cells and glu-cose-6-phosphate dehydrogenase deficient cells to oxidativestress and to glucose depletion is significantly higher than thatof control cells, leading to enhanced apoptosis and to the shor-tened life span of defective erythrocytes (14).

EFFECT OF INFLAMMATION ON ERYTHROCYTESURVIVAL

Impaired erythropoietin production and impaired responsive-ness of erythroid progenitor cells to this hormone are impor-tant abnormalities contributing to the anemia of chronicdisease (ACD), due to the effects of inflammatory cytokines.The contribution of disordered iron metabolism to anaemia,although characteristic of ACD, may be less important. It iswell established that iron absorption is reduced (15), and thatiron administered intravenously is rapidly sequestered in themononuclear phagocyte system (MPS). Iron delivery to thebone marrow is not impaired, erythroid iron utilization is

206 Marx

not markedly depressed in ACD, and recombinanterythropoietin therapy can correct ACD, but it cannot correctthe anemia due to iron deficiency. It is likely that abnormal-ities such as reduced iron absorption and decreased erythro-blast transferrin-receptor expression largely result fromdecreased erythropoietin production and inhibition of its activ-ity by inflammatory cytokines (16).

A shortening of red cell life span is often suggested tooccur in ACD. The inflammatory state may have a negativeinfluence on erythrocyte survival as reported in patients withrheumatoid arthritis (17). This could not be confirmed in aclean study inducing controlled inflammation by surgeryin otherwise healthy patients who underwent total hipreplacement. The disappearance of 51Cr-labeled RBCs wascalculated for the preoperative period and for the period frompostoperative Day 4 on, when blood loss had ceased. The half-life of RBCs did not change if estimated before (29.0� 4.4days) and after (27.4 � 3.6 days, p¼ 0.55)surgery, indicatingthat RBC life span is not influenced by the surgery inducedinflammation (18).

UPTAKE OF EFFETE ERYTHROCYTES BY THEMACROPHAGE SYSTEM

Monocytes are produced in the bone marrow. After maturationand release to the blood, they rapidly disappear from the circu-lation to home in a variety of different tissues where theydevelop into macrophages. These cells have an impressivenumber of important functions in the immune system, whichmay be different for macrophages that settle in differentorgans. Even in the same organ, macrophages can be foundwith a specialized set of functions. Erythrocytes and othereffete blood cells are degraded in the mononuclear phagocyticsystem (MPS), mainly localized in the spleen, the liver, andbone marrow. Macrophages in other organs like lung and skin,however, are also capable of performing erythrophagocytosis.Monocytes (MN) and macrophages are able to process erythro-cytes, containing vast amounts of iron, safely and rapidly.


Many different in vivo systems were used to study ery-throphagocytosis and macrophage iron metabolism. Theseinclude isolation of macrophages from the liver (Kupffer cells),lungs, and peritoneum or using cell lines. As erythrophagocy-tosis and iron handling are already expressed in sets of periph-eral blood monocytes, this can be considered to be a convenientand reliable model to study erythrophagocytosis. To follow thewhole process of erythrophagocytosis, experiments usingother iron sources than intact erythrocytes are less suitable.The uptake of homologous erythrocytes by MPS cells can beinvestigated by incubation of peripheral blood monocytes, orof monocyte-derived macrophages (MDM), with opsonized ery-throcytes (19). The uptake of erythrocytes can be studied usingthe microscope or by quantification of in vivo 59Fe-labeledhemoglobin in rabbit red blood cells (RRBCs) (20). To be recog-nized and taken up by MN or MDM, these erythrocytes wereopsonized with heat-inactivated mouse antirabbit erythrocyteserum. MN or MDM must be incubated with an excess of ery-throcytes for a desired number of hours. Noningested RRBCsare removed by performing hypotonic lysis. The extent of pha-gocytosis can be evaluated by light microscopy on Giemsa-stained cytospins as shown in Fig. 1. This experiment demon-strates that already peripheral blood monocytes have the fullcapacity to perform erythrophagocytosis, providing a good toolto study this phenomenon. The figure also shows disappear-ance of hemoglobin from some engulfed erythrocytes. Erythro-phagocytosis can be expressed as phagocytic index (PI), i.e.,the number of erythrocytes taken up per monocyte. It can becalculated that in a normal human MPS macrophages processone erythrocyte per macrophage per day (21). It is our experi-ence that the ratio of one erythrocyte per MN or MDM is alsomost suitable for in vitro studies as a higher number ofengulfed erythrocytes per phagocyte, as visualized in Fig. 1,rapidly decreases the viability of MN or MDM.

Using in vitro techniques, it was found that phagocyticcapacity of MN was about 50% diminished in erythrocytesof patients with hereditary hemochromatosis if comparedwith normal controls (19). The defect was found in all patientsstudied. Phagocytosis of opsonized erythrocytes is an Fc

208 Marx

receptor-mediated process. The defect in HH could not beattributed to differential expression of Fcg or complementreceptors (22).

ERYTHROCYTE DESTRUCTION AND RELEASEOF IRON FROM HEMOGLOBIN ANDMACROPHAGES

After being engulfed by the macrophage, the red blood cell isdigested enzymatically in the phagolysosome. Heme is liber-ated from hemoglobin by proteolysis and transported to the

Figure 1 Erythrophagocytosis by monocytes isolated from humanblood, incubated with opsonized rabbit erythrocytes. Some ingestederythrocytes show a fainting color due to disappearance of hemo-globin. (From Moura E, Verheul AFM, Marx JJM, unpublishedmaterial.)


cytosol. Iron is removed from heme in the endoplasmatic reti-culum (ER) by heme oxygenase (HO), mainly the inducibleform of this enzyme (HO-1 or HSP32), which produces Fe2þ,biliverdin, and carbon monoxide (CO) (23).

There is recent evidence that heme catabolism canalready take place in the phagolysosome itself as part of thephagosomal membrane is derived by fusion with ER duringits biogenesis (24). This implies that phagosomal membranesshould have the capacity to transport iron as free Fe(II), andnot (only) as heme or hemoglobin, directly to the labile ironpool of the macrophage.

After this step, iron can be released from the macrophageto the plasma or stored in ferritin or hemosiderin. The quan-tity of iron storage in macrophages differs depending on thecause of iron accumulation. Under the microscope stored ironcan be easily recognized, using the Perl blue method, as eithera diffuse blue staining (ferritin) or coarse blue dots (hemosi-derin). Little stainable iron is seen under normal conditions.In secondary iron overload as a result of dyserythropoiesis,hemolysis, or transfusions, however, macrophages are heavilyloaded with iron. In contrast, in hereditary hemochromatosis(HH), little iron is seen in the Kupffer cells and other macro-phages. Figure 2 shows an electron micrograph of healthyvolunteer monocytes that were incubated with opsonizederythrocytes. Shortly after phagocytosis enclosed red cellsmaintain a regular shape and are electron dense (Fig. 2A).Phagocytosed erythrocytes show fusion with small lysosomes(Fig. 2B). After about 90 min most erythrocytes show a floccu-lent aspect (Fig. 2C). After 24 hr, no erythrocyte structurescan be detected anymore in MN.

The MPS have effective ways to protect themselves fromoxidative damage by iron released during heme catabolism.Firstly, there is a rapid release of iron derived from MPS(T1=2¼ 33 min), and secondly potentially toxic excess iron isdeposited in ferritin. Also ferritin-bound iron is released,but at a much slower rate (T1=2¼ 6 days) as demonstratedby in vivo studies (25). At the molecular level, the most impor-tant protection may be provided by HO-1 related hemecatabolism itself, producing equimolar amounts of CO (a

210 Marx

signalling molecule), biliverdin, and free Fe(II) (26). HO-1 is astress responsive molecule that is rapidly induced by free andstable radicals as well as by hypoxia. Heme-derived biliverdinis reduced to bilirubin, which is a potent antioxidant. It wasdemonstrated that supra induction of HO-1 completely pro-tects ischemic kidney against tissue injury by rapid inactiva-tion of the pro-oxidant heme of denatured hemoproteins andconverting it to bilirubin and CO.

Figure 2 Electron micrograph of healthy volunteer monocytes(MN) after erythrophagocytosis. (A) Monocyte suspension directlyafter phagocytosis of opsonized rabbit erythrocytes (asterisk), recog-nized as rounded electron-dense bodies inside MN. Some electron-lucent vacuoles (arrow) are visible (magnification 8300�). (B) Detailsof phagocytosed erythrocyte showing fusion with small lysosomes(arrows) (magnification 17,500�). (C)MN 90min after erythrophago-cytosis showing a remnant of an erythrocyte with a flocculant content(asterisk). Several vacuoles (arrows) are present (magnification11,000�). (From Moura E, Verheul AFM, JJM Marx, unpublishedmaterial.)


Release of iron from monocytes and macrophages hasbeen studied at the quantitative and the qualitative level. Asimilar biphasic pattern, compared to in vivo studies, wasseen in vitro using human monocytes, rat peritonealmacrophages and Kupffer cells: a fast release phase of radio-iron to apotransferrin in the medium (within 1.5 hr), followedby a much slower release of iron (21,27–29). In contrast to thegeneral opinion, it is not true that iron is leaving the MPSonly from a cytosolic labile iron pool, apparently as Fe(II),but also as ferritin and hemoglobin. To detect this phenom-enon, it is necessary to load monocytes or macrophages invitro, not with iron containing immune complexes, but withhemoglobin containing erythrocytes, as mentioned in theprevious chapter, because only this approach is similar tothe in vivo situation. After phagocytosis of [59Fe]-hemoglo-bin-labeled erythrocytes MN or MDM can be investigatedfor release of 59Fe at different time points, and the molecularform of iron can be analyzed by size-exclusion high perfor-mance liquid chromatography (SE-HPLC) (29). Releaseexperiments can be continued for 48 hr with high viabilityof MN and MDM.

Earlier in vitro studies of human MN, loaded with ironby erythrophagocytosis, demonstrated iron in the form ofHb and ferritin inside the cell and in the culture supernatant,suggesting that iron was released in these macromolecularforms (27). Also studies using rat peritoneal macrophagesand Kupffer cells detected iron as ferritin and in a low mole-cular weight form that bound to apotransferrin or to desfer-rioxamine (21,28,30). In some of these experiments, Hb wasalso recovered, but this was considered to be an artifact.There is good evidence, however, that iron indeed leavesmacrophages as Fe(II), ferritin, and hemoglobin under phy-siological conditions as cells maintain a high viability duringsuch experiments, and a progressive shift of 59Fe activity fromHb to ferritin inside the cells demonstrates the presence ofmetabolically active and intact cells (29). This investigationdemonstrated that monocytes from control subjects andpatients with homozygote hereditary hemochromatosis (HH)released iron in the forms of ferritin, Hb, and as nonprotein-

212 Marx

bound low molecular weight iron (LMW-Fe) complexes. Ironrelease as Hb is probably a physiologic process occurringwhenever the erythrocyte catabolizing capacity of macro-phages is exceeded. Remarkably, HH monocytes releasedtwice as much iron in the form of LMW-Fe complexes thancontrol monocytes. The finding of increased release ofLMW-Fe in HH might explain the high transferrin saturationand nontransferrin-bound iron in HH. It also allows under-standing why patients with HH have always normal ironstores in macrophages, including hepatic Kupffer cells, inthe early stages of iron overload.

MAJOR PROTEINS INVOLVED IN IRON RELEASEFROM HEMOGLOBIN AND MACROPHAGES

The pathway of iron release from macrophages, and the majorproteins involved, is indicated in Fig. 3. The hepatocyte isincluded in this scheme as this cell takes care of all erythro-phagocytosis-derived forms of iron molecules in plasma thatare released from spleen macrophages and Kupffer cells.

As can be seen in Fig. 1, phagocytozed erythrocytes startto lose pink color, indicating that hemoglobin is removed. Animportant initial step in this process is degradation ofhemoglobin and heme. As mentioned before, the responsibleenzyme is HO-1. Monocytes, which are not exposed to ery-throphagocytosis before, contain only limited amounts ofHO-1, the rate-limiting enzyme of Hb degradation (31). Theearly release of Hb from macrophages might be due to insuffi-cient amounts of (inducible) HO-1 during the first hours aftererythrophagocytosis. Iron release in the form of Hb is prob-ably a normal physiologic process, not only occurring afterintravascular hemolysis, but also after normal erythrophago-cytosis. Indeed, haptoglobin-bound Hb is present in plasma ofhealthy subjects (normal values <40 mg Hb=L plasma). Thereason why the concentration of Hb in plasma remains lowis the fact that hepatocytes possess receptors for haptoglo-bin-complexed and free Hb and also for hemopexin-complexedand free heme. It has been estimated that in humans 113–157


nmol (21.6–30mg) of haptoglobin–Hb complexes are clearedper liter per minute (T1=2 haptoglobin–Hb in the rat is7min) (32–34). It is not yet known which molecules facilitatethe release of intact Hb from the macrophage.

After HO-mediated release from heme, iron leavesthe phagosome toward the cytosol, probably transportedby natural resistance-associated macrophage protein-1(NRAMP1), a protein exclusively expressed in monocytes andmacrophages (35,36). NRAMP1 can be found in lysosomes, lateendosomes, and phagosomes. Some studies suggest thatNRAMP1 transports Fe(II) intomicrobe-containing phagolyso-

Figure 3 Scheme of erythrophagocytosis by spleen macrophage(similar to hepatic Kupffer cells), and communication betweenmacrophage and hepatocyte via portal circulation. Most importantproteins involved in iron transport and storage are included.DMT1, divalent metal transporter-1; HFE, molecule that is mutatedin hereditary hemochromatosis; HO, heme oxygenase-1; NRAMP1,natural resistance-associated macrophage protein-1; NTBI, non-transferrin bound iron; RBC, red blood cell; SFT, stimulator of Fetransport; TrF, transferrin; TfR, transferrin receptor-1; TfR2, trans-ferrin receptor-2.

214 Marx

somes for oxygen radical mediated killing of bacteria (37).Others, however, found that NRAMP1 transports Fe(II) outof the phagosome (38). Another study demonstrated thatNRAMP1 regulates macrophage iron handling, and probablyfacilitates iron release from macrophages undergoing erythro-phagocytosis in vivo (36). This discrepancy of results may beinterpreted as the ability of NRAMP1 to allow bidirectional,concentration-gradient driven transport of Fe(II).

In the reducing environment of the cytosolic labile ironpool iron should be present as Fe(II). This view is supportedby the fact that all following steps in the pathway of ironrequire Fe(II). Most prominent is the release of iron to theplasma by the recently discovered Fe(II) transporter IREG-1,also described as ferroportin (39). A mutation of ferroportinis associated with an autosomal dominant form of genetichemochromatosis, in which iron is retained by macrophages(40). Ferroportin is identical to IREG1, the transporter ofFe(II) across the basolateral membrane of intestinal cells(41). After entrance into the plasma most iron rapidly bindsto free binding sites of (apo)transferrin. As transferrin bindsiron as Fe(III) autoxidation at transferrin is required. Prob-ably because the highly soluble Fe(II) is potentially toxic incatalyzing the formation of hydroxyl radicals from H2O2,hemopexin is readily available to oxidize Fe(II) to form Fe(III),which binds with high affinity to transferrin. Nevertheless,one should be aware that after entrance into plasma frommacrophages, similar to iron entering plasma from intestinalmucosal cells, nontransferrin bound iron (NTBI) is presentfor some time. The concentration can be high in all forms ofprimary and secondary iron overload, with potential toxicityfor vascular endothelium and other target cells (42,43). MostNTBI, however, is removed from the plasma by hepatocytes,using divalent metal transporter-1 (DMT1) and stimulator ofFe transport (SFT) as transporters of Fe(II) (44–46). As alsoNTBI in the form of Fe(III) seems to be transported associatedferrireductase activity may be needed.

After release from the phagolysosome part of the iron istrapped into ferritin, the professional iron storage protein.Expression of ferritin is regulated at the mRNA level by iron


responsive protein (IRP). At a low concentration of labile cyto-solic iron IRP concentration is high and blocks translation offerritin-mRNA. At a low IRP concentration, however, largeamounts of ferritin are formed, which incorporates Fe(II) thatis stored in the ferritin core as Fe(III) hydroxide complexes.This storage iron can be rapidly released towards the labileiron pool and subsequently to the plasma as soon as iron defi-ciency develops or more iron is needed for erythropoiesis.Many investigations have meanwhile established that intactferritin can be released to plasma, although the molecularmechanism remains obscure. Also ferritin can be rapidlycleared from plasma by hepatocytes, which possess ferritinreceptors. It will be evident that the liver plays a prominentrole as an intermediate sink for all kinds of potentially toxiciron molecules as indicated in Fig. 3.

The protein responsible for the release of Fe(II) from themacrophage to the plasma is the recently identified iron-regulated transporter 1 (IREG-1) (41), by other authorsnamed metal-transporter protein 1 (MTP-1) (47), or ferropor-tin-1 (FPN1) (48). Similar to ferritin, its expression is upregu-lated at the mRNA level if labile intracellular iron increases.Ferroportin is highly expressed in liver, spleen, and bonemarrow macrophages, and at the basolateral membrane ofgut mucosal cells. Ferroportin clearly is the most importantiron transporter out of iron-donor cells towards the plasma.Similarly the transferrin receptor (TfR) is the most important(transferrin-bound) iron transporter of iron-acceptor cells:those cells that need to incorporate iron for proliferation orproduction of (mainly) heme proteins. Mutations of theSLC11A3 gene that decrease the iron-transport function ofFerroportin result in a decrease of iron absorption and anincrease of macrophage iron retention and a high serum ferri-tin level of the patients, a clinical entity now recognized asautosomal dominant hereditary hemochromatosis type 4(40,49). The fact that these patients are not severely anemiccan be related to the possibility of erythroblasts to bind andincorporate ferritin, which may form an important alternativeiron source for hemoglobin production (50). Experiments onFPN1’s role in macrophage iron metabolism, studying the

216 Marx

effect of iron status and erythrophagocytosis on FPN1 expres-sion in J774 macrophages, revealed that FPN1 mRNA levelswere increasing 8-fold after 4 hr and returning to basal levelsby 16 hr after erythrophagocytosis due to increase of erythro-cyte-derived LIP (39). Comparative Northern analyses ofiron-related genes after erythrophagocytosis revealed a16-fold increase in FPN1 levels after 3 hr, a 10-fold increasein HO-1 after 3hr, a 2-fold increase in NRAMP1 levels after6 hr, but no change in DMT1 levels.

The driving force and regulation of iron release frommacrophages remain unclear (51). In normal subjects moder-ate iron stores are seen as ferritin and hemosiderin, both inthe MPS and hepatocytes. Storage iron is rapidly availableif the bone marrow needs increased supplies for erythropoi-esis. This is most striking during phlebotomy treatment inhemochromatosis patients. No anemia develops during inten-sive phlebotomy treatment of 500-mL blood per week causinga huge increase of iron utilization while transferrin ironsaturation remains high. The saturation of plasma transfer-rin in iron overload is apparently not a major modulator ofiron release from the MPS, which is consistent with someexperimental studies (21,28). Other studies showed a clearenhancement of iron release from macrophages related toapotransferrin availability (52). Some speculation on regula-tion of macrophage Fe(II) release, however, is possible. Asiron is transported across the plasma membrane of macro-phages as highly soluble Fe2þ, and iron is bound to transfer-rin as Fe(III), driving forces can be the availability of Fe(II) inthe intracellular iron pool and the rate of Fe(II) oxidation inthe plasma, changing the concentration gradient of Fe(II)across the plasma membrane. Autoxidation of Fe(II) is ratherslow, allowing potentially harmful interaction of iron withtoxic oxygen species. For this reason, nature has designedceruloplasmin (Cp), a plasma multicopper ferroxidase. Oxida-tion of Fe(II) and binding to transferrin occur easily in thepresence of Cp and molecular oxygen at acid pH, which isnot the situation in plasma. Interestingly introduction of ahypoxic atmosphere results in marked Cp-stimulated bindingof iron to apotransferrin at physiological pH, and a remark-


able increase of iron release from macrophages (53). In thesestudies, iron could only be extracted from cells if the intracel-lular labile iron pool (LIP) was sufficiently filled. If the LIP isdepleted, which may easily occur during in vitro experiments,modulation of plasma iron saturation may fail to exhibit aneffect on iron release from macrophages. Therefore, acontinuous supply of iron towards the LIP seems to be essentialto create a positive intracellular Fe(II) gradient. In this context,it should be realized that intracellular ferritin is not a long-term storage protein. There appears to be a continuous synth-esis and proteolysis of intracellular ferritin. Normally half-lifeof intracellular ferritin molecules is less than 24 hr, with largefluctuations. Ferritin life span decreases in iron deficientand increases in iron-loaded cells (54,55). The dynamics ofthe LIP in macrophages will be important for ERP-regulatedproduction of ferritin, proteolysis of ferritin, and iron exportvia ferroportin, the only physiologically relevant sourcesof iron influx being erythrophagocytosis and ferritinbreakdown.

A newly discovered polypeptide, which was identifiedduring a genomic search for genes that are upregulated inhepatocytes under the condition of iron excess, is hepcidin(56). The molecule is produced as an 84 amino acid protein,and is present in plasma as a 25 amino acid peptide. It wasfurther recognized that hepcidin behaves as a hormone thatregulates iron mobilization from cells based on hepatic ironstores. In addition, it was observed that mice, which haveno hepcidin develop severe iron overload (57). The hormoneis able to regulate iron transport from gut mucosal cells andmacrophages to the plasma in conjunction with HFE.Hepcidin probably plays a key role as a genetic modifier con-tributing to the phenotype of patients with hereditary hemo-chromatosis type 1 (HH), which is caused by a mutation in theHFE gene (58). Iron release from gut mucosal cells andmacrophages is increased in HH and decreased duringinflammation. The finding, in animal infection models, ofincreased plasma concentrations of hepcidin suggests thathepcidin may play a role in macrophage iron accumulationduring inflammation (59).

218 Marx

IRON RELEASE FROM MACROPHAGES ININFLAMMATION

A typical feature of anemia in subjects with acute and chronicinflammation is that a low plasma iron and transferrinconcentration, and an increased plasma ferritin level, isassociated with slightly increased macrophage iron stores(60). Plasma iron is low due to a decreased influx of iron fromiron donor cells, in particular macrophages of the MPS (25,61)and gut mucosal cells, the latter being responsible for adecreased iron absorption in for instance rheumatoid arthritisand dialysis patients (15,62,63). A key feature of infection andinflammation is production of proinflammatory cytokines. Invitro experiments with interferon-g (IFN-g) and lipopolysac-charide (LPS)-stimulated monocytic cell lines showed downre-gulation of transferrin receptor (64). This will have littlequantitative impact on MPS iron trafficking as this is fullydependent on erythrophagocytosis. More important is theadditional finding that expression of DMT-1, enabling cellularuptake of NTBI, is upregulated, and ferroportin, enablingiron release by monocytic cells and gut mucosa, is downregu-lated. Preincubation with the anti-inflammatory cytokine IL-10 could counteract these changes. Depression of iron releaseis a consistent finding in inflammatory macrophages.

Macrophage production of ferritin is increased in inflam-mation, mainly due to increased proinflammatory cytokineinduced transcription of heavy molecular weight ferritinsubunits (H-ferritin) (65). As a result more iron will be seque-strated in MPS ferritin (66). Most important for iron homeos-tasis are the iron regulatory proteins (IRP)-1 and 2,cytoplasmic proteins, which, on binding to RNA motifs, callediron responsive elements (IRE), control translation of ferritin,transferrin receptor, and many more iron-related proteins.Treatment of macrophages with proinflammatory cytokinesresults in a nitric oxide-dependent small increase of the activ-ity of IRP1 but a strong reduction of IRP2 activity accompa-nied by increased ferritin synthesis during inflammation(67). NO(�), a redox species of nitric oxide that interacts pri-marily with iron, can activate IRP1 RNA-binding activity


resulting in an increase in TfR mRNA levels and a decrease inferritin synthesis. Treatment of murine macrophages withNO(þ) (nitrosonium ion, which causes S-nitrosylation of thiolgroups), however, resulted in a rapid decrease in RNA bind-ing of IRP2, followed by IRP2 degradation, and these changeswere associated with a decrease in TfR mRNA levels and adramatic increase in ferritin synthesis. These results suggestthat NO(þ)-mediated degradation of IRP2 plays a major rolein macrophage iron metabolism during inflammation (68).

The impact of typical changes in iron metabolism on ane-mia is probably of limited importance, and certainly lessimportant than decreased production of erythrocytes, due toa blunted response to erythropoietin (60). Treatment withRh-erythropoietin is capable of curing anemia of inflamma-tion, which is associated with an improvement of quality oflife (16). Iron treatment of patients with inflammation mustbe avoided due to its growth-promoting effect towardsmicro-organisms and tumour cells and because of its capacityto inhibit T-cell-mediated immune effector pathways (60).Iron as an adjuvant to Rh-erythropoietin therapy, however,is acceptable as it answers the increased demand for iron inparallel with increased hemoglobin production.

The reason why most animal species develop similarmodifications of iron metabolism in infection and inflamma-tion is probably to withhold an essential element for bacterialgrowth. Apparently this was evolutionarily important forsurvival of the species. This also holds true for humans. Theiron-withholding cascade may, however, be of less benefit inthe many immune-related diseases that are not a direct effectof microbial infection (69).

REFERENCES

1. Mohandas N, Groner W. Cell membrane and volume changesduring red cell development and aging. Ann NY Acad Sci1989; 554:217–224.

2. Bosch FH, Werre JM, Roerdinkholder-Stoelwinder B, Huls TH,Willekens FL, Halie MR. Characteristics of red blood cell popu-

220 Marx

lations fractionated with a combination of counterflow centrifu-gation and Percoll separation. Blood 1992; 79:254–260.

3. Willekens FL, Bosch FH, Roerdinkholder-Stoelwinder B,Groenen-Dopp YA, Werre JM. Quantification of loss of haemo-globin components from the circulating red blood cell in vivo.Eur J Haematol 1997; 58:246–250.

4. Dumaswala UJ, Greenwalt TJ. Human erythrocytes shedexocytic vesicles in vivo. Transfusion 1984; 24:490–492.

5. Willekens FL, Roerdinkholder-Stoelwinder B, Groenen-DoppYA, Bos HJ, Bosman GJ, van den Bos AG, Verkleij AJ,Werre JM. Hemoglobin loss from erythrocytes in vivo resultsfrom spleen-facilitated vesiculation. Blood 2003; 101:747–751.

6. Beutler E. Red cell enzyme deficiencies as non-disease. BiomedBiochim Acta 1983; 42:S234–S241.

7. Rettig MP, Low PS, Gimm JA, Mohandas N, Wang J,Christian JA. Evaluation of biochemical changes during invivo erythrocyte senescence in the dog. Blood 1999; 93:376–784.

8. Boas FE, Forman L, Beutler E. Phosphatidylserine exposureand red cell viability in red cell aging and in hemolytic anemia.Proc Natl Acad Sci USA 1998; 95:3077–3081.

9. Recommended method for radioisotope red-cell survivalstudies. International Committee for Standardization inHaematology. Br J Haematol 1980; 45:659–666.

10. Valeri CR, MacGregor H, Giorgio A, Srey R, Ragno G. Compar-ison of radioisotope methods and a nonradioisotope method tomeasure the RBC volume and RBC survival in the baboon.Transfusion 2003; 43:1366–1373.

11. Marx JJM, Verzijlbergen JF. Radionuclide techniques inhaematology. In: van Rijk PP, ed. Nuclear Techniques in Diag-nostic Medicine. Dordrecht: Martinus Nijhoff Publishers,1986:459–517.

12. Mock DM, Lankford GL, Widness JA, Burmeister LF, Kahn D,Strauss RG. Measurement of red cell survival using biotin-labeled red cells: validation against 51Cr-labeled red cells.Transfusion 1999; 39:156–162.


13. Furne JK, Springfield JR, Ho SB, Levitt MD. Simplification ofthe end-alveolar carbon monoxide technique to assess erythro-cyte survival. J Lab Clin Med 2003; 142:52–57.

14. Lang KS, Roll B, Myssina S, Schittenhelm M, Scheel-WalterHG, Kanz L, Fritz J, Lang F, Huber SM, Wieder T. Enhancederythrocyte apoptosis in sickle cell anemia, thalassemia andglucose-6-phosphate dehydrogenase deficiency. Cell PhysiolBiochem 2002; 12:365–372.

15. Weber J, Werre JM, Julius HW, Marx JJ. Decreased ironabsorption in patients with active rheumatoid arthritis, withand without iron deficiency. Ann Rheum Dis 1988; 47:404–409.

16. Spivak JL. Iron and the anemia of chronic disease. Oncology(Huntingt) 2002; 16:25–33.

17. Dinant HJ, de Maat CE. Erythropoiesis and mean red-cell life-span in normal subjects and in patients with the anaemia ofactive rheumatoid arthritis. Br J Haematol 1978; 39:437–444.

18. van Iperen CE, van De WA, de Bruin M, Marx JJ. Total hipreplacement surgery does not influence RBC survival. Trans-fusion 2000; 40:1235–1238.

19. Moura E, Verheul AF, Marx JJ. A functional defect in heredi-tary haemochromatosis monocytes and monocyte-derivedmacrophages. Eur J Clin Invest 1998; 28:164–173.

20. Moura E, Verheul AF, Marx JJ. 59Fe labelling of rabbit ery-throcytes as a continuous source for iron metabolism studies.Lab Anim 1998; 32:284–289.

21. Kondo H, Saito K, Grasso JP, Aisen P. Iron metabolism in theerythrophagocytosing Kupffer cell. Hepatology 1988; 8:32–38.

22. Moura E, Verheul AF, Marx JJ. Evaluation of the role of Fcgamma and complement receptors in the decreased phagocyto-sis of hereditary haemochromatosis patients. Scand J Immunol1997; 46:399–405.

23. Maines MD. The heme oxygenase system: a regulator ofsecond messenger gases. Annu Rev Pharmacol Toxicol 1997;37:517–554.

24. Gagnon E, Duclos S, Rondeau C, Chevet E, Cameron PH,Steele-Mortimer O, Paiement J, Bergeron JJ, Desjardins M.

222 Marx

Endoplasmic reticulum-mediated phagocytosis is a mechanismof entry into macrophages. Cell 2002; 110:119–131.

25. Fillet G, Beguin Y, Baldelli L. Model of reticuloendothelial ironmetabolism in humans: abnormal behavior in idiopathic hemo-chromatosis and in inflammation. Blood 1989; 74:844–851.

26. Maines MD, Panahian N. The heme oxygenase system and cel-lular defense mechanisms. Do HO-1 and HO-2 have differentfunctions? Adv Exp Med Biol 2001; 502:249–272.

27. Custer G, Balcerzak S, Rinehart J. Human macrophage hemo-globin-iron metabolism in vitro. Am J Hematol 1982; 13:23–36.

28. Saito K, Nishisato T, Grasso JA, Aisen P. Interaction of trans-ferrin with iron-loaded rat peritoneal macrophages. Br JHaematol 62:275–286.

29. Moura E, Noordermeer MA, Verhoeven N, Verheul AF, Marx JJ.Iron release from human monocytes after erythrophagocytosisin vitro: an investigation in normal subjects and hereditaryhemochromatosis patients. Blood 1998; 92:2511–2519.

30. Sibille JC, Kondo H, Aisen P. Interactions between isolated hepa-tocytes and Kupffer cells in iron metabolism: a possible role forferritin as an iron carrier protein. Hepatology 1988; 8:296–301.

31. Willis D, Moore AR, Frederick R, Willoughby DA. Heme oxyge-nase: a novel target for the modulation of the inflammatoryresponse. Nat Med 1996; 2:87–90.

32. Hershko C, Cook JD, Finch CA. Storage iron kinetics. II.The uptake of hemoglobin iron by hepatic parenchymal cells.J Lab Clin Med 1972; 80:624–634.

33. Hershko C, Cook JD, Finch CA. Storage iron kinetics. VI. Theeffect of inflammation on iron exchange in the rat. Br J Hae-matol 1974; 28:67–75.

34. Hershko C. The fate of circulating haemoglobin. Br J Haematol1975; 29:199–204.

35. Mackenzie B, Hediger MA. SLC11 family of H(þ)-coupled metal-ion transporters NRAMP1 and DMT1. Pflugers Arch 2004;447:571–579.

36. Biggs TE, Baker ST, Botham MS, Dhital A, Barton CH, PerryVH. Nramp1 modulates iron homoeostasis in vivo and in vitro:


evidence for a role in cellular iron release involving de-acidifi-cation of intracellular vesicles. Eur J Immunol 2001; 31:2060–2070.

37. Kuhn DE, Lafuse WP, Zwilling BS. Iron transport intomycobacterium avium-containing phagosomes from anNramp1(Gly169)-transfected RAW264.7 macrophage cell line.J Leukoc Biol 2001; 69:43–49.

38. Barton CH, Biggs TE, Baker ST, Bowen H, Atkinson PG.Nramp1: a link between intracellular iron transport andinnate resistance to intracellular pathogens. J Leukoc Biol1999; 66:757–762.

39. Knutson MD, Vafa MR, Haile DJ, Wessling-Resnick M. Ironloading and erythrophagocytosis increase ferroportin 1(FPN1) expression in J774 macrophages. Blood 2003;102:4191–4197.

40. Montosi G, Donovan A, Totaro A, Garuti C, Pignatti E,Cassanelli S, Trenor CC, Gasparini P, Andrews NC, Pie-trangelo A. Autosomal-dominant hemochromatosis is asso-ciated with a mutation in the ferroportin (SLC11A3) gene. JClin Invest 2001; 108:619–623.


42. Hershko C, Graham G, Bates GW, Rachmilewitz EA. Non-specific serum iron in thalassaemia: an abnormal serum ironfraction of potential toxicity. Br J Haematol 1978; 40:255–263.

43. de Valk B, Addicks MA, Gosriwatana I, Lu S, Hider RC, Marx JJ.Non-transferrin-bound iron is present in serum of hereditaryhaemochromatosis heterozygotes. Eur J Clin Invest 2000;30:248–251.

44. Wright TL, Brissot P, Ma WL, Weisiger RA. Characterizationof non-transferrin-bound iron clearance by rat liver. J BiolChem 1986; 261:10909–10914.

45. Gutierrez JA, Yu J, Wessling-Resnick M. Characterization andchromosomal mapping of the human gene for SFT, a stimula-

224 Marx

tor of Fe transport. Biochem Biophys Res Commun 1998;253:739–742.

46. Scheiber-Mojdehkar B, Sturm B, Plank L, Kryzer I,Goldenberg H. Influence of parenteral iron preparations onnon-transferrin bound iron uptake, the iron regulatory proteinand the expression of ferritin and the divalent metal transpor-ter DMT-1 in HepG2 human hepatoma cells. Biochem Phar-macol 2003; 65:1973–1978.


48. Donovan A, Brownlie A, Zhou Y, Shepard J, Pratt SJ,Moynihan J, Paw BH, Drejer A, Barut B, Zapata A, Law TC,Brugnara C, Lux SE, Pinkus GS, Pinkus JL, Kingsley PD,Palis J, Fleming MD, Andrews NC, Zon LI. Positional cloningof zebrafish ferroportin1 identifies a conserved vertebrate ironexporter [see comments]. Nature 2000; 403:776–781.

49. Cazzola M, Cremonesi L, Papaioannou M, Soriani N, Kioumi A,Charalambidou A, Paroni R, Romtsou K, Levi S, Ferrari M,Arosio P, Christakis J. Genetic hyperferritinaemia and reticu-loendothelial iron overload associated with a three base pairdeletion in the coding region of the ferroportin gene (SLC11A3).Br J Haematol 2002; 119:539–546.

50. Leimberg JM, Konijn AM, Fibach E. Developing humanerythroid cells grown in transferrin-free medium utilize ironoriginating from extracellular ferritin. Am J Hematol 2003;73:211–212.

51. Knutson M, Wessling-Resnick M. Iron metabolism in the reticu-loendothelial system. Crit Rev Biochem Mol Biol 2003; 38:61–88.

52. Rama R, Sanchez J, Octave JN. Iron mobilization fromcultured rat bone marrow macrophages. Biochim BiophysActa 1988; 968:51–58.

53. Sarkar J, Seshadri V, Tripoulas NA, Ketterer ME, Fox PL.Role of ceruloplasmin in macrophage iron efflux duringhypoxia. J Biol Chem 2003; 278:44018–44024.

54. Truty J, Malpe R, Linder MC. Iron prevents ferritin turnoverin hepatic cells. J Biol Chem 2001; 276:48775–48780.


55. Meyron-Holtz EG, Vaisman B, Cabantchik ZI, Fibach E,Rouault TA, Hershko C, Konijn AM. Regulation of intracellu-lar iron metabolism in human erythroid precursors by interna-lized extracellular ferritin. Blood 1999; 94:3205–3211.

56. Pigeon C, Ilyin G, Courselaud B, Leroyer P, Turlin B, Brissot P,Loreal O. A new mouse liver-specific gene, encoding a proteinhomologous to human antimicrobial peptide hepcidin, is overex-pressed during iron overload. J Biol Chem 2001; 276:7811–7819.


58. Nicolas G, Andrews NC, Kahn A, Vaulont S. Hepcidin, a can-didate modifier of the hemochromatosis phenotype in mice.Blood 2004; 103:2841–2843.

59. Shike H, Lauth X, Westerman ME, Ostland VE, Carlberg JM,Van Olst JC, Shimizu C, Bulet P, Burns JC. Bass hepcidin is anovel antimicrobial peptide induced by bacterial challenge.Eur J Biochem 2002; 269:2232–2237.


61. Konijn AM, Hershko C. Ferritin synthesis in inflammation. I.Pathogenesis of impaired iron release. Br J Haematol 1977;37:7–16.

62. Kooistra MP, Marx JJ. The absorption of iron is disturbed inrecombinant human erythropoietin-treated peritoneal dialysispatients. Nephrol Dial Transplant 1998; 13:2578–2582.

63. Kooistra MP, Niemantsverdriet EC, van Es A, Mol-BeermannNM, Struyvenberg A, Marx JJ. Iron absorption in erythropoie-tin-treated haemodialysis patients: effects of iron availability,inflammation and aluminium. Nephrol Dial Transplant 1998;13:82–88.

64. Ludwiczek S, Aigner E, Theurl I, Weiss G. Cytokine-mediatedregulation of iron transport in human monocytic cells. Blood2003; 101:4148–4154.

226 Marx


66. Fahmy M, Young SP. Modulation of iron metabolism in mono-cyte cell line U937 by inflammatory cytokines: changes intransferrin uptake, iron handling and ferritin mRNA. BiochemJ 1993; 296(pt 1):175–181.

67. Recalcati S, Taramelli D, Conte D, Cairo G. Nitric oxide-mediated induction of ferritin synthesis in J774 macrophagesby inflammatory cytokines: role of selective iron regulatoryprotein-2 downregulation. Blood 1998; 91:1059–1066.

68. Kim S, Ponka P. Role of nitric oxide in cellular iron metabo-lism. Biometals 2003; 16:125–135.

69. Marx JJ. Iron and infection: competition between host andmicrobes for a precious element. Best Pract Res Clin Haematol2002; 15:411–426.


8

New Regulator Molecules in Anemiaof Chronic Diseases

OLIVIER LOREAL and PIERRE BRISSOT

INSERM U-522 and Service des Maladies du Foie,University Hospital Pontchaillou,

Rennes, France

INTRODUCTION

Anemia is a frequent complication occurring in chronic dis-eases, including chronic inflammation and=or infection, neo-plastic states, and end-stage renal diseases. The bioclinicalpicture of this anemia associates, besides the symptomsrelated to the causing disease and the inflammatorysyndrome, a normocytic or slightly microcytic anemia withlow serum iron parameters, including serum iron and trans-ferrin saturation, all contrasting with an hyperferritinemia(for reviews, see Refs. 1–3).

229

For many years, the molecular mechanisms involvedin the development of anemia of chronic disease haveincluded not only disturbances in iron homeostasis, but alsoinhibition of erythroid progenitor proliferation and differen-tiation, blunted erythropoietin response, and a decrease of ery-throcyte life span (for reviews, see Refs. 1, 2, 4). In fact, duringanemia of chronic disease, despite these low serum iron para-meters, iron stores are normal or even high (1,5). Therefore,iron metabolism abnormalities participating in the occurrenceof anemia are related to an inadequate repartition and=orutilization of body iron stores reflecting an abnormal regula-tion of iron metabolism.

The discovery of hepcidin, a molecule directly involved inthe control of iron metabolism, provides new insights on theunderstanding of iron metabolism abnormalities occurringduring chronic diseases and paves the road for new therapeu-tic strategies.

HEPCIDIN: A KEY MOLECULE IN IRONMETABOLISM

Hepcidin is a polypeptide that was initially purified fromhuman plasma ultra filtrates (6) and urine (7) as 25- (themajor form), 22- or 20-amino acid polypeptide. This moleculeexhibits antibacterial properties demonstrated in vitro (6,7)and was, therefore, firstly described as a member of the anti-microbial peptide family and suspected to directly participatein innate antimicrobial defense (6,7). The human gene islocated on chromosome 19 and generates a 412 bp mRNA,which encodes an 84 AA proform in humans (7,8). The synthe-sized polypeptide proform is addressed to the secretory path-way due to the presence of a peptide signal sequence (6,7).Then, the 25 aa mature form is cleaved, thanks to a signalsequence cleavage site for furin and related mammaliansubtilisin=Kex2p-like propeptide convertases, which likelyprocess prohepcidin to the secreted forms (7,8).

Hepcidin is a cationic molecule that contains eightcysteines on 25 amino acids (6,7). Studies performed on

230 Loreal and Brissot

synthetic and refolded peptide, reported that all cysteine resi-dues are engaged in four disulfide bridges (7,9). One of thesebridges is uncommon since it involves two vicinal amino acids.The mature form exhibits a distorted beta-sheet shape with ahairpin loop. The beta-sheet structure, with two antiparallelstructures, is stabilized by disulfide bridges as well as hydro-gen bonds. The positively charged hydrophilic side chains ofthe molecule are likely involved in the disruption of bacterialmembranes and therefore may explain the antimicrobialactivity of the molecule (7,9). The roles of the hairpin andthe uncommon vicinal disulfide bridge remain to be clarified.

Hepcidin is mainly expressed in the liver (7,8) and moreprecisely in hepatocytes (8). Because in drosophila, the fatbody, a functional equivalent of the liver (10), is involved inthe production of most antimicrobial peptides, hepcidin couldrepresent a hepatic homologue of insect cysteine-rich antimi-crobial peptides (11). A potential role of hepcidin in theorganism defense was reinforced by the demonstration thathepcidin mRNA levels were inducible by lipopolysaccharides(LPS) in vivo in mice liver and in vitro in hepatocytes (8).Such strong induction of hepcidin mRNA was also found dur-ing bacterial challenge in bass (12). This was subsequentlyconfirmed in cultured human hepatocytes and the role ofinflammatory cytokine, e.g., IL6, in the induction of hepcidinmRNA expression level was demonstrated (13). In addition, apotential indirect mechanism involving macrophages overpro-ducing cytokines after LPS stimulation has been reported(13).

The relationship between hepcidin and iron metabolismwas firstly reported in mouse. A suppressive subtractivehybridization performed between carbonyl iron overloadedand noniron overloaded mouse livers allowed Pigeon et al.(8) to identify an iron inducible mRNA which correspondedto an Expressed Sequence Tag. Subsequently, when the pep-tide sequence was submitted to Swissprot database by TomasGanz group, it was recognized as encoding the proform of themouse hepcidin (83 aa). In fact, in mouse, there are two differ-ent genes, likely resulting from ancestral duplication, encod-ing two different proteins, which are only 68% homologous

New Regulator Molecules in ACD 231

(8,14,15). In mice as well as in humans, the two genes arevery close to the upstream stimulatory factor-2 (USF2) gene(8,14,15). The relationship between iron metabolism was thenretrieved in parenteral iron overload model and beta2 micro-globulin knock-out model (8). Beta2 microglobulin is a proteinthat interacts with HFE molecule and transferrin receptor1 (16,17). Such an interaction likely explains why beta2microglobulin knock-out animals develop an iron overloadphenotype, which resembles to that found in HFE knock-outanimals (18). In iron overloaded beta2 microglobulin knock-out animals, there was an increase of hepatic hepcidinmRNA level. Conversely, low iron intake in these animalsled to a parallel decrease of liver iron stores and hepcidinmRNA level (8).

Nicolas et al. (14), who worked on animals knocked-outfor the USF2 gene, found unexpectedly that these animalsdeveloped an iron overload close to genetic hemochromatosisphenotype with heavy hepatic and pancreatic iron depositioncontrasting with the absence of splenic iron content increase.By performing a suppressive subtractive hybridizationbetween the livers of USF2 knock-out and wild type animals,they found that the two hepcidin genes were not functional.Then, taking together their observations and those reportedby Pigeon et al. (8), they hypothesized that hepcidin couldbe a hormone involved in the maintenance of iron homeosta-sis by controlling digestive iron absorption and reticuloen-dothelial cell iron egress (14,19). A decrease of hepcidinexpression could favor both digestive iron absorption throughthe enterocytes and iron egress from reticuloendothelial cells,which could explain the iron overload phenotype found inUSF2 knock-out animals (14,19). The role of hepcidin in ironmetabolism was further confirmed in hepcidin 1 transgenicmouse, which exhibited a strong iron deficiency phenotype(20). Finally in humans, Roetto et al. (21) investigating youngpatients with severe juvenile hemochromatosis phenotypereported, for the first time, two homozygous mutations lead-ing to the disappearance of the normal mature hepcidin,thereby confirming the crucial role of the mature form ofhepcidin in iron metabolism.


HEPCIDIN: A KEY MOLECULE IN ANEMIA OFCHRONIC DISEASES

Anemia of chronic disease, as mentioned above, is partlyrelated to a decoupling between iron stores, which are classi-cally normal or high in this case, and iron utilization; 70% ofthe total iron body content being normally localized withinthe erythrocytes. This is related to the poor bioavailability ofplasma iron, which is reflected by the decrease of serum ironand transferrin saturation, transferrin iron being the mostbioavailable iron species for erythropoiesis (22,23). There aretwo major sources of bioavailable iron in plasma (for reviews,see Refs. 22, 23): the iron recycled from senescent erythrocytesafter erythrophagocytosis by reticuloendothelial cells, which isby far the main pool, and iron provided by digestive absorption(only 1–2mg of iron per day). The latter pool compensates theiron losses (digestive, urinary, and cutaneous losses that are‘‘incompressible’’). Therefore, pathological states interactingwith these two processes, and especially—due to its quantita-tive importance—with the iron release from macrophages,may limit the erythropoiesis level and favor the developmentof anemia.

The role of hepcidin overexpression in the alteration ofiron metabolism during chronic diseases has been evoked byNicolas et al. (14,19). Such a role is supported in mice bythe inducibility of hepatic hepcidin mRNA in situationsmimicking in vitro and in vivo the inflammation includingLPS (8) and turpentine exposition (24). This hypothesis hasbeen reinforced, as described above, by the report of a strongincrease of hepcidin mRNA level in human hepatocyte cul-tures directly exposed to LPS or to conditioned media frommacrophages treated with LPS or IL6 (13). It is noteworthythat IL6 treatment of the hepatocytes had a 25-fold highereffect than LPS. The use of conditioned media from macro-phages reproduced the result of direct hepatocyte IL6 treat-ment suggesting the major role of this cytokine in thecontrol of hepcidin mRNA expression (13). In addition, IL1and TNFa do not significantly modify the hepcidin mRNA inthese cultures suggesting that these proinflammatory


cytokines, which stimulate type I acute phase response pro-teins, are not involved in a direct regulation of hepcidinmRNA during the chronic inflammatory disease (13). There-fore, it has been suggested that hepcidin was a member ofthe type II acute phase response proteins.

Three types of observations in diseased patients supportthe relationship between inflammatory processes and hepci-din mRNA expression.

Firstly, a high level of hepcidin protein has been found inurine from patients with chronic inflammatory diseases (13).Secondly, in one patient with acute infection, there wasa strong increase of hepcidinuria. The level of hepcidinuriareturned to normal after resolution of the infectious pro-cess under appropriate antibiotic treatment (13). Thirdly,Weinstein et al. (25) reported that, in two patients with hepaticadenomas overexpressing hepcidin mRNA, the resection ofthese tumors led to the resolution of the chronic anemia, simi-lar to that found in inflammatory chronic disease, exhibited bythese patients.

All these data clearly demonstrate a strong relationshipbetween the expression of hepcidin, a key molecule in ironmetabolism, and inflammatory processes, thus, suggestingthat hepcidin is a crucial molecule in the development ofACD (Fig. 1). Therefore, regulators of its expression as wellas the molecules that interact with hepcidin are supposed tobe potential regulators of ACD development.

HEPCIDIN EXPRESSION REGULATORS

IL-6: The Main Cytokine Involved in HepcidinOver Expression

A serum increase of inflammatory cytokines has been recog-nized for many years during the inflammatory process andhas been involved in the development of ACD. From theexperimental data mentioned above, it appears clearly thatinflammatory cytokines, especially IL6, play a crucial role inthe regulation of hepcidin mRNA expression (13). In addition,despite the fact that cytokines such as IL1 and TNF do not


appear to act as regulators of hepcidin expression, they likelyplay a role during the inflammatory process at least by regu-lating the expression of other proteins involved in iron meta-bolism (for review, see Ref. 1), including ferritin (26). Inaddition, anti-inflammatory cytokines (5) are well-establishedmodifiers of iron metabolism. Molecular mechanisms involvedin the control of hepcidin expression during inflammatoryprocess may involve transcriptional factors such as NFkB

Figure 1 Mechanisms of regulation and action of hepcidin (sche-matic representation; for details, please see the text.) Note that,during ACD, the original signal is likely generated by the inflamma-tion which promotes hepcidin expression. This hepcidin productionincrease may in turn, through its effect on iron metabolism, contri-butes to the maintenance of hepcidin increase. The respective rolesof iron stores (hepatocytic and macrophagic) and biochemical serumiron form on hepcidin expression remain to be determined. Theeffects, during ACD, of (i) iron metabolism gene mutations and poly-morphisms, including HFE and other genes involved in iron absorp-tion and macrophages iron egress genes, and (ii) modulations ofhepatic function remain to be characterized.


and STAT. Indeed, bioinformatics analysis of the 50-flankingregion of the hepcidin gene revealed the presence of putativebinding sites for these two transcriptional factors (8,12). Thebiological role of these potential binding sites remains to bedemonstrated. In addition, it cannot be ruled out thatcytokine-related post-transcriptional events are involved inthe control of hepcidin polypeptide expression.

Erythropoietin: A Downregulator ofHepcidin Expression

Besides cytokines, other plasmatic circulating proteins prob-ably play a major role in the control of hepcidin expression.Thus, erythropoietin, which controls many genes involved inerythropoiesis (for reviews, see Refs. 2, 4, 27), has beenreported to strongly regulate hepcidin mRNA expression.Indeed, it has been shown that hypoxia, secondary to themaintenance of animals in hypobaric condition or to theinduction of hemolytic anemia, reduced strongly the level ofhepcidin mRNA expression (28). This was suggestive thateither hypoxia and=or erythropoietin overproduction, whichare induced under these two conditions, were able to reducehepcidin mRNA (28). It should be noted that the presence ofhypoxia inducible binding site, which is involved in theexpression control of number of genes involved in iron meta-bolism, erythropoiesis, and angiogenesis (29), has not yetbeen reported in the flanking region of hepcidin gene, and adirect role of hypoxia remains to be demonstrated. However,injection of human recombinant erythropoietin to mice hasbeen reported to strongly decrease hepatic hepcidin mRNAexpression (24). This relation between hepcidin, a polypeptidecontrolling the bioavailability of iron and erythropoietin, thehormone that controls erythropoiesis, is likely essential indiseased patients.

Erythropoietin and Cytokines: A Duo in HepcidinRegulation During Chronic Diseases

In patients with the inflammatory process, a slight increase ofplasmatic erythropoietin level has been reported (2). Such an


increase could theoretically promote erythropoiesis directlyand indirectly, through its favoring role in both increasingiron release from macrophages and iron absorption by enter-ocytes. However, despite these two modes of erythropoietinactions, when ACD occurs, the level of erythropoiesis is notsufficient (1) and stored iron remains high. This is at leastrelated to the fact that the degree of erythropoietin levelincrease is frequently not correlated with the intensity of ane-mia (30). In addition, besides erythropoietin, other moleculesmay play a role in controlling both erythropoiesis and ironbioavailability during inflammatory process. Thus, the inhibi-tory effect of cytokines, such as IL1, TNFa, IFNg and IFNa onthe CFU-E colony formation has been previously reported(31–34) and the direct effect of inflammatory cytokines onthe expression of proteins involved in iron metabolism hasbeen underlined (for review, see Ref. 1).

The role of such molecules is well illustrated in patientswith chronic renal disease. In these cases, anemia frequentlyoccurs, and the molecular mechanisms involved in its develop-ment include, as a major factor, not only a strong decrease inerythropoietin synthesis related to the renal disease but alsothe hemodialysis-related iron losses and an increase of seruminflammatory cytokines during this treatment (27,35). Humanrecombinant erythropoietin administration is a revolution inthe treatment of these patients leading, in association withiron supplementation, to a correction of the anemia (36,37).The results obtained experimentally suggest that the efficacyof the erythropoietin treatment during chronic renal diseasecould be partly related to its effects on iron metabolismthrough the downregulation of hepcidin expression, whichcould favor iron bioavailability for stimulated erythropoiesis.However, some patients are resistant to a well-conducted ery-thropoietin and iron supplementation and it has been demon-strated that the coexistence of an intense inflammatorysyndrome is frequent in those cases (38–40). This suggeststhat erythropoietin treatment, including its effect on hepcidinand iron metabolism, may be counteracted by a negative effectof inflammatory cytokines inducing a reduction of the prolif-eration and differentiation of erythroid progenitors and=or a


major upregulation of hepcidin, which orientates the iron ontothe iron storage forms.

Hepatocyte Function and Hepcidin Expression

Regulation of hepcidin mRNA may occur during other physio-logical processes or events. Thus, in vivo, the hepatocyte phe-notype regulates hepcidin expression, which is maximum inthe adult liver in contrast with lower levels during intrauterinedevelopment (7,8,15,20). The role of the hepatocyte phenotypehas been reinforced by in vitro studies. Thus, hepatic cell linesexpressed hepcidin mRNA at low levels (8,41,28). In addition,normal isolated hepatocytes maintained a hepcidin expressiononly for a few days when they expressed differentiated func-tions (8,13), and this expression requires the addition of serumin the culture medium (8,13), suggesting that the presence ofsoluble factors is critical. The role of specific transcriptionalfactors, such as C=EBP alpha and hepatocyte nuclear factor 4has been demonstrated both in vitro by transfection experi-ments and gel shift assay, and in vivo in mouse (41). Thus,C=EBP alpha appears to enhance hepcidin expression both invitro and in vivo. Three putative binding sites have been iden-tified in the hepcidin promoter, only one being functional. ForHNF4, contradictory results have been obtained between invitro and in vivo experiments (41), whereas in vitro transfec-tions experiments suggest that HNF4 favor hepcidin mRNAexpression, HNF4 knock-out mice overexpressed the hepcidinmRNA.

Iron Status as a Regulator of Hepcidin Expression

The knowledge of the relationship between iron metabolismand hepcidin expression has been further extended. It hasbeen firstly demonstrated that hepatic iron store modulationwas not required to produce a modulation in hepcidin mRNAexpression (25,42). Thus, in rats, poor iron diet content wasable to decrease the hepcidin mRNA within 3 days. Duringthis delay, neither the hepatic nor the splenic iron stores weremodified (42). However, in parallel, transferrin saturationwas slightly decreased, suggesting a role of this parameter


in hepcidin mRNA expression control (42). This resultcontrasts with those obtained in vitro in mouse and humanhepatocyte culture demonstrating the absence of hepcidinexpression increase or even a decrease when iron was addedto the culture medium as nontransferrin-bound iron or irontransferrin (8,13,43). It has been hypothesized by Gehrkeet al. (43) that nontransferrin-bound iron could regulate nega-tively the level of hepcidin expression. Secondly, it has beenreported that transferrinemic anemic mice, expressed hepci-din mRNA at lower levels than controls despite strong hepaticiron overload (25). This suggests that there is a signal (or sig-nals) that overcomes the iron overload effect (25). In thesestrongly anemic mice, this signal could be an increase ofserum erythropoietin leading to the hepcidin expression drop,as described above. In addition, if nontransferrin-bound ironplays a role in the regulation of hepcidin expression, thedecrease of iron transferrin bioavailability due to the absenceof endogen transferrin production in these mice could alsotake place, by facilitating the nontransferrin-bound ironappearance in the modulation of hepcidin expression.

HFE Molecule as a Regulator of HepcidinExpression

A relationship between HFE genotype and hepcidin expres-sion has been established. Firstly, Ahmad et al. (44) reportedthat young HFE knock-out animals expressed hepcidin mRNAat lower levels than their controls despite the presence of hepa-tic iron excess, suggesting a relationship between HFE geno-type and hepcidin expression. This result was confirmed inother animal studies (45,46), and it has been demonstratedthat the iron overload phenotype, which develops in HFEknock-out mice, was corrected by a constitutive expression ofhepcidin 1 gene (46). Finally in humans, patients with HFE-1 genetic hemochromatosis (e.g., related to the C282Y muta-tion) and exhibiting an iron overload phenotype were demon-strated to have abnormally low levels (42) or no increase (45)in hepatic hepcidin mRNA despite the development of ironoverload.


All these results strongly suggest that HFE may exert itseffect in iron metabolism by regulating the level of hepcidinmolecule expression (42,44–46). Even if the link between thetwo molecules remains to be clarified, it appears that HFE gen-otype could modulate, by its effect on hepcidin expression, thedevelopment of the anemia in chronic diseases. In this regard,it is of usual clinical observation that genetic hemochromatosispatients with specific inflammatory arthritis do not exhibit adecrease of serum iron and transferrin saturation levels. Inaddition, in hemochromatosis patients, the increase of ironcontent in Kupffer cells, which are hepatic resident macro-phages, occurs only at the end-stage of the iron overloaddisease (47). Therefore, the place of the polymorphisms, as wellas the mutations of the HFEmolecule (48,49) in the expressionof the ACD, remains to be determined. The same holds true forits molecular partners including the beta2 microglobulin, andtransferrin receptor 1 molecule (16,17,48,50,51).

MOLECULAR FUNCTION OF HEPCIDIN

The importance of hepcidin in iron metabolism clearly beingdemonstrated, its molecular function remains to be fully char-acterized. Each of the proteins interacting with—or regulatedby—hepcidin could be a regulator of the ACD.

Does Hepcidin Bind Iron?

Some proteins involved in iron metabolism and secreted inplasma, like hepcidin, have been described to exert antimicro-bial properties. Thus, transferrin was firstly isolated by itsantimicrobial effect (for reviews see Ref. 52). Lactoferrinwas also demonstrated to have antibacterial effects (53,54).A direct chemical interaction between hepcidin and iron couldbe evoked. However, from the molecular structure and experi-mental studies, including circular dichroism (7), to date thereis no argument for such a function. In addition, the antimicro-bial properties of the molecule have been addressed to itscharges.


Hepcidin: The Iron Hormone?

It has been hypothesized that hepcidin is a hormone secretedby the liver and controls the iron metabolism by regulatingat least the iron egress from enterocytes and macrophages(14,19). Such a hypothesis necessarily implicates that hepcidinsecreted by hepatocytes interacts with specific partners—receptors—localized on the cytoplasmic membrane. In addi-tion, the interaction between hepcidin and cellular targetscould lead to: (i) a local biological effect that limits iron egressfrom the cell, or (ii) an induction of transduction signal thatregulates expression or function of other molecules involvedin iron transport and=or iron egress pathways in enterocytesand macrophages, respectively. To date, the molecular targetof hepcidin remains to be characterized. However, we canhypothesize that hepcidin may at least in part exert its effectthrough regulation of expression and=or activity of other mole-cules involved in iron metabolism.

Hepcidin and Proteins of Iron Metabolism

All the molecules involved in iron metabolism could be a finaltarget for hepcidin and therefore represent potential regula-tors of the ACD. Inside this group of molecules, which is nowincreasing considerably, HFE, DMT1, Dcytb, ferroportin,and ceruloplasmin could play a predominant role. In addition,the determination of genetic polymorphisms in iron metabo-lism genes (55,56) suggests strongly that modification ofgenetic sequences could modulate biological function. There-fore, all these genes could participate, independent of hepci-din, in the modulation of the ACD expression.

An interaction between hepcidin and molecular complexincluding beta2 microglobulin, HFE molecule, and transfer-rin receptor 1 has been evoked (14). Indeed, the three lattermolecules form a macromolecular complex expressed on thebasolateral membrane of cryptic enterocytes, placenta, andmacrophages (57). This complex is involved in iron uptakefrom the blood and could play, at the cryptic enterocyte level,the role of iron sensor that determines, at the level of apical


enterocytes, not only the degree of iron uptake through theirapical membrane but also the iron egress through their baso-lateral membrane (57). Due to the similar phenotype exhib-ited by animals that are either HFE (58) or hepcidin geneknock-out (14), physical interaction of hepcidin with this com-plex has been suspected (14). However, it has been nowdemonstrated that HFE knock-out or knock-in animals, andHFE C282Y iron overloaded patients present abnormallylow level of hepcidin mRNA, despite the iron overload(43–45,59). This suggests that the relationship betweenHFE complex and hepcidin may not involve a direct physicalinteraction.

Divalent metal transporter 1 (DMT1), also namedNramp2, has been identified by two different groups (60–62).Mutations in DMT1 gene have been reported to induce irondeficiency and microcytic anemia both in Belgrade rat(60,63,64) and mk=mk mice (64). Two different mRNAs, oneof them exhibiting an iron-responsive element, have been iden-tified (60,61,65). The product of this gene is involved in ironuptake (Fe2þ) at the apical membrane of the enterocytes(60,61), but also in intracellular ironmetabolism for numerouscells, including erythroid progenitors (66). Indeed, DMT1 isalso expressed on the endocytic vesicular membrane and playsa role in the iron transport from the endocytic vesicle to the cellcytoplasm (67,68). Mutations of DMT1 induce iron deficiencyby decreasing both digestive iron absorption and iron incor-poration in hemoglobin through a decrease of cytoplasmic ironin the erythroid progenitors. Therefore, DMT1 mutations orpolymorphism could play a role in the intensity of anemia dur-ing chronic diseases. In addition, from experimental data(69,70) and clinical reports (71,72), it has been suggested thatDMT1 mRNA expression (especially the mRNA exhibitingthe IRE sequence) could be modulated by HFE knock-outor mutation. This modulation of DMT1 expression by HFEgenotype could be related to a direct effect of HFE genotypeon DMT1 expression but also to an indirect effect throughmodulation of hepcidin expression in this case. However,conversely, other authors have reported the absence of DMT1expression in case of an abnormal HFE genotype (59,66). In


addition, new forms of DMT1 mRNA, resulting from alterna-tive splicing, have been identified (73). Therefore, the relation-ship between HFE genotype and DMT1 mRNAs expressionremains to be further characterized.

Dcytb, a protein with ferriductase activity, is expressedat the apical plasmic membrane of enterocytes and plays arole in the iron uptake from the digestive lumen (74). Dcytbreduces Fe3þ to Fe2þ, which, in turn, is taken in charge byDMT1 (74). Thus, modulation of its activity could modulatethe iron flux through the enterocyte. A regulation of itsmRNA expression has been found in animals (70,59) whenHFE genotype is altered.

Ferroportin, also named IREG1 and MTP1 (75–77), is amolecule involved in cellular iron egress. This protein isexpressed at very high level, not only in the basolateralmembrane of the apical enterocytes, but also on the plasmicmembrane of macrophages.

Mutations in the ferroportin gene have been recentlyreported and induce the development of a strong hyperferriti-nemia revealing an iron overload, which involves reticuloen-dothelial cells at very high levels and, at a lower degree,hepatocytes (78–83). This phenotype contrasts with a lowserum iron concentration and normal or low transferrinsaturation, and is very close to that observed during theanemia of chronic disease. Whether hepcidin interacts directlyor indirectly with ferroportin remains to be determined.

A complete activity of ferroportin in the iron egress fromthe cell requires the presence of a ferroxidase activity, whichensures the necessary iron oxidation to permit its uptake bycirculating transferrin. This enzymatic activity is exerted byhephaestin, an anchored protein localized in the basolateralplasma membrane of the enterocytes (84,85), and ceruloplas-min which mostly circulates in the plasma (86,87). Mutationsof each gene induce iron metabolism disturbances. Hephaes-tin mutation in mice leads to the development of the sex-linked anemia, which is associated with the disappearanceof the protein activity, which, in turn, downregulates diges-tive iron absorption (84). In humans, aceruloplasminemiarelated to mutations in the ceruloplasmin gene leads to the


development of a normocytic anemia with low serum iron, andtransferrin saturation contrasting with a strong hyperferriti-nemia related to an iron overload phenotype involving in theliver, parenchymal cells, and macrophages (88–97). Again,this phenotype is close to that observed during anemia ofchronic disease, with poor circulating iron bioavailabilityand increased iron content within the storage cells.

CONCLUSIONS

The mechanisms involved in the development of anemia ofchronic diseases are complex. The discovery of hepcidin andits role in iron metabolism gives a better understanding ofthe iron metabolism abnormalities implicated in the develop-ment of ACD. This molecule, which is regulated by numerousfactors including IL-6, erythropoietin, hepatocyte phenotype,and HFE genotype, modulates the quantity of plasmatic ironbioavailable for erythropoiesis. The molecular mechanismslinking the hepcidin molecule and its biological effects remainto be determined. However, today this molecule, which seemsto act as a hormone, represents likely a new way to developtherapeutic strategies in the context of anemia of chronicdiseases.

REFERENCES



3. Means RT Jr. Advances in the anemia of chronic disease. Int JHematol 1999; 70:7–12.

4. Goodnough LT, Skikne B, Brugnara C. Erythropoietin, iron,and erythropoiesis. Blood 2000; 96:823–833.

5. Weiss G, Bogdan C, Hentze MW. Pathways for the regulationof macrophage iron metabolism by the anti-inflammatory cyto-kines IL-4 and IL-13. J Immunol 1997; 158:420–425.


6. Krause A, Neitz S, Magert HJ, Schulz A, Forssmann WG,Schulz-Knappe P, Adermann K. LEAP-1, a novel highlydisulfide-bonded human peptide, exhibits antimicrobial activ-ity. FEBS Lett 2000; 480:147–150.


8. Pigeon C, Ilyin G, Courselaud B, Leroyer P, Turlin B, BrissotP, Loreal O. A new mouse liver-specific gene, encoding aprotein homologous to human antimicrobial peptide hepcidin,is overexpressed during iron overload. J Biol Chem 2001;276:7811–7819.

9. Hunter HN, Fulton DB, Ganz T, Vogel HJ. The solution struc-ture of human hepcidin, a peptide hormone with antimicrobialactivity that is involved in iron uptake and hereditary hemo-chromatosis. J Biol Chem 2002; 277:37597–37603.

10. Sondergaard L. Homology between the mammalian liver andthe Drosophila fat body. Trends Genet 1993; 9:193–200.

11. Ganz T. The role of hepcidin in iron sequestration duringinfections and in the pathogenesis of anemia of chronicdisease. Isr Med Assoc J 2002; 4:1043–1045.

12. Shike H, Lauth X, Westerman ME, Ostland VE, Carlberg JM,Van Olst JC, Shimizu C, Bulet P, Burns JC. Bass hepcidin is anovel antimicrobial peptide induced by bacterial challenge.Eur J Biochem 2002; 269:2232–2237.

13. NemethE,ValoreEV,TerritoM,SchillerG,LichtensteinA,GanzT. Hepcidin, a putative mediator of anemia of inflammation, is atype II acute-phase protein. Blood 2003; 101:2461–2463.

14. Nicolas G, BennounM, Devaux I, Beaumont C, Grandchamp B,Kahn A, Vaulont S. Lack of hepcidin gene expression andsevere tissue iron overload in upstream stimulatory factor 2(USF2) knockout mice. Proc Natl Acad Sci USA 2001; 10:10.

15. Ilyin GCB, Troadec MB, Pigeon C, Alizadeh M, Leroyer P,Brissot P, Loreal O. Comparative analysis of mouse HEPC1and HEPC2 genes encoding hepcidin, a regulator of ironmetabolism: evidence for different pattern of expression andco-inducibility during iron overload. FEBS Lett. In press.


16. Feder JN, Tsuchihashi Z, Irrinki A, Lee VK, Mapa FA,Morikang E, Prass CE, Starnes SM, Wolff RK, Parkkila S,Sly WS, Schatzman RC. The hemochromatosis founder muta-tion in HLA-H disrupts beta2-microglobulin interaction andcell surface expression. J Biol Chem 1997; 272:14025–14028.

17. Feder JN, Penny DM, Irrinki A, Lee VK, Lebron JA, Watson N,Tsuchihashi Z, Sigal E, Bjorkman PJ, Schatzman RC. Thehemochromatosis gene product complexes with the transferrinreceptor and lowers its affinity for ligand binding. Proc NatlAcad Sci USA 1998; 95:1472–1477.

18. De Sousa M, Reimao R, Lacerda R, Hugo P, Kaufmann SHE,Porto G. Iron overload in ß2-microglobulin-deficient mice.Immunology 1994; 39:105–111.

19. Fleming RE, Sly WS. Hepcidin: a putative iron-regulatoryhormone relevant to hereditary hemochromatosis and theanemia of chronic disease. Proc Natl Acad Sci USA 2001; 98:.

20. Nicolas G, Bennoun M, Porteu A, Mativet S, Beaumont C,Grandchamp B, Sirito M, Sawadogo M, Kahn A, Vaulont S.Severe iron deficiency anemia in transgenic mice expressingliver hepcidin. Proc Natl Acad Sci USA 2002; 99:4596–4601.


22. Andrews NC. Disorders of iron metabolism. N Engl J Med1999; 341:1986–1995.

23. Brissot P, Deugnier Y. Normal iron metabolism. In: McIntyreN, Benhamou J, Bircher J, Rizetto M, Rodes J, eds. OxfordTextbook of Clinical Hepatology. Oxford: Oxford UniversityPress, 1999:1379–1391.

24. Nicolas G, Viatte L, Bennoun M, Beaumont C, Kahn A,Vaulont S. Hepcidin, a new iron regulatory peptide. BloodCells Mol Dis 2002; 29:327–335.

25. Weinstein DA, Roy CN, Fleming MD, Loda MF, Wolfsdorf JI,Andrews NC. Inappropriate expression of hepcidin isassociated with iron refractory anemia: implications for theanemia of chronic disease. Blood 2002; 100:3776–3781.


26. Torti SV, Kwak EL, Miller SC, Miller LL, Ringold GM,Myambo KB, Young AP, Torti FM. The molecular cloningand characterization of murine ferritin heavy chain, a tumornecrosis factor-inducible gene. J Biol Chem 1988; 263:12638–12644.

27. Spivak JL, Pham T, Isaacs M, Hankins WD. Erythropoietin isboth amitogen and a survival factor. Blood 1991; 77:1228–1233.


29. Shih SC, Claffey KP. Hypoxia-mediated regulation of gene exp-ression in mammalian cells. Int J Exp Pathol 1998; 79:347–357.


31. Zoumbos NC, Djeu JY, Young NS. Interferon is the suppressorof hematopoiesis generated by stimulated lymphocytes invitro. J Immunol 1984; 133:769–774.

32. Means RT Jr, Krantz SB. Inhibition of human erythroidcolony-forming units by gamma interferon can be correctedby recombinant human erythropoietin. Blood 1991; 78:2564–2567.


34. Wang CQ, Udupa KB, Lipschitz DA. Interferon-gamma exertsits negative regulatory effect primarily on the earliest stages ofmurine erythroid progenitor cell development. J Cell Physiol1995; 162:134–138.

35. Koury MJ, Bondurant MC. Erythropoietin retards DNAbreakdown and prevents programmed death in erythroidprogenitor cells. Science 1990; 248:378–381.

36. Eschbach JW, Egrie JC, Downing MR, Browne JK, AdamsonJW. Correction of the anemia of end-stage renal disease withrecombinant human erythropoietin. Results of a combinedphase I and II clinical trial. N Engl J Med 1987; 316:73–78.


37. Fishbane S, Frei GL, Maesaka J. Reduction in recombinanthuman erythropoietin doses by the use of chronic intravenousiron supplementation. Am J Kidney Dis 1995; 26:41–46.

38. Nordstrom D, Lindroth Y, Marsal L, Hafstrom I, Henrich C,Rantapaa-Dahlqvist S, Engstrom-Laurent A, Fyhrquist F,Friman C. Availability of iron and degree of inflammationmodifies the response to recombinant human erythropoietinwhen treating anemia of chronic disease in patients with rheu-matoid arthritis. Rheumatol Int 1997; 17:67–73.

39. Greendyke RM, Sharma K, Gifford FR. Serum levels oferythropoietin and selected other cytokines in patients withanemia of chronic disease. Am J Clin Pathol 1994; 101:338–341.

40. Macdougall IC, Cooper AC. Erythropoietin resistance: the roleof inflammation and pro-inflammatory cytokines. Nephrol DialTransplant 2002; 17(suppl 11):39–43.

41. Courselaud B, Pigeon C, Inoue Y, Inoue J, Gonzalez FJ,Leroyer P, Gilot D, Boudjema K, Guguen-Guillouzo C, BrissotP, Loreal O, Ilyin G. C=EBPalpha regulates hepatic transcrip-tion of hepcidin, an antimicrobial peptide and regulator of ironmetabolism. Cross-talk between C=EBP pathway and ironmetabolism. J Biol Chem 2002; 277:41163–41170.

42. Frazer DM, Wilkins SJ, Becker EM, Vulpe CD, McKie AT,Trinder D, Anderson GJ. Hepcidin expression inversely corre-lates with the expression of duodenal iron transporters andiron absorption in rats. Gastroenterology 2002; 123:835–844.

43. Gehrke SG, Kulaksiz H, Herrmann T, Riedel HD, Bents K,Veltkamp C, Stremmel W. Expression of hepcidin in heredi-tary hemochromatosis: evidence for a regulation in responseto serum transferrin saturation and non-transferrin-boundiron. Blood 2003; 102:371–376.


45. Bridle KR, Frazer DM, Wilkins SJ, Dixon JL, Purdie DM,Crawford DH, Subramaniam VN, Powell LW, Anderson GJ,Ram GA. Disrupted hepcidin regulation in HFE-associated


haemochromatosis and the liver as a regulator of body ironhomoeostasis. Lancet 2003; 361:669–673.

46. Nicolas G, Viatte L, Lou DQ, BennounM, Beaumont C, Kahn A,Andrews NC, Vaulont S. Constitutive hepcidin expressionprevents iron overload in a mouse model of hemochromatosis.Nat Genet 2003; 34:97–101.

47. Deugnier YM, Loreal O, Turlin B, Guyader D, Jouanolle H,Moirand R, Jacquelinet C, Brissot P. Liver pathology in genetichemochromatosis: a review of 135 homozygous cases and theirbioclinical correlations. Gastroenterology 1992; 102:2050–2059.

48. Feder JN, Gnirke A, Thomas W, Tsuchihashi Z, Ruddy DA,Basava A, Dormishian F, Domingo R Jr, Ellis MC, Fullan A,Hinton LM, Jones NL, Kimmel BE, Kronmal GS, Lauer P, LeeVK, Loeb DB, Mapa FA, McClelland E, Meyer NC, Mintier GA,Moeller N, Moore T, Morikang E, Wolff RR, et al. A novel MHCclass I-like gene is mutated in patients with hereditary haemo-chromatosis [see comments]. Nat Genet 1996; 13:399–408.

49. Mura C, Raguenes O, Ferec C. HFE mutations analysis in 711hemochromatosis probands: evidence for S65C implication inmild form of hemochromatosis. Blood 1999; 93:2502–2505.

50. Lebron JA, Bennett MJ, Vaughn DE, Chirino AJ, Snow PM,Mintier GA, Feder JN, Bjorkman PJ. Crystal structure ofthe hemochromatosis protein HFE and characterization ofits interaction with transferrin receptor. Cell 1998; 93:111–123.

51. Gross CN, Irrinki A, Feder JN, Enns CA. Co-trafficking ofHFE, a nonclassical major histocompatibility complex class Iprotein, with the transferrin receptor implies a role in intracel-lular iron regulation. J Biol Chem 1998; 273:22068–22074.

52. Crichton R. Cellular iron uptake in mammals. In: RobertCrichton, ed. Inorganic Biochemistry of Iron Metabolism. 2nded. Chichester, England: JohnWiley and Sons, 2001:107–129.

53. Brock JH. Lactoferrin in human milk: its role in iron absorp-tion and protection against enteric infection in the newborninfant. Arch Dis Child 1980; 55:417–421.

54. Kuwata H, Yip TT, Tomita M, Hutchens TW. Direct evidenceof the generation in human stomach of an antimicrobial


peptide domain (lactoferricin) from ingested lactoferrin.Biochim Biophys Acta 1998; 1429:129–141.

55. Lee P, Gelbart T, West C, Halloran C, Beutler E. Seekingcandidate mutations that affect iron homeostasis. Blood CellsMol Dis 2002; 29:471–487.

56. Douabin-Gicquel V, Soriano N, Ferran H, Wojcik F, Palierne E,Tamim S, Jovelin T, McKie AT, Le Gall JY, David V, Mosser J.Identification of 96 single nucleotide polymorphisms in eightgenes involved in iron metabolism: efficiency of bioinformaticextraction compared with a systematic sequencing approach.Hum Genet 2001; 109:393–401.

57. Waheed A, Parkkila S, Saarnio J, Fleming RE, Zhou XY,Tomatsu S, Britton RS, Bacon BR, Sly WS. Association of HFEprotein with transferrin receptor in crypt enterocytes of humanduodenum. Proc Natl Acad Sci USA 1999; 96:1579–1584.

58. Zhou XY, Tomatsu S, Fleming RE, Parkkila S, Waheed A, JiangJ, Fei Y, Brunt EM, Ruddy DA, Prass CE, Schatzman RC,O’Neill R, Britton RS, Bacon BR, Sly WS. HFE gene knockoutproduces mouse model of hereditary hemochromatosis. ProcNatl Acad Sci USA 1998; 95:2492–2497.

59. Muckenthaler M, Roy CN, Custodio AO, Minana B, DeGraaf J,MontrossLK,AndrewsNC,HentzeMW.Regulatorydefects inliverand intestine implicate abnormal hepcidin and Cybrd1 expressionin mouse hemochromatosis. Nat Genet 2003; 34:102–107.

60. Fleming MD, Trenor CC III, Su MA, Foernzler D, Beier DR,Dietrich WF, Andrews NC. Microcytic anaemia mice have amutation in Nramp2, a candidate iron transporter gene. NatGenet 1997; 16:383–386.

61. Gunshin H, Mackenzie B, Berger UV, Gunshin Y, Romero MF,Boron WF, Nussberger S, Gollan JL, Hediger MA. Cloning andcharacterization of a mammalian proton-coupled metal-iontransporter. Nature 1997; 388:482–488.

62. Andrews NC. The iron transporter DMT1. Int J Biochem CellBiol 1999; 31:991–994.

63. Fleming MD, Andrews NC. Mammalian iron transport: anunexpected link between metal homeostasis and host defense.J Lab Clin Med 1998; 132:464–468.


64. Fleming MD, Romano MA, Su MA, Garrick LM, Garrick MD,Andrews NC. Nramp2 is mutated in the anemic Belgrade (b)rat: evidence of a role for Nramp2 in endosomal iron transport.Proc Natl Acad Sci USA 1998; 95:1148–1153.

65. Gunshin H, Allerson CR, Polycarpou-Schwarz M, Rofts A,Rogers JT, Kishi F, Hentze MW, Rouault TA, Andrews NC,Hediger MA. Iron-dependent regulation of the divalent metalion transporter. FEBS Lett 2001; 509:309–316.

66. Canonne-Hergaux F, Zhang AS, Ponka P, Gros P. Characteri-zation of the iron transporter DMT1 (NRAMP2=DCT1) in redblood cells of normal and anemic mk=mk mice. Blood 2001;98:3823–3830.

67. Gruenheid S, Canonne-Hergaux F, Gauthier S, Hackam DJ,Grinstein S, Gros P. The iron transport protein NRAMP2 isan integral membrane glycoprotein that colocalizes with trans-ferrin in recycling endosomes. J Exp Med 1999; 189:831–841.

68. Tabuchi M, Yoshimori T, Yamaguchi K, Yoshida T, Kishi F.Human NRAMP2=DMT1, which mediates iron transportacross endosomal membranes, is localized to late endosomesand lysosomes in HEp-2 cells. J Biol Chem 2000; 275:22220–22228.

69. FlemingRE,MigasMC, ZhouX, Jiang J, Britton RS, Brunt EM,Tomatsu S, Waheed A, Bacon BR, Sly WS. Mechanism ofincreased iron absorption in murine model of hereditary hemo-chromatosis: increased duodenal expression of the iron trans-porter DMT1. Proc Natl Acad Sci USA 1999; 96:3143–3148.

70. Dupic F, Fruchon S, Bensaid M, Borot N, Radosavljevic M,Loreal O, Brissot P, Gilfillan S, Bahram S, Coppin H, Roth MP.Inactivation of the hemochromatosis gene differentially regu-lates duodenal expression of iron-relatedmRNAs betweenmousestrains. Gastroenterology 2002; 122:745–751.

71. Zoller H, Pietrangelo A, Vogel W, Weiss G. Duodenal metal-transporter (DMT-1, NRAMP-2) expression in patients withhereditary haemochromatosis. Lancet 1999; 353:2120–2123.

72. Zoller H, KochRO, Theurl I, Obrist P, Pietrangelo A,Montosi G,Haile DJ, Vogel W, Weiss G. Expression of the duodenal irontransporters divalent-metal transporter 1 and ferroportin 1 in


iron deficiency and iron overload. Gastroenterology 2001;120:1412–1419.

73. Hubert N, Hentze MW. Previously uncharacterized isoforms ofdivalent metal transporter (DMT)-1: implications for regula-tion and cellular function. Proc Natl Acad Sci USA 2002;99:12345–12350.

74. McKie AT, Barrow D, Latunde-Dada GO, Rolfs A, Sager G,Mudaly E, Mudaly M, Richardson C, Barlow D, Bomford A,Peters TJ, Raja KB, Shirali S, Hediger MA, Farzaneh F,Simpson RJ. An iron-regulated ferric reductase associated withthe absorption of dietary iron. Science 2001; 291:1755–1759.

75. DonovanA,BrownlieA,ZhouY,ShepardJ,PrattSJ,MoynihanJ,Paw BH, Drejer A, Barut B, Zapata A, Law TC, Brugnara C,Lux SE, Pinkus GS, Pinkus JL, Kingsley PD, Palis J, FlemingMD, Andrews NC, Zon LI. Positional cloning of zebrafishferroportin1 identifies a conserved vertebrate iron exporter[see comments]. Nature 2000; 403:776–781.


77. Abboud S, Haile DJ. A novel mammalian iron-regulatedprotein involved in intracellular iron metabolism. J Biol Chem2000; 275:19906–19912.

78. Pietrangelo A, Montosi G, Totaro A, Garuti C, Conte D,Cassanelli S, Fraquelli M, Sardini C, Vasta F, Gasparini P.Hereditary hemochromatosis in adults without pathogenicmutations in the hemochromatosis gene [see comments].N Engl J Med 1999; 341:725–732.

79. Cazzola M, Cremonesi L, PapaioannouM, Soriani N, Kioumi A,Charalambidou A, Paroni R, Romtsou K, Levi S, Ferrari M,Arosio P, Christakis J. Genetic hyperferritinaemia and reticu-loendothelial iron overload associated with a three base pairdeletion in the coding region of the ferroportin gene (SLC11A3).Br J Haematol 2002; 119:539–546.

80. DevaliaV,CarterK,WalkerAP,PerkinsSJ,WorwoodM,MayA,Dooley JS. Autosomal dominant reticuloendothelial iron over-


load associated with a 3-base pair deletion in the ferroportin 1gene (SLC11A3). Blood 2002; 100:695–697.

81. Montosi G, Donovan A, Totaro A, Garuti C, Pignatti E,Cassanelli S, Trenor CC, Gasparini P, Andrews NC,Pietrangelo A. Autosomal-dominant hemochromatosis isassociated with a mutation in the ferroportin (SLC11A3) gene.J Clin Invest 2001; 108:619–623.

82. Njajou OT, Vaessen N, Joosse M, Berghuis B, van Dongen JW,Breuning MH, Snijders PJ, Rutten WP, Sandkuijl LA,Oostra BA, van Duijn CM, Heutink P. A mutation in SLC11A3is associated with autosomal dominant hemochromatosis. NatGenet 2001; 28:213–214.

83. Wallace DF, Pedersen P, Dixon JL, Stephenson P, Searle JW,Powell LW, Subramaniam VN. Novel mutation in ferroportin1is associated with autosomal dominant hemochromatosis. Blood2002; 100:692–694.


85. Anderson GJ, Frazer DM, McKie AT, Vulpe CD. The cerulo-plasmin homolog hephaestin and the control of intestinal ironabsorption. Blood Cells Mol Dis 2002; 29:367–375.

86. Osaki S, Johnson DEF. The possible significance of the ferrousoxidase activity of ceruloplasmin in normal human serum.J Biol Chem 1966; 241:2746–2751.

87. Harris ED. Cellular copper transport and metabolism. AnnuRev Nutr 2000; 20:291–310.

88. Yoshida K, Furihata K, Takeda S, Nakamura A, Yamamoto K,Morita H, Hiyamuta S, Ikeda S, Shimizu N, Yanagisawa N.A mutation in the ceruloplasmin gene is associated withsystemic hemosiderosis in humans. Nat Genet 1995; 9:267–272.

89. MoritaH, IkedaS,YamamotoK,MoritaS,YoshidaK,NomotoS,Kato M, Yanagisawa N. Hereditary ceruloplasmin deficiencywith hemosiderosis: a clinicopathological study of a Japanesefamily. Ann Neurol 1995; 37:646–656.


90. Daimon M, Kato T, Kawanami T, Tominaga M, Igarashi M,Yamatani K, Sasaki H. A nonsense mutation of the ceruloplas-min gene in hereditary ceruloplasmin deficiency with diabetesmellitus. Biochem Biophys Res Commun 1995; 217:89–95.

91. DaimonM,SusaS,OhizumiT,Moriai S,KawanamiT,HirataA,YamaguchiH,OhnumaH, IgarashiM,KatoT.Anovelmutationof the ceruloplasmin gene in a patient with heteroallelic cerulo-plasmin gene mutation (HypoCPGM) [in process citation].Tohoku J Exp Med 2000; 191:119–125.

92. Harris ZL, Takahashi Y, Miyajima H, Serizawa M,MacGillivray RT, Gitlin JD. Aceruloplasminemia: molecularcharacterization of this disorder of iron metabolism. Proc NatlAcad Sci USA 1995; 92:2539–2543.

93. Miyajima H, Takahashi Y, Shimizu H, Sakai N, Kamata T,Kaneko E. Late onset diabetes mellitus in patients withhereditary aceruloplasminemia [see comments]. Intern Med1996; 35:641–645.

94. Harris ZL, Durley AP, Man TK, Gitlin JD. Targeted gene dis-ruption reveals an essential role for ceruloplasmin in cellulariron efflux. Proc Natl Acad Sci USA 1999; 96:10812–10817.

95. Hellman NE, Schaefer M, Gehrke S, Stegen P, Hoffman WJ,Gitlin JD, Stremmel W. Hepatic iron overload in aceruloplas-minaemia. Gut 2000; 47:858–860.

96. LoganJI,HarveysonKB,WisdomGB,HughesAE,ArchboldGP.Hereditary caeruloplasmin deficiency, dementia and diabetesmellitus. QJM 1994; 87:663–670.

97. Loreal O, Turlin B, Pigeon C, Moisan A, Ropert M, Morice P,Gandon Y, Jouanolle AM, Verin M, Hider RC, Yoshida K,Brissot P. Aceruloplasminemia: new clinical, pathophysiologi-cal and therapeutic insights. J Hepatol 2002; 36:851–856.


9

Iron Withholding as a DefenseStrategy

EUGENE D. WEINBERG

Department of Biology and Program in MedicalSciences, Indiana University,Bloomington, Indiana, U.S.A.

Taken together, these observations suggest that mechan-isms to sequester iron have evolved as part of the host responseto infection. It is possible that ACD represents a side effect ofthis response when it persists over a period of months. (1)

INTRODUCTION

Excessive and misplaced iron in specific cells and tissues pro-motes infection, neoplasia, cardiomyopathy, arthropathy andan array of endocrine and neurodegenerative diseases. Verte-brate hosts maintain an iron withholding defense system

PART III: ACD RATIONALE: IS ANEMIA PARTOF THE BODY’SPHYSIOLOGICAL RESPONSE TO INFLAMMATION?

255

designed to: (a) prevent accumulation of nonprotein bound(free) iron in sensitive sites and (b) sequester the metal ininnocuous packages. Manifestations of iron withholding aremost evident in defense against diseases that are associatedwith known (or yet to be discovered) microbial cell invaders,with neoplastic cell invaders, and with other inflammatoryconditions.

HISTORICAL DEVELOPMENT OF THE CONCEPTOF IRON WITHHOLDING DEFENSE

Defense Components

An early recognition of a major component of iron withholdingis recorded in King Lear, a tragedy written by WilliamShakespeare in 1605. In Act III, Scene 7, the Earl of Glouce-ster is blinded by enemies of the king. Gloucester’s servantcries out: ‘‘I’ll fetch some flax and whites of eggs to apply tohis bleeding face.’’ That raw egg white contains an anti-infec-tive substance apparently was well known to Shakespeare andhis contemporaries even though microbial pathogens were notto be discovered for another two-and-a-half centuries.

Indeed, egg-laying creatures have been employing theanti-infective factor for countless millennia. Finally, in1944, the active ingredient was found to be a protein thatfunctions to bind ‘‘free’’ iron in egg white (2). The iron-bindingprotein prevents the white from becoming contaminated withmicrobial growth. Eggs are inherently subject to spoilage.Each egg is enclosed by a semipermeable membrane withina fragile porous shell. The pores are needed for exchange ofoxygen and carbon dioxide as the embryo develops, butmicrobes can invade through the pores.

In nature, eggs are deposited in milieu heavily ladenwith bacteria and fungi. The hens employ iron withholdingdefense to protect their priceless embryos from infectious rots.They place a milligram of iron in the yolk for use by therapidly developing chick. To prevent potential spoilagemicrobes that have migrated into the white from multiplyingor even from surviving, the hens place no iron in the white.

256 Weinberg

They include the iron-binding protein as 12% of the egg whitesolids. Moreover, since the protein functions best at alkalinepH values, the biochemically astute hens have adjusted thepH of the white to a value of 9.5.

With recognition in 1944 of the remarkable broad spec-trum antimicrobial power of the egg-white protein, it was pre-dicted that analogous compounds would be found to functionin animal tissues. Within two years, a similar protein wasdiscovered in human serum that prevents many, but notall, microbial invaders from growing in blood (3). Thiscompound comprises 3.5% of serum proteins and also ispresent in lymph and cerebrospinal fluid (CSF). It binds ironstrongly at pH 7.4.

The iron-binding proteins in egg white and in mamma-lian serum originally were termed ‘‘siderophilins.’’ However,shortly after their discovery, the serum protein was foundto have a second important function, the transport of ironamong host cells (4). The serum protein was renamed ‘‘trans-ferrin’’ (Tf) and the egg-white protein, ‘‘ovotransferrin.’’In healthy humans, 25–30% of the iron-binding capacityof serum Tf is employed in transfer of the metal. Seventyto seventy five percent functions to withhold iron frominvaders and to prevent serum, lymph, and CSF fromaccumulating ‘‘free’’ iron.

A third example of host stationing of iron-binding pro-teins at potential sites of invasion is that of lactoferrin (Lf).Initially observed in 1939 in human milk (in which itcomprises 20% of the total protein), Lf was not purified andcharacterized until 1960 (5). During the decade of the1960s, the protein was found in many other mammalian exo-crine secretions: tears, saliva, nasal exudates, bronchialmucus, gastrointestinal fluid, cervico-vaginal mucus, andseminal fluid. Additionally, Lf is a major component of thesecondary specific granules of circulating polymorphonuclearneutrophils (PMNs) (6).

Placement of Tf or Lf in nearly all of our various bodyfluids (with the notable exception of urine) constitutes a veryeffective first line of innate defense against invaders. Never-theless, it is imperative that an enhanced level of iron

Iron Withholding as a Defense Strategy 257

withholding be induced during episodes of microbial orneoplastic cell invasions. An early report on induced defensivestrategies was that of a clinical pathologist in 1932 (7). A largenumber of samples of sera from a great variety of patientswith different illnesses were assayed for iron. Specimens frompersons undergoing inflammatory defense processes werefound to be hypoferremic. As patients recovered, normallevels of iron would return.

Subsequent research by others demonstrated that theseprofound shifts in iron metabolism are independent of dietaryiron and are regulated by cytokines (1,6). During the hypofer-remic phase, additional ferritin is synthesized to provideincreased intracellular iron sequestering capacity (8). Inrecent decades, further components of iron withholdingdefense have been described; they will be discussed later inthe chapter and are listed in Table 1.

Invader Contravention

During the past half century, several thousand laboratorystudies and clinical observations have been published onmicrobial acquisition of host iron. Nearly all groups of poten-tially pathogenic bacteria, fungi, and protozoa require iron forreplication; an exception is Borrelia burgdorferi which usesmanganese in place of iron (9). Potentially pathogenic cellseither conveyed directly from the former to the new host orfrom nonhost environments generally are unable to carrywith them a sufficient quantity of stored iron for distributionto their offspring. That iron loading of hosts would provideencouragement to pathogens was observed as early as themiddle of the 19th century. Armand Trousseau, a prominentprofessor of clinical medicine and a practicing physician inParis between 1820 and 1870 was not aware of specific micro-bial pathogens but was quite familiar with the diagnosis ofspecific infectious diseases. In his lectures on ‘‘true and falsechlorosis’’ (10), he warned his students of the danger of feed-ing iron preparations to patients with quiescent tuberculosisand appeared certain that such a procedure could induce clin-ical episodes of the disease. His astute observations have

258 Weinberg

amply been confirmed by recent studies of animals andhumans whose burden of elevated iron via enteral or parent-eral routes has resulted in enhanced susceptibility to a greatvariety of infectious agents.

Not surprisingly, invaders vary greatly in their ability toobtain iron from iron-sufficient as well as from iron-loadedhosts. A strong association exists between enhanced ability toacquire iron from specific host tissues and elevated virulence.

Table 1 Iron Withholding Defense System

Constitutive components:Siderophilins

Transferrin in plasma, lymph, cerebrospinal fluidLactoferrin in secretions of lachrymal and mammary glandsand of respiratory, gastrointestinal and genital tracts

Ferritin within host cellsProcesses induced at time of invasion:1. Increased synthesis of hepcidin to suppress intestinal iron

absorption and release of recycled macrophage irona

2. Suppression of absorption of up to 80% of dietary ironb

3. Enhancement of TfR expression and iron uptake into macrophagesto result in up to 70% reduction in serum ironb

4. Increased synthesis of ferritin to sequester withheld ironb

5. Release of neutrophils from bone marrow into circulation and theninto site of infectionb

6. Release of apolactoferrin from neutrophil granules followed bybinding of iron in septic sites

7. Macrophage scavenging of ferrated lactoferrin in areas of sepsis8. Hepatic release of haptoglobin and hemopexin (to bind

extracellular hemoglobin and hemin, respectively)9. Synthesis of nitric oxide (from l-arginine) by macrophages to

disrupt iron metabolism of invaders and to downshiftexpression of TfRc

10. Suppression of growth of some microbial cells within macrophagesvia enhanced synthesis of Nramp1 by the host cellsc

11. Induction in B lyphocytes of synthesis of immunoglobulins toiron- repressible cell surface proteins that bind either heme,ferrated siderophilins or ferrated siderophoresd

aActivated by IL-6.bActivated by IL-1, IL-2, IL-6, IFNg, or TNFa.cActivated by IFNg or IL-1.dActivated by IL-4, IL-5, IL-10, or IL-13.


Indeed, the very first bacterial growth factor to have beendiscoveredwas a siderophore, i.e., a type of lowmolecularmasscompound that, for some bacteria and fungi, facilitates ironacquisition from Tf.

In 1912, a mycobacterial species that causes severeenteritis in calves was found to replicate only if provided witha compound synthesized by other groups of mycobacteria (11).The compound, called mycobactin, was shown later to be anhydroxamate siderophore. Other factors of virulence ofmicrobial and neoplastic cell invaders associated with hostiron procurement are described later.

COMPONENTS OF IRONWITHHOLDING DEFENSE

Constitutive Components

Sequestration of Extracellular Iron

As indicated earlier, mammals employ two siderophilins in acomplementary manner to continually purge body fluids of‘‘free’’ iron. Transferrin is responsible for iron binding inserum, lymph, and CSF whereas Lf functions in exocrinesecretions as shown in Table 2 (12,13). Transferrin is synthe-sized mainly in hepatocytes but also in other sites such asglial cells in the central nervous system and Sertoli cells inthe male reproductive tract. Lactoferrin is produced byepithelial cells of various exocrine glands and also by develop-ing neutrophils.

Transferrin and Lf are 78kDa glycoproteins with about44% homology. The molecules each consist of a single peptidechain folded in two lobes. Each lobe has a powerful metalbinding site; active residues are two tyrosines, a histidine,and an aspartate. For completion of the chelate rings, bothTf and Lf require bicarbonate ions. Upon binding of iron,the lobes undergo a conformational transition from an openinterdomain cleft to a closed structure (13). The affinityconstant of 1024 for the iron complex of Lf is about 240 timesstronger than that of Tf.

260 Weinberg

The nutritional function of Tf requires release of iron bythe protein in acidic endosomes. Thus Tf is unable to scavengeiron in infected and inflamed areas in which the pH value maybe lowered by catabolic acids released from metabolicallyactive invaders as well as from defense leukocytes. Unlike Tf,in Lf the interdomain hydrogen bonds do not protonate in mildacidity. Accordingly, Lf can effectively scavenge and retainiron at pH values above 4 (14). When Lf binds sufficient ironto become about 40% saturated, it is endocytosed by macro-phages. Scavenged iron accumulates also in hepatocytes.

Sequestration of Intracellular Iron

Iron not required for cell function or biosynthesis is seques-tered in intracellular ferritin. This protein, 450kDa, can con-tain up to 450 atoms of iron per molecule. Ferritin consists of24 subunits of heavy and light polypeptide chains; the mixvaries in different tissues. As ferritin molecules becomesatiated with iron, they are transformed into hemosiderin.

Table 2 Content of Apolactoferrin in Human Body Fluids

Fluids Concentration (mM) Underlying Condition

Colostrum 100 NormalMilk 20–60 NormalTears 25 NormalSeminal fluid 1.4 NormalVaginal fluid 0.1 Before menses

2.0 After mensesSaliva 0.05 Normal children

0.11 Normal adults0.25 Children: cystic fibrosis

Cerebrospinal fluid 0.00 Normal children0.01 Children: aseptic meningitis0.12 Children: bacterial meningitis

Synovial fluid 0.014 Noninflammatory arthritis0.338 Inflammatory arthritis

Serum 0.005 Normal2.5 Acute sepsis

(Modified from Ref. 13.)


The latter is an insoluble amalgam of degraded protein andferric hydroxide. In mature macrophages of noninflammedhosts, the ratio of hemosiderin to ferritin was observed to be0.03:1 whereas in similar cells exposed to inflammatoryagents, the ratio was increased to 0.28:1 (15).

Components Induced During Disease Episodes

Induction of enhanced iron withholding begins very early inan acute microbial invasion. Note in Table 3, the promptreduction in serum iron during the incubation phase of theinfection (16). The gradual restoration of normal iron metabo-lism as the illness wanes also is recorded in Table 3. Thehypoferremic response in a chronic inflammatory disease isshown in patients in various stages of illness (Table 4) (17).

In the induction phase of enhanced iron withholding, Tfiron saturation generally falls to between 5% and 16% as com-pared with values greater than 20% in normal persons and0–16% in iron deficiency (1). The reduction in Tf iron satura-tion to 5–16% precludes many, but not all, potential extracel-lular pathogens from obtaining growth-essential iron. Otherbiochemical changes that occur early in the invasion are listedin items 1–10 in Table 1.

The induced metabolic changes are initiated by suchcytokines formed by Th-1 lymphocytes as IFN-g, IL-2, andTNF-b. These cytokines activate normal macrophages to formproinflammatory compounds: IL-1, IL-6, and TNF-a (18).Additionally, as noted in items 9 and 10 of Table 1, IFN-g acti-vates macrophages that are being threatened by invasion toalter their iron metabolism to the detriment of the attackers(19). As the assault is overcome, the inflammatory responseis down regulated by formation by Th-2 lymphocytes of IL-4,IL-5, IL-10, and IL-13. These cytokines induce a strong anti-body response as indicated in item 11 of Table 1.

Hepatic synthesis of a 25-amino acid peptide, hepcidin,is increased greatly during inflammatory episodes and infec-tion as well as in iron loading conditions. The peptideinhibits absorption of iron in the small intestine andrelease of recycled iron from macrophages. In primary human

262 Weinberg

Table

4Hypoferremic

Respon

seto

Hod

gkin’s

Disea

seandto

Non

-Hod

gkin’s

Lymphom

a

Hod

gkin’sDisea

seNon

-Hod

gkin’s

Lymphom

aNormal

Stage1

Stage2

Stage3

Stage4

No.

ofpersons

12

67

73

6

Serum

iron

(mM)

22.7

�1.8

16.4

�1.6

11.4

�1.5

8.6

�1.5

6.97

�0.61

10.7

�2.0

Transferrin

saturation

(%)

35

�3.7

27

�3.5

23

�2.6

18

�3.5

16

�1.2

17

�3.3

Values

are

mea

ns�

SE.

(Data

from

Ref.17.)

Table

3Hypoferremic

Respon

seto

Infection

DayPostexposure

01

23

45

67

89

10

Mea

nseru

miron

(mM)

22.3

18.0

15.5

11.2

10.0

6.4

7.3

12.3

13.0

13.2

15.3

Mea

nfever

index

(h�

� F)

01

017

30

12

52

40

1

Fou

rhumanvolunteerswereex

posed

toFra

nciscella

tularensison

day0.They

dev

elop

edtypicalclinicalillnessasindicatedbyfever

ondays3–8.

(Data

from

Ref.16.)


hepatocytes, hepcidin mRNA was found to be induced by IL-6but not by either IL-l or TNF-a (19a).

A key effector molecule of macrophage-mediated antimi-crobial and antitumor activity is nitric oxide (NO). A widevariety of iron-containing molecules including ribonucleotidereductase and mitochondrial Fe–S proteins are targeted byNO. The compound acts similarly to iron chelators to induceiron release from tumor cells and to inhibit macrophage ironuptake from Tf (20). However, expression of the induced NOsynthase is suppressed by intracellular iron. For instance,in a plasmodial coculture system, the iron chelator deferoxa-mine increased both NO formation and parasite killing buthad no effect in the presence of an inhibitor of NO formation.Thus, the clearance of the protozoa apparently was due toenhanced generation of NO rather than to limitation of ironto the pathogen (21).

Nitric oxide is derived from one of the two guanido nitro-gen atoms of l-arginine. This amino acid has long been recog-nized to be important in antitumor defense (22). Unlikemammals, avian hosts lack a complete urea cycle and areunable to form l-arginine. However, their cytokine-stimulated macrophages, as in mammalian hosts, synthesizeNO from exogenous arginine.

Although the concept of iron withholding defense hasbeen recognized for nearly 60 years, complete understandingof all of its various components remains to be achieved. Forexample, the role of the Nramp1 gene in susceptibility orresistance to specific infections in humans remains uncertain.A study in West Africa found that persons with a polymorph-ism in the gene were over-represented in a group with tuber-culosis. An investigation in Brazil failed to confirm thisassociation (23). The Nramp1 gene encodes an integralmembrane glycoprotein expressed exclusively in the lateendosomes=lysosomes and phagosomes=lysosomes of macro-phages. The glycoprotein might suppress microbial growthby extruding iron from invaded phagosomes into the cyto-plasm of the macrophage; alternatively, it might cause deathof the invaders by increasing intraphagosomal iron toenhance microbicidal activity (24).

264 Weinberg

INVADER FACTORS ASSOCIATED WITH IRONWITHHOLDING DEFENSE

Counteraction of Iron Withholding

Bacterial, fungal, and protozoan pathogens have developedone or more of several strategies for securing host iron. Theseinclude (i) lysis of erythrocytes, digestion of hemoglobin, andbinding and assimilation of heme; (ii) cell surface binding offerrated Tf or Lf and extraction and assimilation of the metal;(iii) production of siderophores that extract the metal from Tfwith subsequent binding and assimilated of the ferratedsiderophores or of the metal; and (iv) assimilation of hostintracellular iron derived from pools of low molecular massiron-binding compounds (25).

Successful pathogens often employ differing strategiesdepending on the particular environment. Variation in strat-egy especially is important for strains that live in differenthosts or outside of hosts at various times in their life cycles.Furthermore, the differing biochemical environments invarious tissues of a host also necessitate shifts in mode ofiron acquisition (26).

For example, Salmonella strains utilize siderophores forgrowth in extracellular but not intracellular sites. Helicobac-ter pylori obtains iron directly from Lf when it is replicating inthe gastric lining but uses heme when it invades the gastricwall. Candida albicans, in its yeast phase, employs sidero-phores when growing in the gastrointestinal tract but utilizesheme in its hyphal phase when it is invading systemic tissues.

Extracellular promastigotes of Leishmania chagasi canacquire iron from three sources: heme, Tf, and Lf (27). Thisflexible capability may contribute to the proficiency of theprotozoan to survive in the diverse environments of insectand mammalian hosts. Microbial genera with strains whosegrowth in body fluids, cells, tissues, or intact vertebratehosts is stimulated by excess iron are listed in Table 5 (28).

Although viruses do not utilize iron, infected host cellsneed the metal to synthesize viral particles. Iron withholdingdefense is intensified during viral infections [e.g., Table 6(29)]. Factors that suppress iron withholding enhance viral


progression; factors that strengthen ironwithholding augmenthost antiviral defense (30).

Although normal and neoplastic mammalian cells have asimilar qualitative requirement for iron, the unrestricted pro-liferation of tumor cells would be expected to require anenhanced and probably diversified supply of the metal. Forinstance, in normal human B lymphocytes, expression of Tfreceptors is a multistep, tightly regulated process that

Table 5 Microbial Genera with Strains Whose Growth in BodyFluids, Cells, Tissues, or Intact Vertebrate Hosts Is Stimulatedby Excess Iron

Fungi: Candida, Cryptococcus, Histoplasma, Paracoccidiodes,Pneumocystis, Pythium, Rhizopus, Trichosporon

Protozoa: Entamoeba, Leishmania, Naegleria, Plasmodium, Toxoplasma,Trichomonas, Tritrichomonas, Trypanosoma

Gram positive and acid fast bacteria: Bacillus, Clostridium,Corynebacterium, Erysipelothrix, Listeria, Mycobacterium,Staphylococcus, Streptococcus, Tropheryma

Gram negative bacteria: Acinetobacter, Aeromonas, Alcaligenes,Campylobacter, Capnocytophaga, Chlamydia, Coxiella, Ehrlichia,Enterobacter, Escherichia, Helicobacter, Klebsiella, Legionella,Moraxella, Neisseria, Pasteurella, Proteus, Pseudomonas, Salmonella,Shigella, Vibrio, Yersinia

(Adapted from Ref. 28.)

Table 6 Mean Iron Values of 20 Children with Chicken Pox andSeven Children with Mumps

Stage of Disease

At presentation21 dayslater P value

Serum iron (mM) 5.2 15.2 <0.01Transferrin iron saturation (%) 10.9 26.6 <0.01Serum ferritin (ng mL�1) 59.6 35.6 <0.01

Median age of children: 6 years.Values for hemoglobin, hematocrit, and mean corpuscular volume remainedunchanged during the course of the diseases.(Adapted from Ref. 29.)

266 Weinberg

requires antigen or mitogen stimulation, IL-2, and IL-2 recep-tors. In contrast, malignant B lymphocytes express Tf recep-tors constitutively (31). Similarly, in cells of myogeniclineage, transformed cells were found to have more totaland expressed receptors and higher rates of Tf and iron endo-cytosis than normal cells proliferating at the same rate. More-over, small cell lung cancer cells can synthesize their own Tf,an ability that may account for the very short doubling time ofthis neoplasm even in sites that are not well vascularized.Additionally, in the absence of Tf, some, but not all, leukemiacell lines can obtain iron for DNA synthesis and growth from1–10 mM inorganic iron.

Melanotransferrin (MTf) is a 97kDa homologue of Tfthat is produced in very small amount by normal cells butin much larger quantity by some strains of melanoma andother tumor cell lines (32). However, it appears unlikely thatMTf is needed for enhanced iron acquisition. Melanoma cellsnot only obtain iron via Tf but also employ low affinityabsorption pinocytosis.

Suppression of Iron Withholding

Potential invaders might achieve success not only by develop-ing highly efficient mechanisms of iron extraction from hostsbut also by suppressing iron withholding efforts of the host.For example, Ehrlichia chafeensis is an obligatory intracellu-lar bacterium that infects human macrophages. The ehrli-chiae have been observed to suppress the infected host cell’sdown-regulation of Tf receptor (TfR) mRNA; this actionresulted in a 7-fold increase in TfR within 24 hr of inoculation(19). The pathogen also increased the binding activity of iron-responsive protein 1 to the iron-responsive element of the hostcells, thus preventing degradation of TfR mRNA.

Another bacterial pathogen that induces up-regulation ofTfR expression in infected host cells is Coxiella burnetii. In amurine macrophage-like cell line, up-regulation began within6 hr postinoculation and continued for 24 hr. The processpermitted a 2.5-fold increase in host cell iron and a7-fold increase in bacterial cell number (19).


Nitric oxide synthesis by macrophages can be inhibitedby invaders or their products. For instance, lipophosphogly-can, a surface molecule of promastigotes of Leishmania, sig-nificantly reduced expression of NO synthase mRNA ofmurine macrophages, provided that the glycan was presentprior to IFN-g activation of the host cells (19). Furthermore,the saliva of various sand fly vectors for species of Leishmaniacontains a factor that inhibits NO synthesis in IFN-g-activated human macrophages. The importance of this factorin the establishment of leishmaniasis suggests that, whenidentified, it might become a useful vaccine candidate (19).

Success Only in Absence of Iron Withholding

In contrast to potential pathogens that can overcome or evensuppress host iron withholding defenses, there exist potentialinvaders that have very little iron acquisition capability. Suchmicroorganisms cause disease primarily in iron-loaded hosts.Because of the relatively low percentage of iron-loadedhumans in any given population, patients with these patho-gens are not contagious. Moreover, these impaired pathogensare confined to specific host cells or tissues that provide highlyaccessible iron.

For example, systemic infections caused by Yersiniaenterocloitica and Y. pseudotuberculosis occur almost exclu-sively in persons who have an underlying iron loading condi-tion such as African siderosis, alcoholism, asplenia, excessiveparentereral or oral iron therapy, hemochromatosis, sidero-blastic anemia, thalassemia, or transfusional siderosis (33).These gram negative bacillary pathogens can grow either iniron-enriched host fluids or in macrophages that containelevated iron. The bacteria lack siderophores but contain side-rophore receptors. Thus iron-loaded persons who are beingde-ironed with the siderophore, deferoxamine, are especiallyat risk.

Vibrio vulnificus, an estuarine gram negative bacterialspecies, causes: (i) wound infections associated with sea waterexposure, and (ii) septicemias ensuing from consumption ofcontaminated raw oysters. An iron loading condition suchas alcoholism, chronic hepatitis, hemochromatosis, or

268 Weinberg

thalassemia is present in a very high percentage of patients(33). Virulence of isolates of V. vulnificus is significantlyassociated with ability to obtain iron from Tf molecules thatare highly saturated with the metal.

In a set of eight isolates of V. vulnificus from patients andeight from the marine environment, for instance, none couldgrow in the presence of Tf that had a normal iron saturationof 30%. At 100% Tf iron saturation, six of the patient isolatesand two of the environmental isolates could grow (33). Intra-peritoneal injection of a virulent strain of V. vulnificus intonormal mice required, for a 50% lethal dose, one million bac-terial cells. In contrast, in mice injected with a nontoxic doseof ferric salts, the 50% lethal dose was only one bacterial cell.

Legionella pneumophila is a pleomorphic gram negativebacterium that survives in nature by growing within a varietyof fresh water amoeba and ciliated protozoa (33). It entershumans upon inhalation of contaminated water droplets.The pathogen can evade host defenses only by parasitizingand multiplying within alveolar macrophages. The ensuingpneumonia is not contagious. Evidence that the bacteriumrequires unusually iron-rich macrophages is derived fromboth laboratory and clinical observations.

Laboratory studies have shown that, for extracellulargrowth in vitro, L. pneumophila must be cultured in highlyiron-enriched media. The pathogen produces neither sidero-phores nor siderophilin-binding receptors. Moreover, the bac-terium cannot proliferate within human phagocytic cellswhose iron level remains low due to activation by IFN-g. Simi-larly, chloroquine prevents pathogen growth by raising thepH value of the host cell endosome thereby preventingferrated Tf from releasing the metal to the bacteria.Selected mutants of L. pneumophila, whose ability to acquirehost cell iron is even more impaired than the wild type,completely lose pathogenicity as well as ability to grow insaprophytic protozoa (33).

Risk factors for development of pneumonia due toL. pneu-mophila include alveolar macrophage iron loading due tocigarette smoking, transfusion siderosis, cancer chemotherapyand organ transplantation, and acquired immunodeficiency


disease (AIDS) (33). Intensive chemotherapy for neoplasias orin preparation for organ transplantation can result inhigh levels of cellular accumulation of non-Tf bound iron. InAIDS patients, iron loading of macrophages is well documen-ted. In contrast, hemochromatosis patients have low ironmacrophages and thus have little risk for L. pneumophilapneumonia (33).

HOST FACTORS ASSOCIATED WITH IMPAIREDIRON WITHHOLDING DEFENSE

Behavioral and Nutritional Factors

Enteral supplementation of iron has been observed to increasethe risk of infectious disease. For example, in a salmonellosisoutbreak among 200 formula-fed infants <6 months of age,the 75 cases were three times more likely than the 125 healthyinfants to have received formula with an iron content >179mM (34). In a set of 69 cases of infant botulism, dietary ironeither from formula or other sources apparently contributedto the fulminant form of the disease (35). In a recent surveyof other reports (36), oral iron was associated with increasedrates of clinical malaria in five of nine studies and withincreased morbidity from other infectious diseases in four ofeight investigations. None of the surveyed studies in malarialregions showed benefits of iron supplementation (36).

Likewise, parenteral iron can impair iron withholdingdefense. Injection of 10mg=kg iron (as dextran) in 1587 younginfants resulted within one week in a 7-fold increase in septi-cemias and meningitis caused by E. coli and other gram nega-tive bacteria (37,38). In pregnant women, malarial relapsestriggered by injections of iron dextran were so numerous thatprophylactic chloroquine was recommended for the patientswho were to receive the parenteral iron (39).

Persons who inhale iron are at increased risk for neoplas-tic as well as infectious diseases. Workers in ferriferousindustries have elevated incidence of respiratory tract cancers(40) and of such agents of pneumonia as Acinetobactercalcoaceticus (41). Silicon dioxide, asbestos, and tobacco

270 Weinberg

smoke are heavily contaminated with iron; persons whochronically inhale any of these items have markedly elevatedrates of lung and related cancers (40).

Humans suffering from lack of protein nutrition such asin kwashiorkor may have serum Tf levels of only 10% of nor-mal with iron saturation >100% (41). The great risk of infec-tion in these patients has been exacerbated by health careworkers who fed iron without prior feeding of protein. Simi-larly, rats deprived of protein have low Tf levels; feeding ironto these animals has resulted in severe infections caused byovergrowth of their natural flora (41).

The importance of relatively unsaturated Tf in defenseagainst a common human pathogen, Streptococcus pneumo-niae, was illustrated in a study of 35 patients (42). Of 10 withunbound iron-binding capacity (UIBC)< 130mg=dL, six died,three had complications and only one recovered uneventfully.Of the 25 patients with UIBC>130mg=dL, only four died andtwo had complications.

Genetic Factors

As cited earlier, persons who become iron loaded because ofgene mutations [e.g., hemochromatosis (HH), thalassemia,African siderosis] have increased susceptibility to specificinfectious diseases. Gram negative bacterial infectionsare especially troublesome to persons with HH (as citedabove) and with thalassemia (43). In African siderosis, inwhich macrophage iron is high, tuberculosis is increased(44). However, in HH, macrophage iron is low; thesepersons would be expected to be relatively resistant totuberculosis.

In genetically based iron loading disorders, the risk ofneoplasms, especially of the liver is markedly raised. Forinstance, in a review of 601 autopsies of adults from sevencountries of central and southern Africa, 20% of the malesand 15% of the females had very marked liver siderosis(44,45). These persons had a 17.8-fold increase (CI 1.8–179.2) in death from hepatoma as compared with those whoseliver showed a slight or absent siderosis. Persons who had a


marked increase in hepatic iron had a 5.7-fold increase(CI 1.6–20.9) over the slight or absent iron group.

Thalassemia patients have high liver iron deposits andfibrosis with hemosiderin in both hepatocytes and macro-phages (46). The pattern of liver cell damage is similar to thatof viral hepatitis. In a study of 1087 thalassemic patients,morbidity from neoplasms ranked third behind heart diseaseand infection (43). The neoplasms included hepatic cell carci-noma, leukemias, lymphomas, and neuroblastoma.

In HH, the risk of hepatic carcinomas is elevated asmuch as 200 times (47). As in thalassemia, nonhepatic sitesof neoplasms also occur. For example, in 101 male HHpatients followed for an average of 4.1 years (48), the standar-dized incidence ratio (compared with the general population)of primary liver cancer, esophageal cancer, and skin mela-noma, respectively, was 92.9, 42.9, and 27.8.

In a set of 430 patients who had either multiple mye-loma, breast cancer or colorectal cancer, persons with HHgene mutations of CY=wt or CY=HD combined with homozyg-osity for a TfR gene mutation (substitution of serine for gly-cine at amino acid 142) were significantly prominent (49).For the three cancers, respectively, as compared with wt=wtcontrols, CY=wt carriers had increased risks of 2.2-, 2.2-,and 1.65-fold. Values for CY=HD carriers were 6.5-, 7.3-,and 8.7-fold. In wt=wt controls, the incidence of neoplasmswas not altered by the TfR gene mutation.

As indicated in Table 1 (item 8), haptoglobin (Hp) is aplasma protein responsible for the removal of free hemoglobin(Hb) from the circulation. Following hemolysis, stable Hp–Hbcomplexes are formed which are delivered to hepatocytes. Inhumans, a polymorphism results in three structurally differ-ent phenotypes: Hp-1, Hp-2, Hp 2-2. Hp-1 is a well-definedprotein of 86 kDa, Hp-2 is characterized by heteropolymers(86–300kDa), and Hp 2-2 forms large macromolecular com-plexes (170–1000kDa) (50). The phenotypic distribution inEuropeans is approximately 15% Hp-1, 50% Hp-2, and 35%Hp 2-2. The ability to bind Hb is lowest in Hp 2-2; accordingly,this phenotype, at least in males, is associated with elevationof serum iron, Tf iron saturation, and serum ferritin (50).

272 Weinberg

In a set of male HH patients homozygous for the CY genemutation, Hp 2–2 was found in 49% as compared with 35% inwt=wt controls. However, in patients with African siderosis,nodeviation inHp frequencywasobserved (50). Several studieshave found that elevated iron is a risk factor for HIV-1 infectedadults; the metal is associated with growth of opportunisticbacterial and fungal agents as well as with decreased longevity(51). Not surprisingly, Hp 2–2 patients with AIDS have beenreported to have lower median survival time (Hp-1 and Hp-2,11 years; Hp 2–2, 7.33 years) and higher levels of plasmaHIV-1 RNA (50).

Physiological Factors

In healthy persons, bicarbonate ions assist in prevention ofexcessive iron intake by raising the duodenal pH value whichlowers solubility of the metal. Pancreatic secretion of bicarbo-nate ions is depressed in cystic fibrosis. Patients with this con-dition who survive beyond age 25 have a heightened risk ofdeveloping intestinal, pancreatic, or testicular neoplasms (31).

In persons with normal iron levels, iron withholdingdefense can be compromised by release of the metal from itsregular body compartments; e.g., by hemolysis, bleeding, ordestruction of cells that function to sequester iron deposits.Persons with intestinal mucosal bleeding due to untreatedulcerative colitis or Crohn’s disease have, respectively, a 10–20- or a 4–7-fold increased risk of developing colorectal canceras compared with the general population. Patients withhematuria due to infestation with Schistosoma hematobiumhave an elevated risk of bladder cancer (31). The sustainedhepatonecrosis in chronic hepatitis B carriers is associatedwith raised serum iron and a markedly increased risk of hepa-tocellular carcinoma (52).

During clinical episodes of malaria, the considerabledestruction of both infected and noninfected erythrocytesresults in elevation of serum iron (41). The coincidence of sys-temic salmonellosis with malarial outbreaks and the greaterseverity of the bacterial disease in malarial than in nonmalar-ial patients often has been noted. Control of malarial attacks


by quinine was frequently reported to result in subsidence ofsalmonella bacteremias. In fact, the latter were such commoncomplications of malarial treatment of neurosyphilis in theearly 1930s that some clinicians considered giving salmonellaparatyphoid vaccines to patients who were to be inoculatedwith malarial protozoa (41).

Similarly, the severe hemolysis in the Oroya fever phaseof bartonellosis results in a hyperferremia conducive to devel-opment of systemic salmonellosis and to a variety of other sec-ondary bacterial infections. Inasmuch as up to 40% of patientsdevelop salmonellosis, Barton’s initial impression that salmo-nellae caused the hemolytic disease is understandable. Aswith malaria, the salmonellae are not associated with thearthropod vector of bartonellosis but rather enter and leavethe host by the oral-fecal route. Likewise, hemolytic episodesin patients with hematologic neoplasms result in a hyperfer-remia that is associated with an elevated frequency ofbacterial and fungal infections (41).

Fungal infections are enhanced also by physiologicconditions that cause a ketoacidosis. The inability of Tf toeffectively scavenge ‘‘free’’ iron as the pH value of serumdeclines below 7.4 permits development of mucormycosis (53).

ECOLOGICAL ASPECTS OF IRONWITHHOLDING DEFENSE

The outcome of the contest between host iron withholdingefforts and invader iron acquisition capabilities frequently isa determinant of susceptibility of specific host species and ofspecific tissues and cells. For instance, H. pylori, via its70 kDa Lf binding receptor, can acquire iron directly fromhuman Lf but not from bovine or equine Lf (13,26). Nor canit obtain iron from human, bovine or equine Tf. Thus, thepathogen cannot replicate in cattle or horses nor is it an effec-tive systemic invader of humans. The singular location of H.pylori in human gastric epithelium apparently is a conse-quence of the availability of human Lf and iron in gastricjuice. When H. pylori attempts to migrate into the stomach

274 Weinberg

wall where it could obtain iron from heme, the infected hostcan block the invasion by forming antibodies to the 77kDaheme binding receptor (26).

Binding of host siderophilins has been observed for Tri-chomonas vaginalis and Tritrichomonas foetus (27). The for-mer obtains iron from human Lf whereas the latter canacquire the metal from bovine Lf and, as well, from bovineTf. Thus, T. vaginalis grows primarily in the Lf-rich environ-ment of human vaginal mucus. Disease symptoms begin toexacerbate during menses at which time the vaginal concen-trations of Lf and iron are greater than at midcycle. In thehuman male urethra, the illness is self-limited or asympto-matic. Seminal fluid contains Tf and little Lf; urine containsneither siderophilin. Tritrichomonas foetus grows in thebovine vagina where Lf is available. The protozoan alsoinvades the uterus which has abundant Tf. Fortunately, thisinvader cannot utilize human siderophilins and thus is not ahuman pathogen.

Other examples of microorganisms with narrow hostranges because of specific siderophilin receptors are foundwithin the Neisseriaceae and the Pasteurellaceae (26). Forinstance, strains of Hemophilus somnus form receptors forbovine but not for human Tf; these bacteria are virulent forcattle but not for humans. Actinobacillus pleuropneumoniaesynthesizes a swine-specific Tf receptor and causes a seriouspneumonia only in pigs. The human pathogen, Neisseriameningitides can bind transferrins from humans and homi-nids such as chimpanzees, gorillas, and orangutans but notfrom monkeys or from nonprimate mammals. This pathogencan be induced to multiply in mice provided that the parent-eral inoculum consists of a mix of the bacteria and humanTf (41).

In contrast, some pathogens can readily obtain iron froma diversity of host species. Prominent among bacteria thatinvade a broad range of hosts is Listeria monocytogenes(26). This microorganism can be acquired by humans fromingestion of undercooked tissue of other mammals, birds, fish,or crustacea as well as from raw vegetables. Unable to pro-duce its own siderophores, the bacterium obtains iron by a


receptor that recognizes the chelated iron site of siderophoresof various other microorganisms or of natural catechols wide-spread in the environment. The metal then is reduced by acell surface bound reductase and assimilated. In mammalianhosts, dopamine and norepinephrine can function as xeno-siderophores owing to their ortho-diphenyl chelating site.Indeed, natural catecholamines released into brain tissueduring stress may be a determinant of the tropism of L. mono-cytogenes for our central nervous system.

REFERENCES

1. Lee GR. The anemia of chronic disease. Sem Hematol 1983;20:61–80.

2. Schade AL, Caroline L. Raw hen egg white and the role of ironin growth inhibition of Shigella dysenteriae, Staphylococcusaureus, Escherichia coli, and Saccharomyces cerevisiae.Science 1944; 100:14–15.

3. Schade AL, Caroline L. An iron-binding component in humanblood plasma. Science 1946; 104:340–341.

4. Holmberg CG, Laurell C-B. Investigations in serum copper. I.Nature of serum copper and its relation to the iron-bindingprotein in human serum. Acta Chem Scand 1947; 1:944–950.

5. Johansson BG. Isolation of an iron-containing red protein fromhuman milk. Acta Chem Scand 1960; 14:510–512.

6. Weinberg ED. The development of awareness of iron-withhold-ing defense. Persp Biol Med 1993; 36:215–221.

7. Locke A, Main ER, Rosbach DO. The copper and non-hemoglo-binous iron contents of the blood serum in disease. J ClinInvest 1932; 11:527–542.


9. Posey JE, Gherardini FC. Lack of a role for iron in the Lymedisease pathogen. Science 2000; 288:1651–1653.

276 Weinberg

10. Trousseau A. True and false chlorosis. In: Lectures on ClinicalMedicine. Vol. 5. Philadelphia: Lindsay & Blakiston, 1872:95–117.

11. Twort FW, Ingram GLY. A method for isolating and cultivat-ing the Mycobacterium enteriditis chromicae pseudotuberculo-sis bovis Johne, and some experiments on the preparation of adiagnostic vaccine for pseudotuberculosis enteritis of bovines.Proc R Soc Lond Ser B 1912; 84:517–530.

12. Parkkinen J, von Bonsdorff L, Ebeling F, Sahlstedt L. Func-tion and therapeutic development of apotransferrin. Vox Sang2002; 83(suppl 1):321–326.

13. Weinberg ED. Human lactoferrin: a novel therapeutic withbroad spectrum potential. J Pharm Pharmacol 2001;153:1303–1310.

14. Abdallah FB, Chahine J-MEH. Transferrins: iron release fromlactoferrin. J Mol Biol 2000; 303:255–266.

15. Alvarez-Hernandez X, Felstein MV, Brock JH. The relation-ship between iron release, ferritin synthesis and intracellulariron distribution in mouse peritoneal macrophages. BiochimBiophys Acta 1986; 886:214–222.

16. Pekarek RS, Bostian KA, Bartelloni PJ, Calla FM, Beisel WR.The effects of Francisella tularensis infection on iron metabo-lism in man. Am J Med Sci 1969; 258:14–25.

17. Beamish MR, Jones PA, Trevett D, Evans IH, Jacobs A. Ironmetabolism in Hofdgkin’s disease. Br J Cancer 1972; 26:444–452.

18. Weiss G. Iron and anemia of chronic disease. Kidney Int 1999;55(suppl 69):S12–S17.

19. Weinberg ED. Modulation of intramacrophage iron metabolismduring microbial cell invasion. Microbes Infect 2000; 2:85–89.

19a. Nemeth E, Valore EV, Territo M, Schiller G, Lichtenstein A,Ganz T. Hepcidin, a putative mediator of anemia of inflamma-tion, is a type II acute-phase protein. Blood 2003; 101:2461–2463.

20. Weiss G, Wachter H, Fuchs D. Linkage of cell-mediated immu-nity to iron metabolism. Immunol Today 1995; 16:495–499.


21. Fritsche G, Larcher C, Schennach H, Weiss G. Regulatoryinteractions between iron and nitric oxide metabolism forimmune defense against Plasmodium falciparum infection. JInfect Dis 2001; 183:1388–1394.

22. Weinberg ED. Iron depletion: a defense against intracellularinfection and neoplasia. Life Sci 1992; 50:1289–1297.

23. Pal S, Peterson EM, de la Maza LM. Role of Nramp1 deletionin Chlamydia infection in mice. Infect Immun 2000; 68:4831–4833.

24. Wyllie S, Seu P, Goss JA. The natural resistance-associatedmacrophage protein 1 Slcllal (formerly Nramp1) and ironmetabolism in macrophages. Microbes Infect 2002; 4:351–359.

25. WeinbergED.Acquisition of iron and other nutrients. In:RothJ,ed. Virulence Mechanisms of Bacterial Pathogens. Washington,DC: American Society for Microbiology, 1995:79–94.

26. Weinberg ED. Patho-ecological implications of microbial acqui-sition of host iron. Rev Med Microbiol 1998; 9:171–178.

27. Weinberg ED. Roles of iron in protozoan and fungal infectiousdiseases. J Eukar Microbiol 1999; 46:231–238.

28. Weinberg ED. Iron loading and disease surveillance. EmergInfect Dis 1999; 5:346–352.

29. Cemeroglu AP, Ozsoylu S. Haematological consequences ofviral infections including serum iron status. Eur J Paediatr1994; 153:171–173.

30. Weinberg ED. Iron withholding: a defense against viralinfections. BioMetals 1996; 9:393–399.

31. Weinberg ED. The role of iron in cancer. Eur J Cancer Prev1996; 5:19–36.

32. Richardson DR, Ponka P. The molecular mechanisms of themetabolism and transport of iron in normal and neoplasticcells. Biochim Biophys Acta 1997; 1331:1–40.

33. Weinberg ED. Microbial pathogens with impaired ability toacquire host iron. BioMetals 2000; 13:85–89.

34. Haddock RL, Cousens SN, Guzman CC. Infant diet and salmo-nellosis. Am J Public Health 1991; 81:997–1000.

278 Weinberg

35. Arnon SS. Infant botulism. Annu Rev Med 1980; 31:541–560.

36. Oppenheimer SJ. Iron and its relation to immunity andinfectious disease. J Nutr 2001; 131:616S–635S.

37. Barry DMJ, Reeve AW. Increased incidence of gram negativeneonatal sepsis with intramuscular iron administration.Pediatrics 1977; 60:908–912.

38. Becroft DMO, Dix MR, Farmer K. Intramuscular iron dextranand susceptibility of neonates to bacterial infections. Arch DisChild 1977; 52:778–781.

39. Byles AB, D’Sa A. Reduction of reaction due to iron dextraninfusion using chloroquin. Br Med J 1970; 3:625.

40. Weinberg ED. The development of awareness of the carcino-genic hazard of inhaled iron. Oncol Res 1999; 11:109–113.

41. Weinberg ED. Iron withholding: a defense against infectionand neoplasia. Physiol Rev 1984; 64:65–102.

42. Lambert CC, Hunter RL. Low levels of unsaturated transfer-rin as a predictor of survival in pneumococcal pneumonia.Ann Clin Lab Med 1990; 20:140–146.

43. ZurloMG,DeStefano P, Borgna-Pignatti C,Di PalmaA, PigaA,Melevendi C, Gregorio Fdi, Burattini MG, Terzoli S. Survivaland causes of death in thalassemia major. Lancet 1989; 2:227–230.

44. Gordeuk VR, MacPhail AP, McLaren CE, Deichsel G, BothwellTH. Association of iron overload in Africa with hepatocellularcarcinoma and tuberculosis: Strachan’s 1929 thesis revisited.Blood 1996; 87:3470–3476.

45. Gangaidzo IT, Gordeuk VR. Hepatocellular carcinoma andAfrican iron overload. Gut 1995; 37:727–730.

46. Thakerngpol K, Fucharoen S, Boonyaphipat P, Srisook K,Sahaphong S, Vathanophas V, Stitnimankarm T. Liver injurydue to iron overload in thalassemia: histopathologic and ultra-structural studies. BioMetals 1996; 9:177–183.

47. Niederau C, Fischer R, Sonnenberg A, Stremmel W,Trampisch HJ, Strohmeyer G. Survival and causes of deathin cirrhotic and in non-cirrhotic patients with primaryhemochromatosis. N Engl J Med 1985; 313:1256–1262.


48. Hsing AW, McLaughlin JK, Olsen JH, Mellemkjar L,Wacholder S, Fraumeni JF. Cancer risk following primaryhemochromatosis: a population-based study in Denmark. IntJ Cancer 1995; 60:160–162.

49. Beckman LE, Van Landeghem GF, Sikstrom C, Wahlin A,Markevan B, Hallmans G, Lenner P, Athlin L, Stenling R,Beckman LE. Interaction between haemochromatosis andtransferrin receptor genes in different neoplastic disorders.Carcinogenesis 1999; 20:1231–1233.

50. Delanghe JR, Longlois MR. Haptoglobin polymorphism andbody iron stores. Clin Chem Lab Med 2002; 40:212–216.

51. Weinberg GA, Boelaert JR, Weinberg ED. Iron and HIV infec-tion. In: Friis H, ed. Micronutrients and HIV Infection. BocaRaton, FL: CRC Press, 2002:135–157.

52. Israel JL, McGlynn KA, Hann H-WL, Blumberg BS. Iron-related markers in liver cancer. In: de Sousa M, Brock JH,eds. Iron in Immunity, Cancer and Inflammation. Chichester:John Wiley, 1989:301–316.

53. Artis WM, Fountain JA, Delcher HK, Jones HE. A mechanismof susceptibility to mucormycosis in diabetic ketoacidosis:transferrin and iron availability. Diabetes 1982; 31:1109–1114.

280 Weinberg

10

Iron, Iron Genes, and theImmune System

C.S. CARDOSO

Molecular Immunology, Institute forMolecular and Cell Biology,

Oporto, Portugal

G. WEISS

Department of Internal Medicine,University Hospital,

Innsbruck, Austria

M. DE SOUSA

Molecular Immunology, Institute forMolecular and Cell Biology, and

Molecular Pathology and Immunology,Instituto de Ciencias Biomedicas Abel

Salazar, Oporto, Portugal

INTRODUCTION

The notion that cells of the immunological systemhave a role inthe surveillance of iron toxicitywasfirst put forward in1978 (1).Twenty-five years ago, the exercise of such a function could

281

only be visualized through the association of ‘‘classical’’ ironbinding and storage proteins with the major immune cell sets(Table 1) (2–23). The last seven of those 25 years have seen anextraordinary explosion in the identification of genes and pro-teins involved in the regulation of iron metabolism (for review,see Ref. 24). A significant proportion of the new ‘‘iron genes,’’however, is related to the immune systemand=or has a functionin innate immunity (Table 1) (25). On the other hand, a numberof immunological molecules such as b2 microglobulin (b2m),MHC-class I and HFE have been shown to have a role in ironmetabolism (26–28). Finally, another family of immune cell pro-teins, the cytokines, is now known to have significant roles incellular and systemic iron metabolism (29).

In this chapter, some of the advances made in the under-standing of the role of cells and genes of the immunologicalsystem in iron homeostasis are reviewed.

Material included in earlier textbooks (30) may howeverbe of interest to some readers, especially those interested inthe acceleration of the History of Biology in the second halfof the 20th century. An acceleration catapulted by the discov-ery of the structure of the DNA (31) and the construction ofgenetically modified animals by gene targeting (32) to whichthe study of iron metabolism has not escaped.

The most significant impact of an action of the immuno-logical system on other pathways is necessary to be systemic(21,33). Participation of the immunological system in sys-temic iron metabolism is seen both in the physiology of ironmetabolism and in pathological conditions such as iron defi-ciency and iron overload. Nevertheless by virtue of the capa-city of its circulating cells, namely polymorphs, macrophages,and lymphocytes to migrate to sites of inflammation and=orspecific immune responses, immune system cells have also aunique role in ‘‘localized,’’ regional iron metabolism.

IRON HOMEOSTASIS: A BRIEF SUMMARY

The behavior of iron in aqueous media such as in livingorganisms is largely determined by oxidation–reduction reac-tions between the ferrous (Fe2þ) and ferric (Fe3þ) forms and

282 Cardoso et al.

Table 1 Relevant Proteins and=or Genes Involved in IronHomeostasis and Related Immunological Functions

Relevant ironmetabolism proteinsand genes

Chromosomelocation

Corresponding relevant immunesystem function, cells orproteins

1. ‘‘Classical’’ iron binding and transport proteinsLactoferrin 3q21–q23 Produced by polymorphs. Iron

withholding at low pHBactericidal properties (5) andrecently demonstrated antiviralproperties (6)

H-ferritin 11q12 Synthesized by macrophages andT lymphocytes (2,7)

L-ferritin 19q13 Macrophages (7)Transferrin 3q21 Present in monocyte=macrophages

and T lymphocytes (3)Transferrin

receptor 13q29 Iron entry into activated

lymphocytes; TfR1 is requiredfor DNA synthesis andcell division ofT lymphocytes (8)

2. ‘‘Nonclassical’’ iron binding and transport proteinsNramp1 (SCLA11) Putative iron transporter

expressed in the latephagolysosome of neutrophilsand macrophages conferringresistance to infection withintracellular pathogens (9)

DMT-1 (Nramp2 orDCT1)

12q13 Homologue to Nramp1,ubiquitously expressed, involvedin cellular iron uptake andtransfer (10,11)

Ferroportin 1(IREG1)

2q32 Expressed in splenic, liver, andbone marrow macrophages(12,13). Down-regulated byacute inflammation innormal mice (13)

Hephaestin Xq11–12 UnknownTransferrin

receptor 2 (a and bsubunits)

7q22 Unknown

(Continued)

Iron, Iron Genes, and the Immune System 283

hydrolysis to form hydroxides, which are often insolubleand=or unable to enter in metabolic pathways. In the aerobicextracellular environment, which exists in most multicellularorganisms, the more soluble form of iron (Fe2þ) is readily

Table 1 Relevant Proteins andor Genes Involved in IronHomeostasis and Related Immunological Functions

Relevant ironmetabolism proteinsand genes

Chromosomelocation

Corresponding relevant immunesystem function, cells orproteins

3. ‘‘Modifier’’ proteinsb2-Microglobulin 15q21–q22 Critical to the folding of the MHC-

class I molecules (14). b2mtransactivation is up-regulatedby several cytokines suchTNF-a, IFN-a, b, and g (15)

Ceruloplasmin 3q23–q24 Expressed in lymphocytesand macrophages (16)

Cytokines Multiple Multiple (see Sec. 3 in thischapter)

Dcytb Unknown UnknownHepcidin 19q13 Antimicrobial properties.

Upregulated by IL-6and by medium conditionedby monocytes treated withLPS (17). Upregulated byinflammatory stimuli (18,19)

Heme-oxygenase 1 22q12 Knockout mice have highperipheral blood cell countsand high lymph node andsplenic CD4:CD8 ratios (20)

HFE 6p21 Nonclassical MHC-class Imolecule (21)

IRP-1 9 Ehrlichia chaffensis andE. senetsu (intracellular bacteriain human monocytes andmacrophages) infectionsincrease the binding activityof IRP-1 to the IRE (22)

IRP-2 15 UnknownMHC-class I 6p21 Critical molecule to present viral

peptides (23)

284 Cardoso et al.

oxidized to the much less soluble form (Fe3þ) rendering itunavailable for cellular needs. In evolution, the developmentof complex mechanisms and agents (such as pH and iron bind-ing proteins) was crucial to maintain iron in a soluble formavailable to cells.

Iron is an essential element for all living cells and plays animportant role in fundamental metabolic pathways as oxygentransport, electron transfer, nitrogen fixation, and DNA synth-esis. However, iron in excess of its cellular needs is extremelytoxic, and could generate toxic radicals in biological fluidsthrough the Haber–Weiss and Fenton reactions. Under physio-logical circumstances, the quantity of iron in the body iscarefully controlled in part to secure independence from theexternal milieu (21) and, at the same time, to avoid theaccumulation of amounts of iron that can exceed the capacityof the body storage and consequently to produce toxicradicals.

In evolution, it became of the utmost importance for boththe cells (regulation at cellular level) and the organism (regula-tion at systemic level) to maintain iron homeostasis in order toensure iron supply but toprevent accumulation of excess of iron.

Regulation of Iron at Systemic Level (Fig. 1)

Iron represents 55 and 45 mg=kg of body weight in adult manand woman, respectively. Normally, 60–70% of total bodyiron is present in hemoglobin in circulating erythrocytes.Myoglobin, cytochromes, and other iron-containing enzymescomprise a further 10% and the remaining 20–30% is storedin ferritin. Iron bound to transferrin (Tf) represents lessthan 1% of the total body iron (reviewed in Refs. 34, 35).

Iron Absorption

Food iron occurs largely as either ferric iron or heme iron.Ferric iron is insoluble in solutions with a pH greater than3 and needs to be chelated in the stomach to be availablefor absorption in the less acidic environment of the duode-num. This process is accomplished by intestinal mucins andcertain components of the diet (reviewed in Ref. 36). Heme


iron is the major source of iron, soluble at alkaline pH andobtained from hemoglobin and myoglobin by pancreaticenzymes. Heme enters the intestinal absorptive cells as anintact metalloporphyrin where the porphyrin ring is split byheme-oxygenase with release of the inorganic iron (36).

Iron absorption occurs mostly in the duodenum and jeju-num where the mucosal absorptive cells remain attuned tocurrent body requirements for iron (reviewed in Refs. 37,38). The most important known stimuli to iron absorptionare tissue iron stores, the rate of erythropoiesis, and hypoxia(36). Iron absorption proceeds in two steps: the mucosal

Figure 1 Schematic representation of systemic iron metabolism.The arrows represent the circulation of iron in the human body.The red boxes represent the functional iron compartment, the blueboxes the iron storage compartment, the green box the site of ironabsorption, and the dashed red box represents the periodical bloodloss by menstruating women. NB: The central role of the monocy-te=macrophage system in the recycling of iron from senescent redcells is indicated in the central box.

286 Cardoso et al.

uptake of iron from the lumen of the gut and the transfer ofthis iron from the intestinal cells into the body (39).

Once entering the body, iron stays over a long period andits supply to cells is largely dependent on iron turnover(reviewed in Ref. 40). The pivotal cell for iron recycling andconsequently to the independence of the organism from theexternal milieu (21) is the macrophage.

Iron Turnover=Reutilization

After absorption, iron is carried to the bone marrow as ferriciron bound to Tf, where it binds to the transferrin receptors(TfRs) in red cell precursors and is utilized for hemoglobinsynthesis, for that it is reduced into the ferrous form andtransferred to the protoporphyrin ring. The synthesis andthe subsequent attachment to the globin take place in thebone marrow in the late stages of the development of ery-throid cells (35). Most of the turnover of Tf iron (80%) is uti-lized for erythropoiesis. From bone marrow, reticulocytesare released into the circulation where they develop within24 hr into mature erythrocytes that circulate in the bloodfor about 120 days. Senescent and damaged erythrocytesare removed from the circulation by the macrophages of themononuclear phagocytic system in the spleen, liver, and bonemarrow where iron is catabolized by heme-oxygenase andreturned into the circulation bound to Tf (35,41). The mechan-ism involved in the release of iron from reticuloendothelialcells has not been defined. The remaining iron of thedaily plasma iron turnover is exchanged with nonerythroidtissues mainly in the liver (35). At most, 1% of iron is recycledeveryday (40).

Physiological Iron Loss

The physiological loss of iron is a result of blood loss and des-quamation of the epithelium of skin and intestinal cells andintestinal secretions. However, the amount of iron excretedper day is negligible with the exception of menstruatingwomen who lose blood periodically (36,41).


The Immunological System in Systemic IronMetabolism

Physiology of Iron Handling by Immune Cells

Iron Uptake (Table 2)

Immune cells have developed different mechanisms toacquire iron, either to achieve a sufficient supply of thisessential metal for central cellular functions, such as citricacid cycle, oxidative phosphorylation, or DNA synthesis, touse the metal for immune effector mechanisms such asradical formation via the Haber–Weiss reaction, or to harborand store iron during inflammatory processes in order towithhold this essential growth factor from invading patho-gens and tumor cells as outlined in the chapter by Weinbergin this book.

In contrast to other immune cells, all lymphocytes sub-sets, i.e., B- and T-lymphocytes as well as natural killer (NK)cells, are extremely dependent on TfR mediated iron uptakeand a blockade of this pathway leads to diminished prolifera-tion and differentiation of these cells (34,42). Accordingly,mitogenic stimuli, such as phytohemagglutinin, increase TfRsurface expression on B- and T cells (43–45). However, the lym-phocyte subsets differ in their dependence on iron in generaland specifically on Tf mediated iron uptake. This is of greatfunctional relevance for immunity. In this line, the inductionof experimental iron overload to rats resulted in a shift in the

Table 2 Iron Uptake Mechanism in Immune Cells

Monocyte=macrophages Lymphocytes Neutrophils

Transferrin receptor (TfR) þ þþ þþFerritin receptors þ þ –Non-TfR-mediated ironuptake (DMT-1?)

þ n.k. n.k.

Erythrophagocytosis þ – –Lactoferrin þ þ þþ

þþ, very important; þ, available; –, not important or available; n.k., not known.

288 Cardoso et al.

ratio between T-helper (CD4þ) and T-suppressor=cytotoxic Tcells (CD8þ) with a relative expansion of the latter (46). More-over, even T-helper (Th) subset responds differently to ironperturbations (47). While Th-1 clones are very sensitive totreatment with anti-TfR antibodies resulting in inhibition oftheir DNA synthesis, Th-2 cells are resistant to this procedure.This was partly attributed to the fact that Th-2 clones exhibitlarger chelatable iron storage pools than Th-1 cells. Thus,Th-1 mediated immune effector pathways are much more sen-sitive to changes in iron homeostasis in vivo. The latter may bealso due to the fact that iron perturbations affect the activity ofcytokines secreted by Th-1 cells such as IFN-g (48). This is ofgeneral importance because IFN-g modulates: (i) the expres-sion of critical iron proteins, e.g., in monocytes (see below),and (ii) because of the regulatory properties of this cytokinefor the activation=proliferation of lymphocyte subsets accord-ing to the Th-1=Th-2 paradigm (49).

The expression of TfR is influenced by cytokines such asIL-2 but may also be induced upon decrease of intracellular ironlevels following a proliferative response leading to activation ofiron regulatory protein (IRP) binding to iron responsive ele-ments (IREs) with subsequent TfR mRNA stabilization (29,50).

Once TfRs are expressed, Tf bound iron is taken up effec-tively via endocytosis and subjected to iron storage (ferritinincorporation) or iron consumption (iron-containing enzymes),while a certain percentage of iron (10–20%) remains in thelabile iron pool. Nonetheless, lymphocytes have only a limitedcapacity to sequester iron and hydrophilic iron chelates(51,52). The reported association of the TfR with the T-cellreceptor (TcR) zeta-chain as well as with the zeta-bindingZAP-70 tyrosine kinase is an additional illustration of theclose association between key immunological functions andcellular iron metabolism (53).

Apart from the ‘‘classical’’ iron uptake mechanism viaTf-iron=TfR interactions, nontransferrinmediated iron uptakemechanism has been described in lymphocytes. These include:

i. Non-Tf mediated uptake of chelatable iron, as shownin a malignant B-lymphoblastic cell line or in


phytohemagglutinin stimulated human peripherallymphocytes which acts by a high capacity, lowaffinity, temperature, and calcium dependentmechanism (54,55).

In the meantime, a considerable number of new irontransporters has been identified in various organisms rangingfrom yeast to rodents and humans. Iron transporters whichare of special interest in connection with the immune systeminclude natural resistance associated macrophage proteins(Nramp) 1 and DMT-1 (Nramp2) which are able to transportferrous iron across cell membranes and phagolysosomes, andferroportin 1 all of which are discussed later in this chapter.

ii. All lymphocytes subtypes express receptors for H-but not for L-ferritin, that specifically bind H-ferri-tin (56,57). The expression of these receptors wasobserved on both B- and T cells and peaked duringthe S-phase, suggesting that ferritin receptorexpression may be associated with proliferation(58). However, it is not clear so far whether ornot these receptors may be involved in iron uptakeby lymphocytes or iron exchange, as assumed foriron traffic between hepatocytes and macro-phages, or whether they have functions, thatmay be distinct from iron metabolism.

iii. It has been known for several years that lympho-cytes express receptors, which are capable tointeract with Lactoferrin (Lf), an iron binding pro-tein that is found in milk, mucosa, or neutrophilgranules (59). Lactoferrin receptors are present onalmost all lymphocyte types investigated so far,i.e., CD4þ and CD8þ T cells, B cells or NK-cellsand the highest proportion of Lf receptors wasfound in gd T cells (60). The expression of Lf recep-tors appears to be associated with lymphocyte acti-vation since they are hardly expressed byunstimulated blood lymphocytes. Moreover, Lfreceptors show similar expression kinetics to TfR,although a considerable amount of lymphocytes

290 Cardoso et al.

express only Lf receptors but no TfR. Apart fromthis, Lf has various other immunoregulatory func-tions. These include modulation of cytokine forma-tion, antibody and complement production, andNKfunction (59,61). Recently, Lf has been shown toinhibit mast cell activation (62), and affect thedevelopment of dextran-induced acute colitis (61)and protect development of Mycobacteria infectionin b2-m knock-out mice (63).

Monocyte=macrophages not only acquire iron for theirown needs, participate in systemic iron homeostasis throughtheir role in erythrophagocytosis but also act as a major sto-rage pool of iron which is primarily expanded under inflam-matory conditions or in case of secondary iron overload. Incontrast to lymphocytes, monocyte=macrophages are differen-tiating but no more proliferating cells, thus they are not somuch dependent on TfR mediated iron uptake. On primaryhuman monocytes, TfR expression is rather low or almostabsent, while it is enhanced in cultured macrophages, result-ing in a progressive endosomal uptake of iron Tf complexesand rapid incorporation of iron into ferritin (64).

In rat macrophages, Tf bound iron is taken up by a non-receptor mediated mechanism, which is independent from thedegree of Tf saturation (65). Secondly, human macrophagesare able to take up iron-chelates with a greater efficacy thandiferric Tf by a temperature dependent but pH independentprocess (66). This latter iron uptake process was inhibitedby other divalent-metal chelates which suggests that DMT-1may be centrally involved since this membrane protein hassimilar transport kinetics for iron and other divalent-metalions such as zinc, copper, lead, etc. (10,11).

Moreover, macrophages express Nramp1 in phagolyso-soms that has previously been identified as the gene asso-ciated with resistance towards infections with intracellularpathogens such as Leishmania, Salmonella, or Mycobacteriaspecies (67,68). Nramp1 is able to transport iron across mem-branes (69,70). The notion of intracellular iron trafficking andmobilization of iron from intracellular vesicles is supported by


several investigators (69,71). Macrophages do not produce Lf;however, they express Lf receptors on their surface that bindLf (72). Interestingly, after incubation of macrophages in Lffree media, this milk specific protein was released from cells,previously loaded with Lf thus providing evidence for a recycl-able pool of Lf within macrophages (73). Once taken up, Lfmay play a regulatory role within macrophages by modulat-ing on the one hand iron mediated cytotoxic effector mechan-ism against intracellular pathogens via the formation ofhydroxyl radicals (74), while on the other hand apo-Lf mayprotect macrophages from membrane peroxidation.

In addition, monocyte=macrophages are able to acquireand recirculate iron from erythrocytes. Erythrocytes are takenup by phagocytosis and are then destroyed within monocy-te=macrophages. Macrophages have been shown to phagocy-tize about three times more erythrocytes than monocytes(75). Moreover, upon stimulation of macrophages with TNF-a, the phagocytosis of sialidase-treated erythrocytes is furtherincreased, at least in vitro, which may be due to enhancedexpression of C3bi (CD11b=CD18) receptors following cyto-kine treatment (76). Within the macrophage, iron is thenreleased from hemoglobin and rapidly shifted to reutilizationvia incorporation in iron proteins or storage by ferritin(75,77,78).

As with lymphocytes, the ability of neutrophil granulocytesto proliferate is closely related to the expression of surface TfR.In a recent study, TfR positive HL-60 cells had a tendency toproliferate, while TfR negative cells expressed markers ofdifferentiation and activation such as superoxide radicals orformyl-Met-Leu-Phe receptors (79). Interestingly, neutrophilsmay contribute to the release of iron from Tf via formationand release of superoxide. By this pathway, neutrophils maydeliver metabolic active iron for cytotoxic effector functionssuch as hydroxyl radical formation or lipid peroxidation (80).On the other hand, neutrophils downregulate the binding ofdiferric Tf to their specific surface receptors via the release ofmyeloperoxidase, an enzyme involved in radical formationby the cells (81). Moreover, neutrophils may influence ironhomeostasis in neighboring cells in a paracrine manner since

292 Cardoso et al.

addition of hydrogen peroxide to cells has been shown to stimu-late IRP-1 binding to IREs with the subsequent metabolicchanges on TfR and ferritin expression described above (82).

Lactoferrin is centrally involved in iron handling byneutrophils (59). Lactoferrin is a member of the transferrinfamily. Lf is an iron-binding protein present in various secre-tions (milk, mucosal secretions, sweat). It is also found in spe-cific granules of polymorphonuclear granulocytes, neutrophils(83), from which it is released following activation. Lf exerts abactericidal activity by damaging the outer membrane ofgram-negative bacteria and prevents bacterial biofilm devel-opment (84). The exact function of this protein in vivo remainsto be established. Knockout mice for Lf are viable and fertile,develop normally, and display no overt abnormalities (85).Adult mice on basal or high-iron diet revealed no differenceson transferrin saturation (TfSat) or tissue iron stores in bothdiets between wt and Lf�=– mice (85). Earlier studies on func-tion of Lf concentrates on bactericidal activity of Lf. Recently,other authors have shown that Lf has also a role in antimyco-bacterial infection (63) and antiviral protection (6). Lactofer-rin may be on the one hand directly toxic to microorganismsby limiting iron availability to them, while at the same time,it exerts distinct effects on overall immune-function byregulating the proliferation and activation of lymphocytes,NK-cells, and monocytes (5,86). Leukocyte migration into tis-sues is the hallmark of all types of inflammatory responses.Polymorphonuclear neutrophils are the inflammatory cellswith oxidative and proteolytic potential and are the first cellsinvolved in pathogen recognition at specific sites playing anessential role in host defense against invading microorgan-isms. Recently, Goetz et al. (87) have shown that a neutrophillipocalin NGAL, a 25kDa protein, has specificity and high affi-nity for bacterial catecholate-type ferric siderophores partici-pating in the antibacterial iron-depletion strategy (87).NGAL is a component of neutrophil granules but is alsoexpressed in epithelial cells in response to inflammatorysignals (reviewed in Ref. 88). In addition to its ferric sidero-phore binding capacity, NGAL has been implicated in celldifferentiation, tumorogenesis, and apoptosis (88,89).


Bacteria have evolved multiple sophisticated mechan-isms (siderophores) for acquiring iron in environments wherevery little free iron is available. Cells of the immune systemexercising their unique capacity of migrating to sites of infec-tion and inflammation have a role in securing a ‘‘defensive’’ orprotective function from iron utilization by bacteria.

Iron Release

Macrophages are the key cells securing the recycling ofsenescent red blood cells; however, the mechanism involved inthe release of iron is not clear. Presently, the genes known to beinvolved in the iron release process are HFE and ferroportin 1.

Ferroportin 1: Ferroportin 1 is an iron exporter gene thatis implicated in the basolateral transfer of iron into the circu-lation and was mapped on to chromosome 2q32 by three inde-pendent groups (12,90,91). This protein facilitates the ironexport and is expressed in enterocytes and in macrophages.

Two mutations were found by two independent groups onthe iron transporter gene ferroportin 1 (also called IREG1 orSLC11A3) in two large Dutch (92) and Italian (93) families inwhich hemochromatosis segregates as a dominant trait. Njajouet al. (92) found a missense mutation N144H (A-to-C substitu-tion at nucleotide 734 resulting in an Asn 144 to His substitu-tion) in the heterozygous state in all affected patients. Montosiet al. (93) found an A77Dmutation (C-to-A substitution at exon3 resulting in an Ala 77 to Asp substitution) also in heterozy-gous form in all affected individuals. Accumulation of iron inmacrophages in these patients is an indication that ferroportinmutations affect iron export from macrophages.

HFE: The HFE gene encodes for a 343 amino acid pro-tein, highly similar to the MHC-class I molecules (namelyHLA-A2) (94). The structural organization of the moleculeincludes the signal sequence, the putative peptide-bindingregion (a1 and a2 domains), an immunoglobulin-like domain(a3 domain), a transmembrane region, and a small cytoplas-mic portion (94). One of the most important conserved struc-tural features of the HFE is the 4-cysteine residues that formdisulfide bridges in the a2 and a3 domains. The correct confor-mation of the a3 domain is required for noncovalent interac-

294 Cardoso et al.

tion with b2m and correct cell surface expression of the mole-cule. The HFE gene has two common missense mutations:C282Y and H63D (94). The C282Y mutation is a substitutionof a guanidine to an adenine at nucleotide 845, resulting in asubstitution of a tyrosine for a cysteine at amino acid 282. TheC282Y mutation prevents the correct folding of the a3 domainand hence interferes with its interaction with b2m abolishingthe cell surface expression of the molecule (95,96). The H63Dmutation is a substitution of a cytosine to a guanidine atnucleotide 187, resulting in a substitution of an aspartatefor a histidine at amino acid 63. The H63D mutation is pre-dicted to occur in the a1 domain that is located on the loopbetween the third and fourth b strands of the peptide-bindingdomain and unlike the C282Ymutation does not affect the cellsurface expression of the molecule (95,96). The two mutationscan also be designated C260Y and H41D if the first 22 aminoacids of the signal sequence of the mature protein are omitted.The C282Y mutation in homozygous form is present in themajority of hereditary hemochromatosis (HH) Caucasianpatients (97). HH is a genetic disease of iron overload that ischaracterized by excessive accumulation of iron in parenchymalcells in various organs but a paucity of iron in Kupffer cellsand macrophages (98). The role of HFE in iron metabolismwas suggested by studies showing that HFE is physicallyassociated with transferrin receptor 1 (99–105), but theprecise role by which HFE regulates iron metabolism is stillcontroversial.

Iron Release from Macrophagesof Hereditary Hemochromatosis Patients

Early studies investigating the release of the erythro-cyte-derived iron from purified human monocytes from con-trols and HH patients showed that iron is released in theform of ferritin, hemoglobin, and nonprotein bound low mole-cular weight iron (LMW-Fe) in both groups of subjects (77).Although HH monocytes phagocytosed less than half thenumber of erythrocytes taken up when compared with controlmonocytes, HH monocytes release twice as much iron in theform of LMW-Fe complex than controls (77).


More recently, Montosi et al. (106) observed in macro-phages from HH patients (all carriers of the C282Y mutation)transfected with Wt HFE, an increase of iron content in thecells by both increasing Fe-Tf and Ft pool when comparedwith HH macrophages transfected with control vector. It isof note that one of the characteristics of HH macrophages isthe iron-deficient phenotype.

Drakesmith et al. (107) showed that in THP-1, a maturemonocytic cell line, the cells accumulate iron in response toHFE, whereas HeLa cells and a less mature monocytic cellline reduce their iron content. The authors also show thatHFE inhibits iron release independently of binding to TfRand the H63D mutation has lost the ability to inhibit ironrelease despite binding to TfR.

Iron Metabolism and the Immune System UnderPathological Conditions

New Regulatory Molecules for IronHomeostasis in Inflammation

Several immune system cells may be involved insystemic conditions of iron deficiency under inflammatoryconditions: the macrophages by accumulating iron and bythe failure to release the metal for erythropoiesis. Thesemechanisms have been the subject of earlier reviews(108,109) and Chapter 4 in this book.

Macrophages stimulated by LPS seem to induce in turnthe production of hepcidin by hepatocytes. Presently, hepcidinmay be a significant molecule in the development of anemia inresponse to infection (110). The point we wish to reiteratehere is that in response to infection an immune system cell,again the macrophage, may directly affect erythropoiesis bysequestration of iron, and indirectly through the upregulationof hepcidin. In addition, LPS stimulated mice downregulatethe expression of ferroportin 1 as do human monocytic cellsfollowing cytokine treatment, suggesting a fundamental rolefor this protein in the iron sequestration by macrophages dur-ing an infection (13,111). Studies done in a murine macro-phage cell line (RAW) showed that DMT-1=Nramp2 is

296 Cardoso et al.

localized in phagosomal membrane and it is also associatedwith erythrocyte-containing phagosomes (112). This proteinis closely related to Nramp1, which is expressed exclusivelyin phagosomes and has an antimicrobial activity againstintracellular pathogens (113,114).

Hepcidin (Table 3 and Refs. 17–19, 109, 115–120): Hepci-din is a small disulfide-bond peptide with 20, 22, or 25 amino

Table 3 Differences in the Expression of Hepcidin in Mice andHumans

Expression levels Observation References

Low levels ofexpression

In mice after induced anemia bybleeding or by phenylhydrazinetreatment

19

In sla, mk and hpx mice 109In mice after 12 days of hypoxia 19In Hfe–=– mice 115–117HepG2 and Hep3B humanhepatoma cell lines underhypoxia conditions

19

Downregulated in HepG2 inresponse to Fe-NTA in aconcentration dependent manner

118

Human hepatocytes cultured with10mM ferric ammonium citrateor iron saturated Tf resulted in50% decrease in hepcidin mRNA

17

In untreated HH patients 116In patients with a new type ofsevere juvenilehemochromatosis carrying twodifferent nonsense mutations inhepcidin gene

119

High levels ofexpression

In mice undergoingexperimentally induced ironloading in the liver

18

In b2m�=– mice under normal diet 18Adenoma tissue from two patientswith iron refractory anemia dueto type 1 glycogen storage disease

120

Mice injected with LPS 18Mice injected with turpentine 19


acids. It is produced exclusively in the liver and detectablein serum and urine (121,122). This protein has a good anti-microbial activity (122). Mice with a targeted disruption ofupstream stimulatory factor 2 (USF2) do not express hepci-din genes and develop hepatic iron overload associated withdecreased iron in tissue macrophages (123) suggesting thathepcidin could act to limit the intestinal iron absorptionand macrophage iron release. It should, however, be notedthat recent work by Gobin et al. (15) has shown thatUSF2 binds to the E box of the promoter of b2m and regu-lates the b2m transactivation (15). Inversely, transgenicmice over expressing hepcidin in the liver have decreasedbody iron levels and died at birth of severe iron deficiencyanemia (124). Recently, Nicolas et al. (125) have shown thatliver iron accumulation in mice lacking the Hfe gene can beprevented by over-expression of the hepcidin transgene,reinforcing the evidence that hepcidin has a fundamentalrole in iron homeostasis. Differences in the expression ofhepcidin were observed both in iron related conditionsand inflammatory stimuli (Table 3). Lower hepcidin geneexpression was found in mice after induction of anemia,both by bleeding (iron-deficient anemia) and by phenylhy-drazine treatment (hemolytic anemia). Mice carrying muta-tions that disrupt intestinal iron transport (sla and mk) oriron uptake by erythroblasts (mk and hpx) are anemic(reviewed in Ref. 24) and have also decreased expression ofhepcidin (110). In addition to animal models, Nemeth et al.(17) report an increased excretion of hepcidin in urine ofpatients with anemia of inflammation. In a patient with epi-dymidites and sepsis, the same authors observed a very highurinary hepcidin excretion on day-1, which decreased toundetectable levels as the infection was resolved withtreatment.

Treatment of hepatocytes with medium conditioned bymacrophages incubated with LPS increased hepcidin mRNAup to 25-fold, but LPS alone caused only a small (2–3-fold)increase (17). Hepcidin was induced by IL-6 but not by IL-1a or TNF-a, suggesting that hepcidin is a type II acute-phaseprotein. Interestingly, Shike et al. (126) showed that in white

298 Cardoso et al.

bass liver, infection with the fish pathogen Streptococcusiniae increased hepcidin mRNA expression (4500-fold).

Lower expression of hepcidin mRNA was observed inmice after 12 days of hypoxia (19) and in human hepatomacell lines (HepG2 and Hep3B) cultured under hypoxia condi-tions (19). In hereditary hemochromatosis patients, hepcidinwas found significantly decreased when compared to controls(116,118). Similar findings were observed in Hfe knockoutmice (115–117). Further, hepcidin mRNAwas found increasedin mice in response to inflammatory stimuli such as LPS (18)and treatment with turpentine (19).

Thus, hepcidin may be a central regulatory molecule forthe development of anemia of inflammation: (i) when regu-lated by cytokines (IL-6), (ii) in modulating iron absorptionby the duodenum, and (iii) possibly by interfering with macro-phage iron recycling.

Ferroportin 1: Ferroportin 1 is downregulated by acuteinflammation in a murine LPS model of the acute-phase reac-tion (13). Mice subjected to LPS treatment downregulatedthe expression of Ferroportin 1 in cells from the reticulo-endothelial system of the spleen, liver, and bone marrowand required signaling through Toll-like receptors (13). Thisobservation was also found in a Leishmania donovani modelof chronic infection (13). In addition, innate immunity, proin-flammatory stimuli, such as IFN-g, TNF-a? and=or LPS,downregulated ferroportin 1 expression in human monocyticcells, THP-1, and U937, and reduced iron release frommonocytes (111). Toll-like receptors play a major role inpathogen recognition and initiation of inflammatory andimmune responses. Activation of toll-like receptors by micro-bial products leads to the activation of signaling pathwaysthat result in the induction of antimicrobial genes and inflam-matory cytokines (reviewed in Ref. 25).

DMT-1 (also called DCT1 or Nramp2): Two independentgroups identified DMT1 in 1997 (10,11). DMT-1 is a highlyhydrophobic integral membrane glycoprotein composed of 12transmembrane domains that possess several structuralcharacteristics of ion channels and transporters. DMT-1mRNA expression is ubiquitinous (10). However, its levels of


expression are relatively high in the brain, thymus, proximalintestine, kidney, and the bone marrow (10). More recently,DMT-1 was found in the apical surface in duodenal entero-cytes (127,128).

DMT-1 (Nramp2) is part of the family of Nramp mole-cules that is a family remarkably conserved throughout evolu-tion with homologs in bacteria, yeast, and plants (129).Nramp1 and Nramp2 proteins have 77% overall similarity.DMT-1 expression in macrophages is induced by cytokinesand paralleled increased uptake of non-Tf bound iron (asdetailed in Sec. 3).

Immune Regulation of Iron Homeostasis: Lessonsfrom Gene Knockout Mouse Models

Interestingly, some of the most significant contributions tounderstand that classical ‘‘immunological molecules’’ partici-pate in the regulation of iron metabolism have derived fromthe discovery of iron overload in gene knockout animal models.A detailed description of the animal models of iron overloadwith immunological abnormalities is done in the next sections.Similar to animal models, immunological abnormalities werealso found in hemochromatosis patients that are not alteredby iron depletion suggesting that they preceded the develop-ment of iron overload.

Iron Overload in Immunological Gene KnockoutMice Models (Table 4 and Refs. 18, 19, 26–28,130–138)

b2-Microglobulin Knockout Mice (b2m�=–)

b2m�=– mice have severely decreased cell surface expres-sion of the MHC-class I molecules and consequently they lackthe CD8þ T population (139). The existence of hepatic ironoverload in b2m�=– was revealed to be similar to that foundin HH: pathologic iron deposits occur predominantly inhepatic parenchymal cells (26,130). The b2m�=– has a 4-foldincrease in plasma iron concentration, increased TfSat withreduced plasma apo-Tf and increased hepatic iron when

300 Cardoso et al.

compared with normal control mice (131,132). Mucosaluptake of ferric iron and the subsequent mucosal transfer intothe plasma is inappropriately increased in b2m�=– (131,132).The characterization of the b2m knockout mice provided sig-nificant support for MHC-class I involvement in iron regula-tion. The iron overload in these mice could be attenuated byreconstitution with foetal liver cells derived from 13-day-oldliver harvested from normal mice of the same strain. A shiftin iron stores from the hepatic parenchyma to Kupffer cellsoccurred within 1month following reconstitution, and adecrease in total iron stores was observed after 3 months.However, TfSat and mucosal iron transfer rate remained highsuggesting that the primary defect of iron overload was notcorrected (132).

Table 4 Iron Overload in Immunological Gene Knockout MiceModels

Animal models Immunological abnormalities References

b2m–=– No cell surface expression of MHC-class I. Lack of CD8þ T cells

26, 130–132

Hfe–=– or C282Y KI Lack (or mutated) of a nonclassicalMHC-class I molecule (Hfe)

27

gd–=– Lack of gd lymphocytes 133–135Reduction of TNF-a production 136

TNFR2�=– Increased susceptibility to infection 137Hfe–=–b2m–=– Lack of classical MHC-class I at

cell surface134

Lack of a nonclassical MHC-class I(Hfe)

Lack of CD8þ T cellsb2m–=–Rag1–=– Lack of mature T- and

B-lymphocytes138

Lack of classical MHC-class I atcell surface

H2Kb–=– Db–=– Lack of classical MHC-class Imolecules

28

USF2–=– Lack of hepcidin, a peptide withantimicrobial activity

18,19

Hp–=–Rag1–=– Lack of mature T andB lymphocytes

186


HFE Knockout Mice (Hfe�=–)and C282Y Knock-in Mice (C282Y KI)

In the case of Hfe�=– mice, the gene was disrupted result-ing in mice being homozygous for the null allele with subse-quent loss of protein function and iron overload similar tothat seen in the HH revealing that Hfe (a nonclassical MHC-class I protein) is the protein defective in HH (27,94,133,134). In this model, mice exhibit abnormally high TfSatand excessive iron accumulation in liver, predominantly onhepatocytes, recapitulating the biochemical abnormalitiesand histopathology of HH (27,133,134). Hfe�=– mice have anincreased intestinal iron absorption (133). Interestingly, theC282Y KI mice have a less severe phenotype than the Hfe�=–

mice (135).

gd Knockout Mice (gd TcR�=–)

The gd TCR�=– mice lack the gd intraepithelial lympho-cytes and after 21 days of iron supplementation diet demon-strated increased iron liver staining as compared to micewith the same genetic background (C57BL=6J), although toa lesser extent than b2m�=– mice (136). In addition, gdTCR�=– mice had a marked reduction of TNF-a productionby intraepithelial lymphocytes when compared to controls(136) suggesting a role for TNF-a in intestinal iron regulation.

Tumor Necrosis Factor Receptor 2 KnockoutMice (TNFR2�=–)

Mice being fed with an iron rich diet and being nullizy-gous for the TNFR2 showed a significant increase in hepaticand splenic iron levels which was not observed in TNFR1�=–

mice (137) while Hfe�=– mice demonstrated only increasediron accumulation in the liver.

HFE and b2-Microglobulin Double KnockoutMice (Hfe�=–b2m�=–)

Mice lacking the Hfe and b2m molecules lack also cellsurface MHC-class I expression and CD8þ T lymphocytes.Liver iron deposition was significantly higher in mice lackingboth Hfe and b2m when compared to mice lacking Hfe alone

302 Cardoso et al.

(134). This observation reinforces the original finding withb2m�=– indicating that MHC-class I molecules and CD8þ Tlymphocytes are important in iron homeostasis.

b2-Microglobulin and RecombinaseActivating Gene 1 Double Knockout Mice(b2m�=–Rag1�=–)

b2m�=–Rag1�=– mice lack mature T- and B-lymphocytesand the MHC-class I expression. These mice accumulateexcess dietary iron in parenchymal cells of the liver. Heavyiron deposition also occurs in other tissues, such as thepancreas and the heart. Older mice (20–28 weeks) under aniron-enriched diet develop fibrosis of the heart and pancreas(138). Heigthed iron overload is also Hp�/� Rag�/� mice (186).

MHC-class I knockout mice(H2Kb–=–Db–=–)

The more decisive evidence for the role of MHC-class Iitself in iron homeostasis comes from a recent study ofH2Kb–=–Db–=– mice (28). These mice have significantly higherhepatic nonheme iron content, predominantly in hepatocytes,when compared to the B6 control mice. Single H2Kb andH2Kd knockout mice have an intermediate iron overloadphenotype, providing additional evidence for the participationof MHC-class I alleles in the process (28).

Upstream Stimulatory Factor 2 KnockoutMice (USF2�=–)

USF2�=– mice lack the hepcidin gene expression (19), arecently identified disulfide-bonded peptide exhibiting anti-microbial activity (18). USF2�=– mice progressively developmultivisceral iron overload with massive iron deposition inliver (hepatocytes), pancreas, and, to a lesser extent, in heart(19). In contrast, the splenic iron content was lower inUSF2�=– mice than controls (19). As described earlier, thisprotein has a fundamental role both in iron metabolism andin inflammatory processes.

More recently, two studies have reported differentdegrees of iron overload in three different mouse strains (with


different genetic backgrounds) in which Hfe (140) or b2m(141) genes were knocked out. Interestingly, those mousestrains have different numbers of lymphocytes (142).

Immunological Abnormalities in HH Patients

Iron overload in immunological gene knockout mice (Table 4)reflects more clearly the importance of the immunological sys-tem in the development of iron overload than the reverse, i.e.,the finding of immunological abnormalities in iron overloadconditions. Findings in HH patients that are not altered byiron removal represent a strong suggestion that the basicfindings in mice apply to some diseases in man. Conversely,some models of iron overload in mice have abnormalities ofthe relative proportions of circulating T cells, indicatingthat iron or iron related proteins and enzymes may influenceT-cell differentiation (187). Poss and Tonegawa (20) studiedmice lacking heme-oxygenase 1 (Hmox1�=–). Hmox1 is themolecule responsible to catabolize cellular heme to biliverdin,carbon monoxide, and free iron. Hmox1�=– mice are anemicand have serum iron values severely reduced by 2 weeks ofage (20). Although Hmox1�=– mice displayed iron deficiency,they also had significantly increased levels of serum ferritin.Organ iron loading was observed in renal cortical tubules,Kupffer cells, hepatocytes, and hepatic vascular tissue (20).These mice also have disturbances of the immunological systemcharacterized by high peripheral white blood cell counts, highlymph node, and splenic CD4:CD8 T-cell ratio with numerousactivated CD4þ T cells (20), suggesting that iron relatedproteins and enzymes may influence T-cell differentiation.

Lymphocyte Populations

Both the finding of abnormalities in the relative proportionsof the two major T lymphocyte subpopulations (CD4þ andCD8þ T cells) in HH patients and the observation of an ironoverload similar to that seen in HH in the b2m knockoutmice lacking MHC-class I and CD8þ T lymphocytespreceded the discovery of the HH gene (reviewed in Ref.41). Reimao et al. (143) have shown in 1991 that patients

304 Cardoso et al.

with high CD4=CD8 ratios (>2.9) after intensive phlebotomytreatment displayed a faster re-entry of iron into the serumtransferrin pool reaching abnormal (>60%) TfSat valuesmore rapidly than patients with normal CD4=CD8 ratios(~1). The former required a longer course of phlebotomytreatment and mobilized significantly more iron than thosewith normal CD4=CD8 ratios (144). The iron mobilized byphlebotomy correlated significantly with CD8þ T cells butnot with CD4þ T cells (144). In addition, a highly significantcorrelation between CD4=CD8 ratios and iron stores wasobserved in HLA-A3 patients, this correlation being mainlyattributable to CD8þ T cells (145). It is of note that through-out serial phlebotomy treatment, each individual manifested agreat stability of its CD4=CD8 ratios (143,144), and in its rela-tive or absolute numbers of CD4þ and CD8þ T-cell popula-tions (145). The lower percentages of CD8þ T cells seen inthe peripheral blood were later on shown to be associated withlow numbers of the same cells in the liver (146). In addition,the low numbers of CD8þ T cells observed in the liver wereassociated with higher levels of hepatic tissue iron (146).

The early immunological abnormalities found in CD8þ Tcells in HH patients included a defective p56lck activity (147),an increased number of CD8þ T cells lacking the costimulatorymolecule CD28 (a finding not related to age) and abnormallyhigh percentages of activated T cells (HLA-DRþ) (148).

T-Cell Receptor Repertoire in HH

Lymphocyte Numbers and TcR Repertoire asMarkers of Clinical Heterogeneity in IronOverload Patients

Low lymphocyte counts were found to be associated withhigher grades of iron overload in hereditary hemochromatosislinked to HFE but not with African iron overload (149). In addi-tion, the frequency of Va=b expansions in 32 HH patients homo-zygous for the C282Ymutation was significantly higher in thosewith iron overload related organ pathology than those withoutpathology (150). These expansions were not related to intensivetreatment anddonot alter the size of theT-cell populations (150).


HFE and the TcR in Humans

To analyze the influence of the HFE mutations, indepen-dently of iron overload, in the shaping of the TcR repertoire,the frequency of expansions was examined in 274 healthy sub-jects genotyped for the C282Y and H63D mutations (150). Toour knowledge, this study constituted the largest study ofTcR repertoire in normal subjects reported to date. No differ-ences were observed in the frequency of individuals with expan-sions among carriers of the HFE mutations and subjectscarriers of only the wild-type allele. As reported by others(151–153), the frequency of expansions among the Va=b chainsstudied were mainly within the CD8þ T-cell subset and wererare or absent within the CD4þ T-cell pool independently ofthe HFE genotype. Interestingly, C282Y carriers had no expan-sions in two particular TcR chains: Vb5.2 and Vb12 within theCD8þ T-cell pool with statistically significant lower levels ofVb12þCD8þ T cells. This finding suggests that the C282Ymutation affects the selection of Vb12þCD8þ T cells (150).

Monocyte=Macrophages

Other abnormalities in the immunological system of HHpatients in particular in the monocyte=macrophage popula-tion were observed. Low TNF-a production by peripheralblood macrophages on stimulation with LPS when comparedwith control subjects (154) was seen. Cairo et al. (155) alsofound a significant increase in IRP activity in monocytes fromuntreated HH patients when compared with control subjects.IRP activity after phlebotomy returned to that observed incontrols (155). Interestingly, subjects with secondary ironoverload with a tissue iron burden similar to HH patientshave an IRP activity significantly decreased suggesting thatthis abnormality is characteristic of the hereditary form ofhemochromatosis (155). As already mentioned, release oferythrocyte-derived iron from purified human monocytesshowed that although HH monocytes phagocytosed less thanhalf the number of erythrocytes taken up by control mono-cytes they released twice as much iron in the form of LMW-Fe complex than controls (77).

306 Cardoso et al.

Cytokine Profile in HH Patients

More recently, Fabio et al. (156) in a study of 17 asympto-matic C282Y homozygous patients found a significantdecrease of the total lymphocyte count, CD4þ T cells,CD8þCD28þ T cells, and NK cells. The reduction of CD28þlymphocytes was inversely related to TfSat. In addition, anincreased ability of T cells to produce all the cytokines studied(IFN-g, TNF, IL-2, IL-4, IL-5, IL-10, IL-13) with a moremarked increase in IL-4, IL-5, and IL-10 production by theCD3þCD8þ subset was found (156).

INTERPLAY BETWEEN IRON METABOLISM ANDCYTOKINE ACTIVITIES

Infections and inflammatory diseases induce iron sequestra-tion in macrophages and also decrease iron absorption inthe small intestine. Decreasing availability of iron by the hostmay deny this essential element to invading pathogens andmay inhibit their multiplication and other biologicalprocesses. In addition, macrophages also require iron as acofactor for the execution of important antimicrobial effectormechanisms.

As iron affects immune cell proliferation, cytokineseither directly or via the formation of radicals by immunecells are able to regulate iron homeostasis both by transcrip-tional and post-transcriptional mechanisms. Maintenance ofcellular iron homeostasis is largely exerted at the post-tran-scriptional=translational level by interaction of cytoplasmaticproteins, the so-called IRP-1 and -2, with RNA stem loopstructures, IRE. IREs have been identified within the mRNAs30 and the 50 untranslated regions. The latter are found inmRNAs coding for the central proteins for iron storage(H-chain and L-chain ferritin), iron consumption (erythroidamino levulinic acid synthase, e-ALAS, the key enzyme inheme-biosynthesis), the iron exporter ferroportin 1, and mito-chondrial aconitase. The mRNA coding for the major ironuptake protein, TfR, bears five IREs within its 30 untranslatedregion (for review, see Refs. 157, 158 and Chapter 1 in this


book). Iron deficiency in cells stimulates the binding affinityof IRPs to IREs thus resulting in blocking of ferritin and e-ALAS expression by affecting the formation of the translationinitiation complex. Conversely, binding of IRPs to the IREswithin the 30 untranslated region of TfR mRNA results inincreased expression of this protein by prolonging TfR mRNAhalf-life. In contrast, iron overload in cells reduces the targetaffinity of IRPs to IREs which then causes de-repression offerritin and e-ALAS translation, while TfR mRNA is degradedwhich in turn results in limitation of TfR mediated ironuptake while iron storage (ferritin synthesis) or ironconsumption (heme synthesis) is induced.

Regulation of Iron Homeostasis by Cytokines

Th-1 and Th-2 cytokines as well as LPS regulate iron homeos-tasis in activated macrophages by affecting this IRP=IREnetwork, nonetheless these immune mediators also stimulateIRP-independent pathways in order to alter the expression ofcritical iron genes.

Tumor necrosis factor-a and IL-1 are very potent indu-cers of ferritin transcription by an as yet not fully elucidatedtranscriptional mechanism (159,160). In addition, IL-1 andIL-6 stimulate ferritin translation by interacting with a con-sensus region within the 50-untranslated region of ferritinmRNA termed as ‘‘acute phase box’’ (161). The importanceof these in vitro observations is underscored by a studydemonstrating that treatment of mice with TNF-a results inthe induction of hyperferritinemia and hypoferremia, a situa-tion which very much resembles the changes of iron homeos-tasis observed in anemia of chronic disease (29,162,163).

Although these proinflammatory cytokines ratherdecrease TfR mediated iron uptake (164–167), their in vivoapplication results in the development of hypoferremia(168). This suggests that either the induction of ferritinsynthesis with subsequent storage of iron is a sufficientdriving force to get iron into cells or that proinflammatorycytokines may stimulate TfR independent iron uptakemechanisms by macrophages, e.g., via uptake of hemoglobin,

308 Cardoso et al.

ferritin, or lactoferrin, via erythrophagocytosis, or by stimu-lating the expression and transport capacity of transmem-brane iron carriers such as DMT-1 (111,169). Alternatively,proinflammatory cytokines also inhibit the export of iron fromreticuloendothelial cells by inhibiting ferroportin 1 expressionleading to iron retention within immune cells (13,111). More-over, the role of the liver protein hepcidin, its regulation bycytokines, and its endocrine effects toward modulation ofiron homeostasis in the gut or the reticuloendothelial systemis becoming an increasingly attractive target for further inves-tigation (185).

The major regulatory Th-1 mediated cytokine IFN-g hasdistinct effects on iron homeostasis. Although IFN-g stimu-lates ferritin transcription, it inhibits H-ferritin translation,which can be referred to activation of IRP binding by the cyto-kine. Activation of IRP-1 binding affinity is due to stimulationof nitric oxide (NO) formation by IFN-g (165,167,170).Moreover, IFN-g is also known to induce the formation ofradicals in monocyte=macrophages, and reactive oxygen spe-cies, such as hydrogen peroxide or superoxide anion that havebeen shown to modulate differently IRP-1 activity (82,171).Thus, IFN-g induced oxygen radical and NO formation willstimulate IRP-1 binding to the ferritin IRE, which thenregulates ferritin translation (172).

However, IFN-g treatment of monocytes also blockstransferrin mediated iron uptake by downregulation of TfRexpression (166,167,173) via negative regulatory signals ofthe cytokine towards TfR transcription and post-transcrip-tional expression. Nonetheless, at the same time, IFN-g andLPS increase the expression of DMT-1 and thus nontransfer-rin mediated uptake of iron into monocytic cells (111,169).Moreover, the retention of iron within monocytes=macrophages is warranted by IFN-g=LPS mediated downre-gulation of ferroportin 1 expression (111).

At least in T cells, another Th-1 cytokine, IL-2, upregu-lates TfR expression transcriptionally and post-transcription-ally which may be a prerequisite for a growth promoting effectof IL-2 towards T cells (174). As outlined above, Nramp1 isexclusively expressed in monocytes and neutrophils and a


putative modulator of iron homeostasis due to its potential totransport iron across the phagolysosmal membrane andbecause of the fact that its expression is modulated by cyto-kines (67,68,71). Although iron handling by macrophages isaffected by knocking out Nramp1 functionality (71), its contri-bution to deregulation of iron traffic under inflammatoryconditions and thus to the pathophysiology of anemia ofchronic disease (ACD) has still to be shown.

Interestingly, anti-inflammatory cytokines derived fromTh-2 cells, such as IL-4, IL-10, and IL-13, counteract theeffects of IFN-g on iron homeostasis in activated macrophages(167). Treatment of macrophages with IL-4 and=or IL-13 priorto stimulation with IFN-g suppresses NO formation and thesubsequent IRP activation, concomitantly enhancing ferritintranslation. Conversely, TfR mRNA levels increase followingpretreatment of IFN-g stimulated macrophages with theanti-inflammatory cytokines. This may be referred to IL-4=IL-13 mediated antagonization of the inhibitory signal,which is induced by IFN-g and inhibits TfR expression byan IRP independent pathway. In contrast, anti-inflammatorycytokines such as IL-10 slightly inhibit IFN-g=LPS mediatedDMT-1 expression and iron accumulation. Moreover, thera-peutic administration of IL-10 to subjects suffering fromCrohn’s disease resulted in the development of hyperferritine-mia and anemia in subjects receiving higher IL-10 dosagesduring this randomized, double blinded, placebo-controlledstudy. All abnormalities of erythropoiesis and iron homeosta-sis returned to normal within 4 weeks after the end of IL-10therapy without any additional intervention. Thus, theobserved changes were linked to IL-10 mediated inductionof ferritin expression in monocytic cells via stimulation ofIRP activity (175).

Thus, Th-2 collaborates with proinflammatory cytokines(TNF-a, IL-1, IL-6, IFN-g) in the development of the typicaldiversion of iron traffic during chronic inflammatory processes(Fig. 2). The induction of hypoferremia=hyperferretinemia issupposed to be a major pathogenetic mechanism contributingto the development of ACD (as outlined elsewhere in thisbook).

310 Cardoso et al.

The negative regulatory interaction between iron andimmunity is evident especially in patients with secondaryiron overload leading to the accumulation of iron within theRES, while in HH, the iron is primarily stored within hepato-cytes. Secondary iron overload patients like those withAfrican siderosis or ‘‘Bantu disease’’ have increased suscept-ibility to infections with intracellular pathogens such astuberculosis that are paralleled by a negative associationbetween iron stores and the Th-1 response (176).

Regulation of Cell Mediated Immune EffectorFunction by Iron

As cytokines influence iron homeostasis, iron interferes withcytokine activities and cell mediated immune effectormechanisms of macrophages, thus altering the immuneresponse toward invading pathogens. One central mechanismbeing responsible for this fact is a direct inhibitory effect or

Figure 2 Pathways for the regulation of macrophage iron acquisi-tion and homeostasis by pro- and anti-inflammatory cytokines.


iron towards the activity of IFN-g. Iron loading of macro-phages results in an inhibition of IFN-g mediated pathwaysin macrophages such as formation of the proinflammatorycytokine TNF-a, expression of MHC-class II antigens, forma-tion of neopterin, a degradation product of GTP which iswidely used as a clinically valuable parameter to monitor cellmediated immune activation in vivo and in vitro, and finally,tryptophan degradation via IFN-g mediated induction ofindole-amine-2, 3-dioxygenase or ICAM-1 expression (48,177).As a consequence of this, iron loaded macrophages lose theirability to kill intracellular pathogens by IFN-gmediated path-ways, such as Legionella, Listeria, Ehrlichia, Candida, andalso viruses, in vitro and in vivo (for review, see Refs. 29,178 and Chapter 9 in this book). Part of this can be attributedto the reduced formation of NO in the presence of iron. Thisis of importance since NO is an essential effector moleculeof macrophages to fight infectious pathogens and tumorcells (179,180). Iron blocks the transcription of inducible NO-synthase (iNOS or NOSII), the enzyme being responsible forcytokine inducible high-output formation of NO by hepatocytesor macrophages (181,182). By inhibiting the binding affinity ofNF-IL-6 and of hypoxia inducible factor 1 to the iNOS promoteriron impairs iNOS transcription and reduces its inducibility bycytokines (182,183). Therefore, iron overload stands at the inter-face between a consequence of defects and at the same time, acause of impaired cellular immune effector pathways. Iron over-load may be detrimental for fighting infectious and=or malig-nant diseases as outlined in the chapters on ‘‘putativenegative effects of ACD correction’’ and ‘‘iron withholding as adefense strategy’’ in this book. In addition, as detailed in the sec-tion on Conclusion, iron can be viewed as an impelling forcebetween innate and adaptive immunity.

CONCLUSION

Iron as an Impelling Force Between Two Kinds ofImmunity?

Recent advances in immunology have led to the acknowledge-ment that the protective capacity of the immune system from

312 Cardoso et al.

the threat of pathogens and malignant cells is far wider thanthat conferred exclusively by peptide presentation to T cells byantigen presenting cells or by specific antibodies by B cells (25).

Iron and Innate Immunity

Iron withholding from bacteria by the host had been thoughtfor a long time to be an important defense mechanism (5). Butit is with the studies reviewed in this chapter of the responseof iron genes to inflammatory stimuli that the control of iron‘‘commerce’’ by the macrophage becomes an integral part ofinnate immunity (19,25). In response to LPS, the phagocyticsystem through IL-6 controls hepcidin synthesis and release(17). Upregulation of ferroportin 1 in response to LPS orLPS=IFNg (111) also determines iron holding by the macro-phage. Other known components of innate immunity target-ing iron availability to pathogens include neutrophilproducts, namely lactoferrin (5) and NGAL (87).

Iron ‘‘at Peace’’

Hans Wigzell is using the expression ‘‘at peace’’ as part of thedefinition of self, in the context of discrimination between selfand nonself (Wigzell, personal communication). In physiologi-cal conditions, it is that same phagocytic component of theimmune system that secures ‘‘the peace’’ that comes from sta-bility of iron recycling in higher vertebrates, rendering anorganism relatively independent of variations in iron avail-ability in the external environment (21). The recognition ofsenescent red blood cells and the reutilization of the irontherein for erythropoiesis by macrophages (Fig. 1) are one ofthe most successful examples of recycling of an essentialnutrient in a complex biological system.

Iron and Adaptive Immunity

It is, however, with the identification of HFE as an MHC-classI related gene and the discovery of iron overload in miceknockout for b2m (26), double knockout for b2m and Rag 1(138), b2m and Hfe (134), and the hepatic overload seen in


H2k-mice (28) that it becomes evident that molecules thus farconsidered central to the evolution of adaptive immunity arealso key molecules of iron metabolism.

In addition, macrophage supernatants after erythropha-gocytosis are costimulatory of T lymphocyte activation viaCD3 (77), thymus cell differentiation is altered in hpxþ=–mice (187), H-ferritin has an inhibitory effect on CD3 activa-tion (184).

Caveat

The exact mechanisms whereby immune genes and gene pro-ducts contribute to iron recycling ‘‘at peace,’’ iron deficiencyanemia ‘‘at war’’ with pathogens and iron overload in thepresence of defective immune genes are presently not comple-tely clear (29).

Meanwhile, it is undeniable that the immune system hasas one of its key functions the control of iron metabolism.Whether ‘‘peace’’ was the first achievement in evolution andonly after recycling of red blood cells was established, macro-phages acquired the control of iron availability for pathogenscannot be decided at present. The reverse, in our view, seemsequally or even more plausible. Phagocytes could have‘‘started’’ by withholding iron from bacteria and only later,complex organisms may have benefited from incorporatingin their physiology that primary capacity, extending it to acontrolled release of iron for erythropoiesis. Furthermore, itwould have become greatly economical to adapt the use ofsome of the products of that central function as nutrients orcostimulatory, or regulatory, molecules of lymphocyte differ-entiation, and activation in adaptive immunity.

In conclusion, this chapter is written at a time very dif-ferent from the time when such a function for the immunesystem could only belong in the realm of hypotheses. Thefurther dissection of the mechanisms involved dependsgreatly on technology advances in real time microscopy. Thefiner definition of mechanisms is likely to take us to a bettercontrol of iron deficiency and, very likely to improvementsin the treatment of infection in the face of the growing

314 Cardoso et al.

resistance of bacteria to antibiotics, and of viruses to IFN.Perhaps results indicating the successful use of lactoferrinin viral (6) and mycobacterial infections (63) represent a pio-neering approach to the therapy of infection. Pioneering, asall pioneers know only too well, will have to wait: hopefullyperhaps not too long.

ACKNOWLEDGEMENTS

Carla S. Cardoso is the recipient of an Innova FoundationFellowship. Work by the authors supported by grants fromthe FCT, the Calouste Gulbenkian Foundation, the EU andthe Innova Foundation.

REFERENCES

1. de Sousa M. Lymphoid cell positioning: a new proposal for themechanism of control of lymphoid cell migration. Soc Exp BiolSymp 1978; 32:393–410.

2. Dorner MH, Silverstone AE, Nishiya K, de Sostoa A, Munn G,de Sousa M. Ferritin synthesis by human T-lymphocytes.Science 1980; 209:1019–1021.

3. Nishiya K, Chiao JW, de Sousa M. Iron binding proteins inselected human peripheral blood sets: immunofluorescence.Br J Haematol 1980; 46:235–245.

4. de Sousa M, Martins da Silva B, Dorner M, Munn G, NishiyaK, Grady RW, Silverstone A. Iron and the lymphomyeloidsystem: rationale for considering iron as a target of ‘‘immunesurveillance’’. In: Saltman P, Hegenauer J, eds. The Biochem-istry and Physiology of Iron. Elsevier North Holland Inc,1982:687–698.


6. van der Strate BW, Beljaars L, Molema G, Harmsen MC,Meijer DK. Antiviral activities of lactoferrin. Antiviral Res2001; 52:225–239.


7. Vezzoni P, Levi S, Gabri E, Pozzi MR, Spinazze S, Arosio P.Ferritins in malignant and non-malignant lymphoid cells.Br J Haematol 1986; 62:105–110.

8. Neckers LM, Cossman J. Transferrin receptor induction inmitogen-stimulated human T lymphocytes is required forDNA synthesis and cell division and is regulated byinterleukin 2. Proc Natl Acad Sci USA 1983; 80:3494–3498.

9. Vidal SM, Malo D, Vogan K, Skamene E, Gros P. Naturalresistance to infections with intracellular parasites: isolationof a candidate for Bcg. Cell 1993; 73:469–476.

10. Gunshin H, Mackenzie B, Berger UV, Gunshin Y, RomeroMF, Boron WF, Nussberger S, Gollan JL, Hediger MA. Clon-ing and characterization of a mammalian proton-coupledmetal-ion transporter. Nature 1997; 388:482–488.

11. Fleming MD, Trenor CC III, Su MA, Foernzler D, Beier DR,Dietrich WF, Andrews NC. Microcytic anaemia mice have amutation in Nramp2, a candidate iron transporter gene.Nat Genet 1997; 16:383–386.



14. Bjorkman PJ, Parham P. Structure, function, and diversity ofclass I major histocompatibility complex molecules. Annu RevBiochem 1990; 59:253–288.

15. Gobin SJ, Biesta P, Van Den Elsen PJ. Regulation of humanbeta 2-microglobulin transactivation in hematopoietic cells.Blood 2003; 101:3058–3064.

16. Yang F, Naylor SL, Lum JB, Cutshaw S, McCombs JL,Naberhaus KH, McGill JR, Adrian GS, Moore CM, BarnettDR, et al. Characterization, mapping, and expression of thehuman ceruloplasmin gene. Proc Natl Acad Sci USA 1986;83:3257–3261.

316 Cardoso et al.

17. Nemeth E, Valore EV, Territo M, Schiller G, Lichtenstein A,Ganz T. Hepcidin, a putative mediator of anemia ofinflammation, is a type II acute-phase protein. Blood 2003;101:2461–2463.

18. Pigeon C, Ilyin G, Courselaud B, Leroyer P, Turlin B, BrissotP, Loreal O. A new mouse liver-specific gene, encoding aprotein homologous to human antimicrobial peptide hepcidin,is overexpressed during iron overload. J Biol Chem 2001;276:7811–7819.

19. Nicolas G, Chauvet C, Viatte L, Danan JL, Bigard X,Devaux I, Beaumont C, Kahn A, Vaulont S. The gene encod-ing the iron regulatory peptide hepcidin is regulated by ane-mia, hypoxia, and inflammation. J Clin Invest 2002; 110:1037–1044.

20. Poss KD, Tonegawa S. Heme oxygenase is required formammalian iron reutilization. Proc Natl Acad Sci USA1997; 94:10919–10924.

21. Cardoso CS, de Sousa M. HFE, the MHC and hemochromato-sis. Paradigm for an extended function for MHC class I.Tissue Antigens 2003; 61:263–275.

22. Barnewall RE, Ohashi N, Rikihisa Y. Ehrlichia chaffeensisand E. sennetsu, but not the human granulocytic ehrlichiosisagent, colocalize with transferrin receptor and up-regulatetransferrin receptor mRNA by activating iron-responsiveprotein 1. Infect Immun 1999; 67:2258–2265.

23. Doherty PC, Christensen JP. Accessing complexity: the dyna-mics of virus-specific T cell responses. Annu Rev Immunol2000; 18:561–592.

24. Andrews NC. Iron homeostasis: insights from genetics andanimal models. Nat Rev Genet 2000; 1:208–217.

25. Janeway CA Jr, Medzhitov R. Innate immune recognition.Ann Rev Immunol 2002; 20:197–216.

26. de Sousa M, Reimao R, Lacerda R, Hugo P, Kaufmann SH,Porto G. Iron overload in beta2-microglobulin-deficient mice.Immunol Lett 1994; 39:105–111.

27. Zhou XY, Tomatsu S, Fleming RE, Parkkila S, Waheed A,Jiang J, Fei Y, Brunt EM, Ruddy DA, Prass CE, Schatzman


RC, O’Neill R, Britton RS, Bacon BR, Sly WS. HFE geneknockout produces mouse model of hereditary hemochroma-tosis. Proc Natl Acad Sci USA 1998; 95:2492–2497.

28. Cardoso EM, Macedo MG, Rohrlich P, Ribeiro E, Silva MT,Lemonnier FA, de SousaM. Increased hepatic iron inmice lack-ing classicalMHCclass Imolecules. Blood 2002; 100:4239–4241.


30. Jacobs A, Worwood M. Iron in Biochemistry and Medicine.London and New York: Academic Press, 1974:764.

31. Watson JD, Crick FHC. Molecular structure of nucleic acid.Nature 1953; 171:737–738.

32. Koller BH, Smithies O. Altering genes in animals by genetargeting. Ann Rev Immunol 1992; 10:705–730.

33. Ojcius DM, Delarbre C, Kourilsky P, Gachelin G. MHC andMHC-related proteins as pleiotropic signal molecules. FASEBJ 2002; 16:202–206.

34. de Sousa M. Immune cell functions in iron overload. Clin ExpImmunol 1989; 75:1–6.

35. Ponka P, Lok CN. The transferrin receptor: role in health anddisease. Int J Biochem Cell Biol 1999; 31:1111–1137.

36. Conrad ME, Umbreit JN. Iron absorption and transport-anupdate. Am J Hematol 2000; 64:287–298.

37. Conrad ME, Umbreit JN, Moore EG, Parmley RT. Intestinaliron absorption and hemochromatosis. In: Barton JC,Edwards CQ, eds. Haemochromatosis: Genetics, Pathophy-siology, Diagnosis and Treatment. Cambridge: CambridgeUniversity Press, 2000:118–130.

38. Wood RJ, Han O. Recently identified molecular aspects ofintestinal iron absorption. J Nutr 1998; 128:1841–1844.

39. Marx JJ. Mucosal uptake, mucosal transfer and retention ofiron, measured by whole-body counting. Scand J Haematol1979; 23:293–302.

40. Kuhn LC. Molecular regulation of iron proteins. BaillieresClin Haematol 1994; 7:763–785.

318 Cardoso et al.

41. de Sousa M, Cruz E, Costa L, Santos M. A role for the immu-nological system in iron homeostasis: ‘‘But how?’’. In:Ferreira GC, Moura JJG, Franco R, eds. Iron Metabolism:Inorganic Biochemistry and Regulatory Mechanisms. Wiley-VCH Verlag GmbH, 1999:199–210.

42. Seligman PA, Kovar J, Gelfand EW. Lymphocyte prolifera-tion is controlled both by iron availability and regulation ofiron uptake pathways. Pathobiology 1992; 60:19–26.

43. Mainou-Fowler T, Brock JH. Effect of iron deficiency on theresponse of mouse lymphocytes to concanavalin A: impor-tance of transferrin bound iron. Immunology 1985; 54:325–332.

44. Keyna U, Nusslein I, Rohwer P, Kalden JR, Manger B. Therole of the transferrin receptor for the activation of humanlymphocytes. Cell Immunol 1991; 132:411–422.

45. Brekelmans P, van Soest P, Leenen PJ, van Ewijk W. Inhibi-tion of proliferation and differentiation during early T celldevelopment by anti-transferrin receptor antibody. Eur JImmunol 1994; 24:2896–2902.

46. de Sousa M, Reimao R, Porto G, Grady RW, Hilgartner MW,Giardina P. Iron and lymphocytes: reciprocal regulatoryinteractions. Curr Stud Hematol Blood Transf 1992; 58:171–177.

47. Thorson JA, Smith KM, Gomez F, Naumann PW, Kemp JD.Role of iron in T cell activation: Th-1 clones differ from Th-2clones in their sensitivity to inhibition for DNA synthesiscaused by IGG MAbs against transferrin receptor and theiron chelator desferrioxamine. Cell Immunol 1991; 134:126–127.

48. Weiss G, Fuchs D, Hausen A, Reibnegger G, Werner ER,Werner-Felmayer G, Wachter H. Iron modulates interferon-g effects in the human myelomonocytic cell line THP-1. ExpHematol 1992; 20:605–610.

49. Romagnani S. The Th1=Th2 paradigm. Immunol Today 1997;18:263–267.

50. Brock JH. Iron in infection, immunity, inflammation and neo-plasia. In: Brock JH, Halliday JW, Pippard MJ, Powell LW,


eds. Iron Metabolism in Health and Disease. Philadelphia:Saunders Company Ltd, 1994:353–391.

51. Bomford A, Young S, Williams R. Intracellular forms of ironduring transferrin iron uptake by mitogen stimulated humanlymphocytes. Br J Haematol 1986; 62:487–494.

52. Djeha A, Brock JH. Uptake and intracellular handling of ironfrom transferrin and iron chelates by mitogen stimulatedmouse lymphocytes. Biochim Biophys Acta 1992; 1133:147–152.

53. Salmeron A, Borroto A, Fresno M, Crumpton MJ, Ley SC,Alarcon B. Transferrin receptor induces tyrosine phosphory-lation in T cells and is physically associated with the TCRzeta-chain. J Immunol 1995; 154:1675–1683.

54. Seligman PA, Kovar J, Schleicher RB, Gelfand EW. Transfer-rin-independent iron uptake supports B-lymphocyte growth.Blood 1991; 78:1526–1531.

55. Hamazaki S, Glass J. Non-transferrin dependent 59Fe uptakein phytohemagglutinin-stimulated human peripheral lym-phocytes. Exp Hematol 1992; 20:436–441.

56. Konijn AM, Meyron-Holtz EG, Levy R, Ben-Bassat H,Matzner Y. Specific binding of placental acidic isoferritinsto cells of the T-cell line HD-MAR. FEBS Lett 1990;263:229–234.

57. Fargion S, Fracanzani AL, Brando B, Arosio P, Levi S,Fiorelli G. Specific binding sites for H-ferritin on humanlymphocytes: modulation during cellular proliferation andpotential implication in cell growth control. Blood 1991; 78:1056–1061.

58. Moss D, Powell LW, Arosio P, Halliday LW. Effect of cellproliferation on H-ferritin receptor expression in human Tlymphoid (MOLT-4) cells. J Lab Clin Med 1992; 120:239–243.

59. Brock JH. Lactoferrin. A multifunctional immunoregulatoryprotein? Immunol Today 1995; 16:417–419.

60. Mincheva-Nilsson L, Hammarstrom S, Hammarstrom ML.Activated human gamma delta T-lymphocytes express func-tional lactoferrin receptors. Scand J Immunol 1997; 46:609–618.

320 Cardoso et al.

61. Haversen L, Ohlsson BG, Hahn-Zoric M, Hanson LA,Mattsby-Baltzer I. Lactoferrin down-regulates the LPS-inducedcytokine production in monocytic cells via NF-kappa B. CellImmunol 2002; 220:83–95.

62. He S, McEuen AR, Blewett SA, Li P, Buckley MG, LeufkensP, Walls AF. The inhibition of mast cell activation by neutro-phil lactoferrin: uptake by mast cells and interaction withtryptase, chymase and cathepsin G. Biochem Pharmacol2003; 65:1007–1015.

63. Schaible UE, Collins HL, Priem F, Kaufmann SH. Correctionof the iron overload defect in beta-2-microglobulin knockoutmice by lactoferrin abolishes their increased susceptibilityto tuberculosis. J Exp Med 2002; 196:1507–1513.

64. Hamilton TA, Gray PW, Adams DO. Expression of the trans-ferrin receptor on murine peritoneal macrophages in differ-ent stags of functional activation. J Immunol 1984;132:2285–2290.

65. Rama R, Sanchez J. Transferrin uptake by bone marrowmacrophages is independent of the degree of iron saturation.Br J Haematol 1992; 82:455–459.

66. Olakamni O, Stokes JB, Britigan BE. Acquisition of ironbound to low molecular weight chelates by human monocytederived macrophages. J Immunol 1994; 153:2691–2703.

67. Blackwell JM, Searle S, Goswami T, Miller EN. Understand-ing the multiple functions of NRAMP1. Microbes Infect 2000;2:317–321.

68. Forbes JR, Gros P. Divalent metal transport by NRAMPproteins at the interface to host-pathogen interaction. TrendsMicrobiol 2001; 9:397–403.

69. Atkinson PG, Barton CH. High level expression ofNRAMP1G169 in RAW264.7 cell transfectants: analysisof intracellular iron transport. Immunology 1999; 96:656–662.

70. Kuhn DE, Baker BD, Lafuse WP, Zwilling BS. Differentialiron transport into phagosomes isolated from the RAW264.7macrophage cell lines transfected with NRAMP-1 Gly169 orNRAMP1Asp169. J Leukoc Biol 1999; 66:113–119.


71. Mulero V, Searle S, Blackwell JM, Brock JH. Solute carrier11a1 (Slc11a1; formerly Nramp1) regulates metabolism andrelease of iron acquired by phagocytic, but not transferrin-receptor-mediated, iron uptake. Biochem J 2002; 363:89–94.

72. Ismail M, Brock JH. Binding of lactoferrin and transferrin tothe human promoncytic cell line U937. J Biol Chem 1993;268:21618–21625.

73. Britigan BE, Serody JS, Hayek MB, Charniga LM, Cohen MS.Uptake of lactoferrin bymononuclear phagocytes inhibits theirability to form hydroxyl radical and protects them from mem-brane peroxidation. J Immunol 1991; 147:4271–4277.

74. Lima MF, Kierszenbaum F. Lactoferrin effects of phagocyticcell function. The presence of iron is required for the lactofer-rin molecule to stimulate intracellular killing by macro-phages but not to enhance the uptake of particles andmicroorganisms. J Immunol 1987; 139:1647–1651.

75. Custer G, Balcerzak S, Rinehart J. Human macrophagehemoglobin-iron metabolism in vitro. Am J Hematol 1982;13:23–36.

76. Kitagawa S, Yuo A, Yagisawa M, Azuma E, Yoshida M,Furukawa Y, Takahashi M, Masuyama J, Takaku F. Activa-tion of human monocyte functions by tumor necrosis factor.Rapid priming for enhanced release of superoxide and ery-throphagocytosis, but no direct triggering of superoxiderelease. Exp Hematol 1996; 24:559–567.

77. Moura E, Noordermeer MA, Verhoeven N, Verheul AFM,Marx JJM. Iron release from human monocytes aftererythrofagocytosis in vitro: an investigation in normal sub-jects and hereditary hemochromatosis patients. Blood 1998;92:2511–2519.

78. Costa LM, Moura EM, Moura JJ, de Sousa M. Ironcompounds after erythrophagocytosis: chemical characteriza-tion and immunomodulatory effects. Biochem Biophys ResCommun 1998; 247:159–165.

79. Kanayasu-Toyoda T, Yamaguchi T, Uchida E, Hayakawa T.Commitment of neutrophilic differentiation and proliferationof HL-60 cells coincides with expression of transferrinreceptor. Effect of granulocyte colony stimulating factor on

322 Cardoso et al.

differentiation and proliferation. J Biol Chem 1999; 274:25471–25480.

80. Brieland JK, Fantone JC. Ferrous iron release from transfer-rin by human neutrophil-derived superoxide anion: effect ofpH and iron saturation. Arch Biochem Biophys 1991;284:78–83.

81. Clark RA, Pearson DW. Inactivation of transferrin iron bind-ing capacity by the neutrophil myeloperoxidase system. J BiolChem 1989; 264:9240–9247.

82. Pantopoulos K, Hentze MW. Rapid responses to oxidativestress mediated by iron regulatory protein. EMBO J 1995;14:2917–2924.

83. Esaguy N, Aguas AP, Silva MT. High-resolution localizationof lactoferrin in human neutrophils: labeling of secondarygranules and cell heterogeneity. J Leuk Biol 1989; 46:51–62.

84. Singh PK, Parsek MR, Greenberg EP, Welsh MJ. A compo-nent of innate immunity prevents bacterial biofilm develop-ment. Nature 2002; 417:552–555.

85. Ward PP, Mendoza-Meneses M, Cunningham GA, ConneelyOM. Iron status in mice carrying a targeted disruption oflactoferrin. Mol Cell Biol 2003; 23:178–185.

86. Crouch SPM, Slater KJ, Fletcher J. Regulation of cytokinerelease from mononuclear cells by iron binding protein lacto-ferrin. Blood 1992; 80:235–240.

87. Goetz DH, Holmes MA, Borregaard N, Bluhm ME,Raymond KN, Strong RK. The neutrophil lipocalin NGAL isa bacteriostatic agent that interferes with siderophore-mediated iron acquisition. Mol Cell 2002; 10:1033–1043.

88. Kjeldsen L, Cowland JB, Borregaard N. Human neutrophilgelatinase-associated lipocalin and homologous proteins inrat and mouse. Biochim Biophys Acta 2000; 1482:272–283.

89. Devireddy LR, Teodoro JG, Richard FA, Green MR. Inductionof apoptosis by a secreted lipocalin that is transcriptionallyregulated by IL-3 deprivation. Science 2001; 293:829–834.

90. Donovan A, Brownlie A, Zhou Y, Shepard J, Pratt SJ,Moynihan J, Paw BH, Drejer A, Barut B, Zapata A, Law TC,


Brugnara C, Lux SE, Pinkus GS, Pinkus JL, Kingsley PD,Palis J, Fleming MD, Andrews NC, Zon LI. Positional cloningof zebrafish ferroportin1 identifies a conserved vertebrate ironexporter. Nature 2000; 403:776–781.

91. McKie AT, Marciani P, Rolfs A, Brennan K, Wehr K, BarrowD, Miret S, Bomford A, Peters TJ, Farzaneh F, Hediger MA,Hentze MW, Simpson RJ. A novel duodenal iron-regulatedtransporter, IREG1, implicated in the basolateral transferof iron to the circulation. Mol Cell 2000; 5:299–309.

92. Njajou OT, Vaessen N, Joosse M, Berghuis B, van DongenJWF, Breuning MH, Snijders PJLM, Rutten WPF, SandkuijlLA, Oostra BA, van Duijn CM, Heutink P. A mutation inSLC11A3 is associated with autosomal dominant hemochro-matosis. Nat Genet 2001; 28:213–214.

93. Montosi G, Donovan A, Totaro A, Garuti C, Pignatti E,Cassanelli S, Trenor CC, Gasparini P, Andrews NC,Pietrangelo A. Autosomal-dominant hemochromatosis isassociated with a mutation in the ferroportin (SLC11A3)gene. J Clin Invest 2001; 108:619–623.

94. Feder JN, Gnirke A, Thomas W, Tsuchihashi Z, Ruddy DA,Basava A, Dormishian F, Domingo Jr R, Ellis MC, FullanA, Hinton LM, Jones NL, Kimmel BE, Kronmal GS, LauerP, Lee VK, Loeb DB, Mapa FA, McClelland E, Meyer NC,Mintier GA, Moeller N, Moore T, Morikang E, Wolff RK. Anovel MHC class I like gene is mutated in patientswith hereditary haemochromatosis. Nat Genet 1996; 13:399–406.

95. Feder JN, Tsuchihashi Z, Irrinki A, Lee V, Mapa FA,Morikang E, Prass CE, Starnes SM, Wolff RK, Parkkila S,Sly WS, Schatzman RC. The hemochromatosis founder muta-tion in HLA-H disrupts b2-microglobulin interaction and cellsurface expression. J Biol Chem 1997; 272:14026–14028.

96. Waheed A, Parkkila S, Zhou XY, Tomatsu S, Tsuchihashi Z,Feder JN, Schatzman RC, Britton RS, Bacon BR, Sly WS.Hereditary hemochromatosis: effects of C282Y and H63Dmutations on association with beta2-microglobulin, intracel-lular processing, and cell surface expression of the HFEprotein in COS-7 cells. Proc Natl Acad Sci USA 1997;94:12384–12389.

324 Cardoso et al.

97. Porto G, de Sousa M. Variation of hemochromatosis preva-lence and genotype in national groups. In: Barton JC,Edwards CQ, eds. Haemochromatosis: Genetics, Pathophy-siology, Diagnosis and Treatment. Cambridge: CambridgeUniversity Press, 2000:51–62.

98. Edwards CQ, Dadone MM, Skolnick MH, Kushner JP. Her-editary haemochromatosis. Clin Haematol 1982; 11:411–345.

99. Parkkila S, Waheed A, Britton RS, Bacon BR, Zhou XY,Tomatsu S, Fleming RE, Sly WS. Association of the transfer-rin receptor in human placenta with HFE, the protein defec-tive in hereditary hemochromatosis. Proc Natl Acad Sci USA1997; 94:13198–13202.

100. Feder JN, Penny DM, Irrinki A, Lee VK, Lebron JA, WatsonN, Tsuchihashi Z, Sigal E, Bjorkman PJ, Schatzman RC. Thehemochromatosis gene product complexes with the transfer-rin receptor and lowers its affinity for ligand binding. ProcNatl Acad Sci USA 1998; 17:1472–1477.

101. Lebron JA, Bennett MJ, Vaughn DE, Chirino AJ, Snow PM,Mintier GA, Feder JN, Bjorkman PJ. Crystal structure of thehemochromatosis protein HFE and characterization of itsinteraction with transferrin receptor. Cell 1998; 93:111–123.

102. Gross CN, Irrinki A, Feder JN, Enns CA. Co-trafficking ofHFE, a nonclassical major histocompatibility complex classI protein, with the transferrin receptor implies a role in intra-cellular iron regulation. J Biol Chem 1998; 273:22068–22074.

103. Lebron JA, Bjorkman PJ. The transferrin receptor-bindingsite on HFE, the class I MHC-related protein mutated inhereditary hemochromatosis. JMol Biol 1999; 289:1109–1118.

104. Salter-Cid L, Brunmark A, Li Y, Leturcq D, Peterson PA,Jackson MR, Yang Y. Transferrin receptor is negativelymodulated by the hemochromatosis protein HFE: Implica-tions for cellular iron homeostasis. Proc Natl Acad Sci USA1999; 96:5434–5439.

105. Waheed A, Parkkila S, Saarnio J, Fleming RE, Zhou XY,Tomatsu S, Britton RS, Bacon BR, Sly WS. Association ofHFE protein with transferrin receptor in crypt enterocytesof human duodenum. Proc Natl Acad Sci USA 1999; 96:1579–1584.


106. Montosi G, Paglia P, Garuti C, Guzman CA, Bastin JM,Colombo MP, Pietrangelo A. Wild-type HFE proteinnormalizes transferrin iron accumulation in macrophagesfrom subjects with hereditary hemochromatosis. Blood2000; 96:1125–1129.


108. Konijn AM, Hershko C. The anaemia of inflammation andchronic disease. In: de Sousa M, Brock JH, eds. Iron in Immu-nity, Cancer and Inflammation. England: J Wiley and SonsLtd, 1989:111–144.

109. Weiss G. Pathophysiology and treatment of anemia of chronicdisease. Blood Rev 2002; 16:87–96.

110. Roy CN, Weinstein DA, Andrews NC. E. Mead JohnsonAward for research in pediatrics lecture: the molecular biol-ogy of the anemia of chronic disease: a hypothesis. PediatrRes 2003; 53:507–512.

111. Ludwiczek S, Aigner E, Theurl I, Weiss G. Cytokine mediatedregulation of iron transport in human monocytic cells. Blood2003; Jan 9 [epublished ahead of print].

112. Jabado N, Canonne-Hergaux F, Gruenheid S, Picard V, GrosP. Iron transporter Nramp2=DMT-1 is associated with themembrane of phagosomes in macrophages and Sertoli cells.Blood 2002; 100:2617–2622.

113. Govoni G, Vidal S, Gauthier S, Skamene E, Malo D, Gros P.The Bcg=Ity=Lsh locus: genetic transfer of resistance to infec-tions in C57BL=6J mice transgenic for the Nramp1 Gly169allele. Infect Immun 1996; 64:2923–2929.

114. Vidal S, Belouchi AM, Cellier M, Beatty B, Gros P. Cloningand characterization of a second human NRAMP gene onchromosome 12q13. Mamm Genome 1995; 6:224–230.


326 Cardoso et al.

116. Bridle KR, Frazer DM, Wilkins SJ, Dixon JL, Purdie DM,Crawford DH, Subramaniam VN, Powell LW, Anderson GJ,Ram GA. Disrupted hepcidin regulation in HFE-associatedhaemochromatosis and the liver as a regulator of body ironhomoeostasis. Lancet 2003; 361:669–673.

117. Muckenthaler M, Roy CN, Custodio AO, Minana B, DeGraafJ, Montross LK, Andrews NC, Hentze MW. Regulatorydefects in liver and intestine implicate abnormal hepcidinand Cybrd1 expression in mouse hemochromatosis. NatGenet 2003; Apr 21 [epublished ahead of print].

118. Gehrke SG, Kulaksiz H, Herrmann T, Riedel HD, Bents K,Veltkamp C, Stremmel W. Expression of hepcidin in heredi-tary hemochromatosis: evidence for a regulation in responseto serum transferrin saturation and non-transferrin-boundiron. Blood 2003; Mar 13 [epublished ahead of print].


120. Weinstein DA, Roy CN, Fleming MD, Loda MF, Wolfsdorf JI,Andrews NC. Inappropriate expression of hepcidin is asso-ciated with iron refractory anemia: implications for theanemia of chronic disease. Blood 2002; 100:3776–3781.


122. Krause A, Neitz S, Magert HJ, Schulz A, Forssmann WG,Schulz-Knappe P, Adermann K. LEAP-1, a novel highlydisulfide-bonded human peptide, exhibits antimicrobialactivity. FEBS Lett 2000; 480:147–150.


124. Nicolas G, Bennoun M, Porteu A, Mativet S, Beaumont C,Grandchamp B, Sirito M, Sawadogo M, Kahn A, Vaulont S.


Severe iron deficiency anemia in transgenic mice expressingliver hepcidin. Proc Natl Acad Sci USA 2002; 99:4596–4601.

125. Nicolas G, Viatte L, Lou DQ, Bennoun M, Beaumont C, KahnA, Andrews NC, Vaulont S. Constitutive hepcidin expressionprevents iron overload in a mouse model of hemochromatosis.Nat Genet 2003; Apr 21 [epub ahead of print].

126. Shike H, Lauth X, WestermanME, Ostland VE, Carlberg JM,Van Olst JC, Shimizu C, Bulet P, Burns JC. Bass hepcidin isa novel antimicrobial peptide induced by bacterial challenge.Eur J Biochem 2002; 269:2232–2237.

127. Canonne-Hergaux F, Gruenheid S, Ponka P, Gros P. Cellularand subcellular localization of the Nramp2 iron transporterin the intestinal brush border and regulation by dietary iron.Blood 1999; 93:4406–4417.

128. Griffiths WJ, Kelly AL, Smith SJ, Cox TM. Localization ofiron transport and regulatory proteins in human cells. QJM2000; 93:575–587.

129. Cellier M, Prive G, Belouchi A, Kwan T, Rodrigues V, ChiaW, Gros P. Nramp defines a family of membrane proteins.Proc Natl Acad Sci USA 1995; 92:10089–10093.

130. Rothenberg BE, Voland JR. Beta2 knockout mice developparenchymal iron overload: a putative role for class I genesof the major histocompatibility complex in iron metabolism.Proc Natl Acad Sci USA 1996; 93:1529–1534.

131. Santos M, Schilham MW, Rademakers LH, Marx JJ,de Sousa M, Clevers H. Defective iron homeostasis inbeta2-microglobulin knockout mice recapitulates hereditaryhemochromatosis in man. J Exp Med 1996; 184:1975–1985.

132. Santos M, Clevers H, de Sousa M, Marx JJ. Adaptiveresponse of iron absorption to anemia, increased erythropoi-esis, iron deficiency, and iron loading in beta2-microglobulinknockout mice. Blood 1998; 91:3059–3065.

133. Bahram S, Gilfillan S, Kuhn LC, Moret R, Schulze JB,Lebeau A, Schumann K. Experimental hemochromatosisdue to MHC class I HFE deficiency: immune status andiron metabolism. Proc Natl Acad Sci USA 1999; 96:13312–13317.

328 Cardoso et al.

134. Levy JE, Montross LK, Andrews NC. Genes that modify thehemochromatosis phenotype in mice. J Clin Invest 2000;105:1209–1216.

135. Levy JE, Montross LK, Cohen DE, Fleming MD, AndrewsNC. The C282Y mutation causing hereditary hemochromato-sis does not produce a null allele. Blood 1999; 94:9–11.

136. Ten-Elshof AE, Brittenham GM, Chorney KA, Page MJ,Gerhard G, Cable EE, Chorney MJ. gd intraepithelial lym-phocytes drive tumor necrosis factor-a responsiveness tointestinal iron challenge: relevance to hemochromatosis.Immunol Rev 1999; 167:223–232.

137. Meyer PN, Gerhard GS, Yoshida Y, Yoshida M, Chorney KA,Beard J, Kauffman EJ, Weiss G, Chorney MJ. Hemochroma-tosis protein (HFE) and tumor necrosis factor receptor 2(TNFR2) influence tissue iron levels: elements of a commongut pathway?. Blood Cells Mol Dis 2002; 29:274–285.

138. Santos MM, de Sousa M, Rademakers LH, Clevers H, MarxJJ, Schilham MW. Iron overload and heart fibrosis in micedeficient for both beta2-microglobulin and Rag1. Am J Pathol2000; 157:1883–1892.

139. Koller BH, Marrack P, Kappler JW, Smithies O. Normaldevelopment of mice deficient in beta 2M, MHC class Iproteins, and CD8þ T cells. Science 1990; 248:1227–1230.

140. Fleming RE, Holden CC, Tomatsu S, Waheed A, Brunt EM,Britton RS, Bacon BR, Roopenian DC, Sly WS. Mouse straindifferences determine severity of iron accumulation in Hfeknockout model of hereditary hemochromatosis. Proc NatlAcad Sci USA 2001; 98:2707–2711.

141. Sproule TJ, Jazwinska EC, Britton RS, Bacon BR, FlemingRE, Sly WS, Roopenian DC. Naturally variant autosomaland sex-linked loci determine the severity of iron overloadin beta 2-microglobulin-deficient mice. Proc Natl Acad SciUSA 2001; 98:5170–5174.

142. Parrot DMV. Strain variation in mortality and runt disease inmice thymectomizedat birth.TransplantBull 1962; 29:474–476.

143. Reimao R, Porto G, de Sousa M. Stability of CD4=CD8 ratiosin man: new correlation between CD4=CD8 profiles and iron


overload in idiopathic haemochromatosis patients. C R AcadSci III 1991; 313:481–487.

144. Porto G, Reimao R, Goncalves C, Vicente C, Justica B, deSousa M. Haemochromatosis as a window into the study ofthe immunological system: a novel correlation betweenCD8þ lymphocytes and iron overload. Eur J Haematol1994; 52:283–290.

145. Porto G, Vicente C, Teixeira MA, Martins O, Cabeda JM,Lacerda C Goncalves R, Fraga J, Macedo G, Silva BM, AlvesH, Justica B, de Sousa M. Relative impact of HLA phenotypeand CD4–CD8 ratios on the clinical expression of hemochro-matosis. Hepatology 1997; 25:397–402.

146. Cardoso EMP, Hagen K, De Sousa M, Hulcrantz R. Hepaticdamage in C282Y homozygotes relates to low numbers ofCD8þ cells in the liver lobuli. Eur J Clin Invest 2001;31:45–53.

147. Arosa FA, da Silva AJ, Godinho IM, Ter Steege JCA, Porto G,Rudd CE, De Sousa M. Decreased CD8-p56lck activity inperipheral blood T lymphocytes from patients with heredi-tary haemochromatosis. Scand J Immunol 1994; 39:426–432.

148. Arosa FA, Oliveira L, Porto G, da Silva BM, Kruijer W,Veltman J, de Sousa M. Anomalies of the CD8þ T cell poolin haemochromatosis: HLA-A3 linked expansions ofCD8þCD28– T cells. Clin Exp Immunol 1997; 107:548–554.

149. Porto G, Cardoso CS, Gordeuk V, Cruz E, Fraga J, Areias JCOliveira J, Bravo F, Gangaidzo IT, MacPhail AP, Gomo ZA,Moyo VM, Melo G, Silva C, Justica B, de Sousa M. Clinicaland genetic heterogeneity in hereditary haemochromatosis:association between lymphocyte counts and expression ofiron overload. Eur J Haematol 2001; 67:110–118.

150. Cardoso C, Porto G, Lacerda R, Resende D, Rodrigues P,Bravo F, Oliveira JC, Justica B, de Sousa M. T-cell receptorrepertoire in hereditary hemochromatosis: a study of 32hemochromatosis patients and 274 healthy subjects. HumImmunol 2001; 62:488–499.

151. Arosa FA, Porto G, Cabeda JM, Lacerda R, Resende D, CruzE, Cardoso C, Fonseca M, Simoes C, Rodrigues P, Bravo F,Oliveira JC, Alves H, Fraga J, Justica B, de Sousa M.

330 Cardoso et al.

Expansions of CD8þCD28– and CD8þTcRVbeta5.2þ T cellsin peripheral blood of heavy alcohol drinkers. Alcohol ClinExp Res 2000; 24:519–27.

152. Fitzgerald JE, Ricalton NS, Meyer AC, West SG, Kaplan H,Behrendt C, Kotzin BL. Analysis of clonal CD8þ T cell expan-sions in normal individuals and patients with rheumatoidarthritis. J Immunol 1995; 154:3538–3547.

153. Posnett DN, Sinha R, Kabak S, Russo C. Clonal populationsof T cells in normal elderly humans: the T cell equivalent to‘‘benign monoclonal gammapathy’’. J Exp Med 1994; 179:609–618.

154. Gordeuk VR, Ballou S, Lozanski G, Brittenham GM.Decreased concentrations of tumor necrosis factor-alpha insupernatants from homozygotes for hereditary hemochroma-tosis. Blood 1992; 79:1855–1860.


156. Fabio G, Zarantonello M, Mocellin C, Bonara P, Corengia C,Fargion S, Fiorelli G. Peripheral lymphocytes and intracellu-lar cytokines in C282Y homozygous hemochromatosispatients. J Hepatol 2002; 37:753–761.

157. Hentze MW, Kuhn LC. Molecular control of vertebrate ironmetabolism: mRNA based regulatory circuits operated byiron, nitric oxide and oxidative stress. Proc Natl Acad SciUSA 1996; 93:8175–8180.

158. Rouault TA, Klausner R. Regulation of iron metabolism ineukaryotes. Curr Top Cell Regul 1997; 35:1–6.

159. Torti SV, Kwak EL, Miller SC, Miller LL, Ringold GM,Myambo KB, Young AP, Torti FM. The molecular cloningand characterization of ferritin heavy chain, a tumornecrosis factor inducible gene. J Biol Chem 1988; 263:12638–12644.

160. Wei YS, Miller SC, Tsuji Y, Torti S, Torti FM. Interleukin 1induces ferritin heavy chain in human muscle cells. BiochemBiophys Res Commun 1989; 169:298–296.


161. Rogers JT. Ferritin translation by Interleukin-1 and interleu-kin-6: the role of sequences upstream of the start codons ofthe heavy and light subunit genes. Blood 1996; 87:2525–2537.

162. Konijn AM, Carmel N, Levy R, Hershko C. Ferritin synthesisin inflammation. Mechanism of increased ferritin synthesis.Br J Haematol 1981; 49:361–368.


164. Tsuji Y, Miller LL, Miller SC, Torit SV, Torti FM. Tumornecrosis factor-a and interleukin 1-a regulate transferrinreceptor in human diploid fibroblasts. J Biol Chem 1991;266:7257–7261.

165. Weiss G, Goossen B, Doppler W, Fuchs D, Pantopoulos K,Werner-Felmayer G, Wachter H, Hentze MW. Translationalregulation via iron-responsive elements by the nitricoxide=NO-synthase pathway. EMBO J 1993; 12:3651–3657.

166. Weiss G, Bogdan C, Hentze MW. Pathways for the regulationof macrophage iron metabolism by the anti-inflammatorycytokines IL-4 and IL-13. J Immunol 1997; 158:420–425.

167. Mulero V, Brock JH. Regulation of iron metabolism in murineJ774 macrophages: role of nitric oxide dependent and inde-pendent pathways following activation with gamma inter-feron and lipopolysaccharide. Blood 1999; 94:2383–2389.

168. Alvarez-Hernandez X, Licega J, McKay I, Brock JH. Induc-tion of hypoferremia and modulation of macrophage ironmetabolism by tumor necrosis factor. Lab Invest 1989;61:319–322.

169. Wardrop SL, Richardson DR. Interferon-gamma and lipopo-lysaccharide regulate the expression of Nramp2 and increasethe uptake of iron from low relative molecular mass com-plexes by macrophages. Eur J Biochem 2000; 267:6586–6593.

170. Drapier JC, Hirling H, Wietzerbin H, Kaldy P, Kuhn LC.Biosynthesis of nitric oxide activates iron regulatory factorin macrophages. EMBO J 1993; 12:3643–3650.

171. Cairo G, Castrusini E, Minotti G, Bernelli-Zazzera A. Super-oxide and hydrogen peroxide-dependent inhibition of iron

332 Cardoso et al.

regulatory protein activity: a protective stratagem againstoxidative injury. FASEB J 1996; 10:1326–1335.

172. Kim S, Ponka P. Nitrogen monoxide-mediated control offerritin synthesis: implications for macrophage iron homeos-tasis. Proc Natl Acad Sci USA 2002; 99:12214–12219.

173. Byrd T, Horwitz MA. Regulation of transferrin receptorexpression and ferritin content in human mononuclearmacrophages. Coordinate upregulation by iron transferrinand down-regulation by interferon gamma. J Clin Invest1993; 91:969–976.

174. Seiser C, Teixera S, Kuhn LC. Interleukin-2 dependent tran-scriptional and posttranscriptional regulation of transferrinreceptor mRNA. J Biol Chem 1993; 268:13074–13080.

175. Tilg H, Ulmer H, Kaser A, Weiss G. Role of IL-10 for induc-tion of anemia during inflammation. J Immunol 2002; 169:2204–2209.

176. Lounis N, Truffot-Pernot C, Grosset J, Gordeuk VR, BoelaertJR. Iron and Mycobacterium tuberculosis infection. J ClinVirol 2001; 20:123–126.

177. Recalcati S, Pometta R, Levi S, Conte D, Cairo G. Response ofmonocyte iron regulatory protein activity to inflammation:abnormal behavior in genetic hemochromatosis. Blood 1998;91:2565–2572.

178. Kontoghiorghes GJ, Weinberg ED. Iron: mammalian defensesystems, mechanisms of disease, and chelation therapyapproaches. Blood Rev 1995; 9:33–45.

179. MacMicking J, Xie QW, Nathan C. Nitric oxide and macro-phage function. Annu Rev Immunol 1997; 15:323–350.

180. Bogdan C. Nitric oxide and the regulation of gene expression.Trends Cell Biol 2001; 11:66–75.


182. Mellilo G, Taylor LS, Brooks A, Musso T, Cox GW, Varesio L.Functional requirement of the hypoxia responsive element in


the activation of the inducible nitric oxide synthase promoterby the iron chelator desferrioxamine. J Biol Chem 1997;272:12236–12242.

183. Dlaska M, Weiss G. Central role of transcription factor NF-IL6 for cytokine and iron-mediated regulation of murineinducible nitric oxide synthase expression. J Immunol 1999;162:6171–6177.

184. Gray CP, Franco AV, Arosio P, Hersey P. Immunosuppres-sive effects of melanoma-derived heavy-chain ferritin aredependent on stimulation of IL-10 production. Int J Cancer2001; 92:843–850.

185. Nemeth E, Tuttle MS, Powelson J, Vaughn MB, Domovam A,Ward DM, Gantz T, Kaplan J. Hepcidin regulates iron effluxby binding to ferroportin and inducing its internalization.Science 2004; Oct 28 epub ahead of print.

186. Miranda CJ, Markui H, Andrews NC, Santos MM. Contribu-tion of BetaZ-microglubin dependent molecules and lympho-cytes to iron regulation. Blood 2004; 103:2847–2849.

187. Macedo MF, de Sousa M, Ned RM, Mascarenhas C, AndrewsNC, Correia-Neves M. Transferrin is required for early T-celldifferentiation. Immunology 2004; 112:543–549.

334 Cardoso et al.

11

Clinical Approach to the Patientwith Anemia of Chronic Disease

VICTOR R. GORDEUK

Department of Medicine, Center for Sickle CellDisease, Howard University,

Washington, D.C., U.S.A.

INTRODUCTION

Dr. Robert Schilling has proposed that the ‘‘anemia of chronicdisease’’ is better termed the ‘‘anemia of chronic inflamma-tion,’’ for it is a hypoproliferative anemia syndrome thatoccurs as the result of the chronic inflammatory responseto an underlying disorder such as infection, malignancy,immune-mediated disease, or trauma (1). This is the mostcommon form of anemia in hospitalized adults in nontropicalcountries. The characteristic clinical findings are (1) anunderlying chronic inflammatory process, (2) mild or moder-ate anemia, (3) hypoferremia, and (4) normal or increased

PART IV: DIAGNOSIS OF ACD

335

iron stores in the bone marrow. The diagnosis of the anemia ofchronic disease is to a certain extent an exercise in exclusionof other forms of anemia. Therefore, in evaluating a patientwith the possible diagnosis of the anemia of chronic disease,it is essential to perform the history, physical examination,and laboratory tests necessary to exclude other forms of ane-mia such as immune hemolysis, nutritional and hormonaldeficiencies, hemoglobinopathies, red cell enzyme and struc-tural defects, or bone marrow failure. Treatment is usuallydirected at the underlying condition rather than the anemia.It is important to consider and make the diagnosis of anemiaof inflammation, for unnecessary therapy with iron salts dur-ing infection or inflammation is potentially toxic.

CHRONIC INFLAMMATORY PROCESS

The diagnosis of the anemia of chronic disease requires thepresence of an underlying chronic inflammatory condition.Inflammation leads to the release of certain cytokines, suchas tumor necrosis factor-a and interferon-g, which have aninhibitory effect on erythroid progenitor cells in the bone mar-row. Inflammation also profoundly affects iron metabolismthrough cytokines such as interferon-g, interleukin-1, andtumor necrosis factor-a, leading to sequestration of iron inthe storage compartment in cells of the mononuclear–phagocyte system and decreased delivery of iron to erythroidprecursors by transferrin. These same cytokines appear tosuppress the production of erythropoietin or interfere withits effect on erythroid progenitors. Inhibition of progenitorcells, iron limitation to the development of maturing eryth-rocytes, and reduced erythropoietin activity in turn appearto be major mechanisms of the anemia of chronic inflammation(2–4).

No only is the production of erythrocytes by the bonemarrow impaired, but erythrocyte survival is also modestlyreduced in chronic inflammation. Because circulating ery-throcytes have a life span of about 120 days under normal cir-cumstances and about 90 days during inflammation, an acute

336 Gordeuk

inflammatory process will not by itself lead to a reducedhemoglobin concentration; chronicity of the inflammatoryprocess (weeks to months) is necessary for the developmentof anemia (5).

Chronic inflammation related to a number of conditionsmay underlie the anemia of chronic disease. Infectious causesof chronic inflammation include osteomyelitis, tuberculosis,helminthic infections, tropical infections such as Leishmania-sis, chronic abscesses, periodontal disease, endocarditis, andHIV disease. Collagen vascular diseases such as rheumatoidarthritis and systemic lupus erythematosis and other systemicinflammatory processes such as temporal arteritis, polymyal-gia rheumatica, inflammatory bowel disease, and ischemicheart disease may be accompanied by an inflammatory-related anemia. Certain hematologic malignancies and solidtumors are accompanied by a systemic inflammatory responseand the development of an anemia not related to bone marrowreplacement or chemotherapy. Repeated trauma and frequentsurgery can also be associated with chronic inflammation andthe development of anemia (4–6).

At times, the patient with anemia possibly due to chronicdisease does not have the clinical diagnosis of an inflamma-tory process, but workup of the anemia reveals compatiblelaboratory tests and no other cause for the anemia. In suchcases, the diligent search for an underlying inflammatory pro-cess may unearth a treatable condition for which effectivetherapy will be associated with resolution of the anemia. Itis important to remember that in any patient with a lowhemoglobin concentration, the anemia is often multifactorial,and a component of chronic inflammation may combine withnutritional deficiencies, hemolysis, ineffective erythropoiesis,or bone marrow failure in promoting anemia. Laboratory teststhat are supportive of a systemic inflammatory process includeelevated erythrocyte sedimentation rate and increased serumconcentrations of C-reactive protein and a 2 globulins (2,4–6).Systemic inflammation is also associated with decreasedserum concentrations of albumin and transferrin andincreased serum concentrations of haptoglobin, fibrinogen,ceruloplasmin, amyloid protein A, and ferritin (5,7,8).

Clinical Approach to the Patient 337

HYPOPROLIFERATIVE ANEMIA OFNORMOCYTIC OR MICROCYTICMORPHOLOGY

The reduction of the hemoglobin concentration in the anemiaof chronic disease is usually mild to moderate (reviewed inRef. 5). The anemia tends to develop over the first 1–2 monthsof inflammatory illness and thereafter to not progress (9). Thehematocrit is usually maintained between 25% and 40%, withthe degree of anemia paralleling the severity of the inflamma-tory response. Microscopy of the peripheral blood smear oftenshows normocytic, normochromic red blood cells, althoughmild microcytosis and hypochromia develop in 30–50% ofpatients (10). The automated cell counter reveals normal toborderline low mean corpuscular volume and, usually, a nor-mal red cell distribution width. Occasionally, the anemia andmicrocytosis may be more marked, with hematocrits less than25% and mean corpuscular volumes less than 72 fL (3,11).Typical changes in the complete blood count of patients withthe anemia of chronic disease are contrasted and comparedto changes in patients with iron-deficiency anemia and iron-loading anemias in Table 1.

The anemia of chronic disorders is hypoproliferative incharacter, meaning that it is characterized by neither hemoly-sis nor ineffective erythropoiesis. Hemolysis is typicallyaccompanied by an increase in the reticulocyte count alongwith a reduction in the serum haptoglobin concentration andvariable increases in the serum concentrations of lactate dehy-drogenase and unconjugated bilirubin. With ineffective ery-thropoiesis, also known as intramedullary hemolysis, thereare increased erythroid precursors in the bone marrow butmost of them die before being released to the peripheral blood.Examples include sideroblastic anemias, congenital dysery-thropoietic anemias, and megaloblastic anemias. Ineffectiveerythropoiesis is characterized by changes in serum concen-trations of haptoglobin, lactate dehydrogenase, and unconju-gated bilirubin typical of hemolysis but with no increase inthe reticulocyte count. Therefore, laboratory tests supportiveof anemia of chronic disease, i.e., a hypoproliferative anemia

338 Gordeuk

in the setting of systemic inflammation, include a reticulocytecount that is not increased, normal serum concentrationsof lactate dehydrogenase and unconjugated bilirubin, andnormal to increased serum concentration of haptoglobin.

The white blood cell and platelet counts are usually notdecreased in the anemia of chronic disease. Also, the serumerythropoietin concentration tends to be in the normal orincreased range depending on the degree of anemia. Thereis some evidence that the erythropoietin response to theanemia of chronic disease is somewhat blunted, but the levelsare not severely reduced as in the anemia of chronic renalfailure or postbilateral nephrectomy (3,12).

ABSENCE OF OTHER CAUSES OF AHYPOPROLIFERATIVE ANEMIA

Other causes of a hypoproliferative anemia than chronicinflammation need to be considered and ruled out in theworkup of the anemia of chronic disease. Examples includeiron-deficiency anemia, anemia related to chronic renal fail-ure and the associated erythropoietin deficiency, and mild

Table 1 Typical Ranges in the Complete Blood Count in theAnemia of Chronic Disease Compared to Iron Deficiency Anemiaand Iron-Loading Anemias

Anemia ofchronicdisease

Irondeficiency

anemia

Chronicdisease and

irondeficiency

Iron-loadinganemias

Normalrange

Hemoglobin(g=dL)

8.5–12 4–12 4–12 7–12 12–17

Hematocrit (%) 25–36 12–36 12–36 21–36 36–52Mean

corpuscularvolume (fL)

72–95 55–90 55–90 50–120 80–100

Red celldistributionwidth

11–16 16–25 13–18 16–26 11–16


anemias associated with deficiencies of thyroxine or cortisol.Also, hypoproliferative bone marrow disorders such as myelo-dysplastic syndromes (for example, refractory anemia without excess blasts), pure red cell aplasia, aplastic anemia,and chronic treatment with myelosuppressive drugs need tobe ruled out. Myelodysplasia, aplastic anemia, and myelosup-pressive drugs may be associated with some degree of throm-bocytopenia and neutropenia, changes that do not usuallyaccompany the anemia of chronic disease. Also, myelodysplasia, aplastic anemia, pure red cell aplasia, and therapy withmyelosuppressive drugs are often associated with an increasein the serum iron concentration and transferrin saturationrather than the decrease typical of a chronic inflammatorycondition.

CHANGES IN IRON METABOLISM

Key general features of normal iron metabolism includetightly controlled absorption of iron by the duodenum and effi-cient reutilization of iron, derived from the hemoglobin ofsenescent erythrocytes removed from the circulation, for theproduction of hemoglobin in erythroid precursors in the bonemarrow. Macrophages remove senescent erythrocytes fromthe circulation by endocytosis and destroy them in the phago-lysosome in a process that includes the degradation of hemoglo-bin and release of iron from heme. Most of the iron released inthis process is then transported from the macrophage to theplasma for binding by transferrin and delivery to developingred blood cells in the bone marrow. Inflammation profoundlyalters these normal iron metabolic processes in two ways: (1)inflammation leads to reduced absorption of iron by the duode-num, and (2) inflammation causes macrophages, after erythro-phagocytosis of senescent erythrocytes, to store iron in theform of ferritin and hemosiderin rather than to transfer ironto plasma transferrin. The indirect and direct measurementsof iron status used to evaluate patients with the anemia ofchronic disease in the clinic reflect the fundamental altera-tions in iron metabolism induced by chronic inflammation.

340 Gordeuk

Characteristic features of the anemia of chronic diseaseinclude reduced serum iron concentration, total iron bindingcapacity, and transferrin saturation (7,10,20), normal to ele-vated serum ferritin concentration (8) and normal serumtransferrin receptor concentration (13). Bone marrow exami-nation reveals normal to elevated iron stores (9). The reducedserum iron concentration and transferrin saturation reflectthe inflammatory stimulus to reduce iron absorption by theenterocytes and to decrease the release of iron to plasma byerythrocyte-phagocytosing macrophages. The increased bonemarrow macrophage iron stores reflect the inflammatory sti-mulus to store iron in macrophages in the form of ferritinand hemosiderin rather than to release it to the plasma. Thenormal serum transferrin receptor levels reflect the fact thatinflammation is characterized by neither of the major stimulifor increased cellular transferrin receptor expression—increased erythropoiesis and intracellular iron deficiency.The elevated serum ferritin concentration reflects both the sti-mulus for increased storage of iron in macrophages and theincreased synthesis of ferritin in response to inflammatorycytokines such as tumor necrosis factor-a. Typical changes inindirect measures of iron status in the anemia of chronic dis-orders, iron-deficiency anemia, and iron-loading anemias arecompared and contrasted in Table 2.

DIAGNOSIS OF IRON DEFICIENCY IN THESETTING OF INFLAMMATION

At times, it can be challenging to determine if a patient withan inflammatory condition also has iron deficiency, as theinflammatory response will tend to raise the serum ferritinconcentration into the normal range. A low serum ferritinconcentration (<10–12mg=L) is diagnostic of iron deficiency.Serum ferritin concentrations typically range from 100 to2000 mg=L in patients with the anemia of chronic disease,while they are usually less than 20 mg=L in patients with irondeficiency not complicated by inflammation. In patients withthe combination of inflammation and iron deficiency, serum


Table

2T

yp

ical

Ran

ges

inIn

dir

ect

Mea

sure

sof

Iron

Sta

tus

inth

eA

nem

iaof

Ch

ron

icD

isea

seC

omp

are

dto

Iron

-Defi

cien

cyA

nem

iaan

dIr

on-L

oad

ing

An

emia

s

An

emia

ofch

ron

icd

isea

se

Iron

defi

cien

cyan

emia

Ch

ron

icd

isea

sean

dir

ond

efici

ency

Iron

-loa

din

gan

emia

sN

orm

al

ran

ge

Ser

um

iron

(mg=D

l)5–60

4–40

4–60

100–250

50–150

Tot

al

iron

bin

din

gca

paci

ty(m

g=d

L)

100–250

350–500

150–300

150–250

250–425

Ser

um

tran

sfer

rin

(mg=d

L)

80–250

250–410

140–380

125–208

200–340

Tra

nsf

erri

nsa

tura

tion

(%)

5–15

1–15

5–15

75–100

15–45

Ser

um

ferr

itin

(mg=L

)75–2,0

00

1–20

10–100

300–4,0

00

20–300

Ser

um

tran

sfer

rin

rece

pto

r(m

g=L

)0.8

–3.1

2.0

–20.0

2.5

–10.0

5–50

0.8

–3.1

Tra

nsf

erri

nre

cep

tor=

log

ferr

itin

rati

o0.2

–1.4

2.0

–20.0

1.5

–10.0

1.3

–20

0.3

–2.5

342 Gordeuk

ferritin concentrations may range from less than 20 to100mg=L (see Table 2). Indeed, in the context of a systemicinflammatory process, a serum ferritin concentration of lessthan 75–100mg=L is suggestive of associated iron deficiency(8,13,14). However, almost all studies show an overlap inthe serum ferritin concentrations between iron-deficient andiron-replete anemic patients with an underlying chronicinflammatory condition.

In the context of the anemia of inflammation, the serumtransferrin receptor concentration may be useful to diagnoseaccompanying iron deficiency. In a careful study of rheumatoidarthritis patients with anemia in whom the presence or absenceof iron deficiency was confirmed by bone marrow examination,an elevated transferrin receptor concentration was more accu-rate than serum ferritin concentration in identifying iron defi-ciency, having a greater than 90% sensitivity and specificity(15). The ratio of serum concentrations of transferrin receptorand ferritin is an accurate reflection of body iron stores in theabsence of inflammation (16) and this ratio has been proposedas an improvement over the serum ferritin concentration orthe serum transferrin receptor concentration as single teststo identify iron deficiency in the context of inflammation (13).The use of the ratio of transferrin receptor to ferritin to identifyiron deficiency in the context of the anemia of inflammation hasrecently been reviewed by Dr. Yves Beguin (2003). Some stu-dies have found the ratio to be a useful distinguishing test whileothers have found it to be no more useful than unadjustedvalues for serum ferritin concentration, transferrin receptorconcentration, and total iron binding capacity or transferrinconcentration in assessing whether a patient with an inflam-matory process and anemia has iron deficiency.

Ultimately, the gold standard for making the diagnosis ofiron deficiency is to perform a bone marrow aspirate and stainthe spicules for iron with Prussian blue; absent macrophageiron confirms that there is iron deficiency. If iron deficiencyis found to be present in a patient with a chronic inflamma-tory process, it is imperative to evaluate the patient to iden-tify a potential site of chronic blood loss and to correct theproblem if possible. This evaluation should include the upper


and lower gastrointestinal tracts as well as the genitourinarytract. Neoplasms may be diagnosed in this process at a stagewhen they are still curable.

HEPCIDIN

Hepcidin is a recently discovered 25-amino acid peptide, whichis released by hepatocytes in response to inflammation andinfection. Induction of hepcidin in experimental models leadsto the typical changes in iron metabolism of inflammation-reduced absorption of iron by enterocytes and decreasedrelease of iron by macrophages to plasma transferrin, whileknock-out or suppression of hepcidin leads to opposite effects(17). A number of investigators have proposed that hepcidinmay be a key player in the pathogenetic pathway for the ane-mia of chronic disease (17–19). As reflected in urinary concen-trations, hepcidin production is increased in patients with theanemia of chronic disease compared to healthy controls andpatients with iron-deficiency anemia. Recently, kits for themeasurement of hepcidin concentrations in plasma and serumhave become available. It is possible, therefore, that serumhepcidin measurements may in the future find a niche in thediagnosis of the anemia of chronic disease.

CONCLUSION

The diagnosis of the anemia of chronic disease can be made byconfirming a hypoproliferative anemia and finding consistentiron metabolism measurements in the setting of a chronicinflammatory process. Appendix A presents a summary ofthe clinical approach to the diagnosis of the anemia of chronicdisease.

ACKNOWLEDGMENTS

This study was supported in part by NIH research grant no.UH1-HL03679-05 from the National Heart, Lung and Blood

344 Gordeuk

Institute and the Office of Research on Minority Health andby Howard University General Clinical Research CenterGrant No. MO1-RR10284.

APPENDIX

Summary of clinical approach to diagnose anemia of chronicdisease

1. Document a mild-to-moderate hypoproliferativeanemia of normocytic to mildly microcytic morphology

� Hemoglobin 8.5–12.0 g=dL� MCV 72–90 fL� Reticulocyte count not increased� Lactate dehydrogenase and unconjugated bilirubin

concentrations not increased� Haptoglobin concentration not decreased� White blood cell and platelet counts not decreased� Erythropoietin concentration normal or increased

2. Rule out other causes of hypoproliferative anemia

� Chronic renal failure with marked erythropoietindeficiency

� Hypothyroidism� Cortisol or adrenal insufficiency� Myelodysplastic syndromes not associated with

ineffective erythropoiesis� Aplastic anemia and pure red cell aplasia� Chronic use of bone marrow suppressive drugs

3. Show compatible changes in measurements of ironmetabolism

� Reduced serum iron concentration, total iron bind-ing capacity, transferrin concentration and trans-ferrin saturation

� Normal serum transferrin receptor concentration� Normal to increased serum ferritin concentration� Normal to increased bone marrow macrophage iron


Document the presence of an underlying chronic inflam-matory process

� Infection—for example: osteomyelitis, tuberculosis,helminthic infections, Leishmaniasis, chronic abs-cesses, periodontal disease, and HIV disease

� Noninfectious inflammatory processes—for exam-ple: rheumatoid arthritis, systemic lupus erythema-tosis, temporal arteritis, polymyalgia rheumatica,inflammatory bowel disease, and ischemic heartdisease

� Hematologic malignancies and solid tumors accom-panied by a systemic inflammatory response

� Repeated trauma or frequent surgery� Laboratory evidence of inflammation—for example:

increased erythrocyte sedimentation rate andincreased serum concentrations of C-reactive pro-tein, haptoglobin, fibrinogen, ceruloplasmin, amy-loid protein A, ferritin, and a 2 globulins; decreasedserum concentrations of albumin and transferrin

REFERENCES

1. Schilling RF. Anemia of chronic disease: a misnomer [editor-ial]. Ann Intern Med 1991; 115:572–573.

2. Davis D, Charles PJ, Potter A, Feldmann M, Maini RN, ElliottMJ. Anaemia of chronic disease in rheumatoid arthritis: invivo effects of tumour necrosis factor a blockade. Br J Rheu-matol 1997; 36:950–956.


4. Fitzsimmons EJ, Brock JH. The anaemia of chronic diseaseremains hard to distinguish from iron deficiency anaemia insome cases. BMJ 2001; 322:811–812.

5. Lee GR. The anemia of chronic disorders. In: Lee GR, BithellTC, Foerster J, Athens JW, Lukens JN, eds. Wintrobe’s Clini-cal Hematology. Philadelphia: Lea & Febiger, 1993:840–851.

346 Gordeuk

6. Hutter JW, van der Velden U, Varoufaki A, Huffels RA, HoekFJ, Loos BG. Lower numbers of erythrocytes and lower levelsof hemoglobin in periodontitis patients compared to controlsubjects. J Clin Periodontol 2001; 28:930–936.

7. Bainton DF, Finch CA. The diagnosis of iron deficiencyanemia. Am J Med 1964; 37:62.

8. Lischitz DA, Cook JD, Finch CA. A clinical evaluation of serumferritin as an index of iron stores. N Engl J Med 1974;290:1213.

9. Cartwright GE. The anemia of chronic disorders. SeminHematol 1966; 3:351.

10. Cartwright GE, Wintrobe MM. The anemia of infection. AdvIntern Med 1952; 5:162.

11. LoPresti PJ, Olson L, Auerbach M. Case report: severe micro-cytosis associated with the anemia of chronic disease. Md MedJ 1996; 45:762–764.


13. Punnonen K, Irjala K, Rajamaki A. Serum transferrin receptorand its ratio to serum ferritin in the diagnosis of irondeficiency. Blood 1997; 89:1052–1057.

14. Haurani FI. Interpretation of serum ferritin in anemia ofchronic disease. Am J Hematol 2002; 69:296.

15. Fitzsimons EJ, Houston T, Munro R, Sturrock RD, Speeken-brink ABJ, Brock JH. Erythroblast iron metabolism andserum soluble transferrin receptor values in the anemia ofrheumatoid arthritis. Arthritis Rheum (Arthritis Care Res)2002; 47:166–171.

16. Cook JD, Flowers CH, Skikne BS. The quantitative assess-ment of body iron. Blood 2003; 101:3359–3364.

17. Ganz T. Hepcidin, a key regulator of iron metabolism andmediator of anemia of inflammation. Blood 2003; 102:783–788.

18. Fleming RE, Sly WS. Hepcidin: a putative iron-regulatory hor-mone relevant to hereditary hemochromatosis and the anemia


of chronic disease. Proc Natl Acad Sci USA 2001; 98:8160–8162.

19. Nemeth E, Valore EV, Territo M, Schiller S, Lichtenstein A,Ganz T. Hepcidin, a putative mediator of anemia of inflamma-tion, is a type II acute-phase protein. Blood 2003; 101:2461–2463.

20. Cartwright GE, Hugeley CM, Ashenbrucker H et al. Studieson free erythrocyte protoporphyrin, plasma iron and plasmacopper in normal and anemic subjects. Blood 1948; 3:501–525.

21. Beguin Y. Soluble transferrin receptor for the evaluation oferythropoiesis and iron status. Clin Chim Acta 2003; 329:9–22.

348 Gordeuk

12

Usefulness of Old and NewDiagnostic Tests in ACD

KARI PUNNONEN

Department of Clinical Chemistryand Laboratory Hematology, KuopioUniversity Hospital, Kuopio, Finland

ALLAN RAJAMAKI

Department of Clinical Chemistryand Laboratory Hematology, TurkuUniversity Hospital, Turku, Finland

DISTINGUISHING THE ANEMIAOF CHRONIC DISEASE FROM OTHERFORMS OF ANEMIA

The diagnosis of anemia of chronic disease (ACD) is largelybased on exclusion of other causes of anemia. The hemolyticanemias can usually be easily excluded on the basis of theabsence of signs of hemolysis, such as elevated reticulocyte

Abbreviations: ACD: anemia of chronic disease; CRP: C-reactive protein;ESR: erythrocyte sedimentation rate; IDA: iron deficiency anemia; IL-6:interleukin-6; MCV: mean corpuscular volume; TfR: transferrin receptor;TNF-alpha: tumor necrosis factor-alpha.

349

count and reduced serum haptoglobin concentration. Anincreasingly important type of anemia, pernicious anemia,is due to inadequate absorption of vitamin B12. Perniciousanemia is macrocytic and megaloblastic, and in addition itcan be distinguished from ACD on the basis of reduction inserum levels of vitamin B12 or related metabolites.

Often it is important to distinguish between ACD andiron-deficiency anemia (IDA) (1–3). Reliable diagnostic proce-dures are important because IDA needs to be carefully evalu-ated. For example, IDA may be caused by gastrointestinalbleeding due to a malignancy, and thorough investigation ofthe patient may reveal a malignancy that is curable bysurgery. In this regard, the clinical status of the patient shouldnot be overlooked. If a patient has clear signs of bleeding, thediagnostic procedures should focus on the cause of bleedingregardless of the results of the laboratory tests.

In inflammatory diseases, a highly important factor is thechange in the patient’s hemoglobin value in relation to diseaseactivity. Patients with worsening anemia despite stable dis-ease activity should be evaluated carefully. The diagnosticevaluation of anemic patients should always include theassessment of the risk of a hematological malignancy. Thefinding of thrombocytopenia or leukopenia in addition to ane-mia should raise the concern that a hematological malignancymay be present.

Due to the biological basis of ACD, the diagnostic testsshould include evaluation of inflammatory status (1,3–5). Theerythrocyte sedimentation rate (ESR) is a test that reflectsthe inflammatory response regardless of the underlying cause.For example, the ESR is a useful marker of disease activity inrheumatoid arthritis. The use of serum proteins as markers ofthe acute phase response associated with bacterial infections iswidespread, but the clinical practices vary considerably. Insome countries, serumC-reactive protein (CRP) is in use, whilein others orosomucoid may be used. Of the modern inflamma-tory markers, serum interleukin-6 (IL-6) and=or tumornecrosis factor-alpha (TNF-alpha) may be used to define theinflammatory status of an anemic patient. Currently, theclinical use of these modern markers is very limited as the

350 Punnonen and Rajamaki

analytical applications are not available on any automatedclinical chemistry analyzers. Therefore, ESR andCRP continueto be the clinically most important markers of inflammation.

RED BLOOD CELL MORPHOLOGYAND TRADITIONAL CLASSIFICATIONOF ANEMIAS ON THE BASIS OF MEANCORPUSCULAR VOLUME

In ACD, peripheral blood cell morphology may have somecharacteristic findings. To start with, the elevated ESR canbe morphologically visualized as the characteristic rouleauxformation of red blood cells. Furthermore, ACD has been tra-ditionally considered to be a normocytic anemia. However, itshould be emphasized that ACD may well be microcytic(1,3). In a retrospective analysis of patients who underwentbone marrow examination along with red blood cell analysis,

Figure 1 The distribution of MCV values in ACD and IDApatients.

Usefulness of Old and New Diagnostic Tests in ACD 351

approximately 10% of patients with ACD had a mean corpus-cular volume (MCV) value that was below 80 fL, whileapproximately 30% of patients with IDA had an MCV valueabove 80 fL (Fig. 1). From the practical point of view, thismeans that the MCV value and the simple classification ofanemias as microcytic or normocytic should not be used as astrict guide to decide the focus of laboratory tests.

In normocytic as well as microcytic anemias, both IDAand ACD have to be considered as potential diagnoses. InACD, morphological changes may be observed not only inred blood cells but also in neutrophils, which may be hyper-granular in association with bacterial infections. All in all,the changes in neutrophil and red blood cell morphology inACD are nonspecific and, therefore, the definitive diagnosismust be based on other findings and tests.

THE TRADITIONAL MARKERS OF IRONSTATUS INCLUDING FERRITIN,TRANSFERRIN, AND SERUM IRONIN DIAGNOSIS OF ACD

The most widely used serum protein markers of iron metabo-lism are serum ferritin and transferrin (2,6). Ferritin is amacromolecule that consists of a protein shell (apoferritinwith 24 subunits) and an iron core. In general, ferritin synth-esis and serum concentrations are increased as a result ofincreased iron in the reticuloendothelial system. Apoferritinsynthesis is induced by iron and it is characteristic for serumferritin that it reflects the amount of iron stores ranging all theway from diminished stores to excess of iron. Therefore, serumferritin measurements are useful not only in evaluation of irondeficiency but also in identification of patients with hemochro-matosis or transfusional iron overload. In IDA, serum ferritinconcentrations are reduced and in anemic patients the deple-tion of iron stores can be confirmed by a ferritin concentrationthat is below a certain limit. The use of conventional referencelimits is complicated by the fact that serum ferritin concentra-tion shows a sex difference in the reference ranges, and it is


questionable whether the reference limits are appropriate fordiagnostic purposes. Typically, in a variety of hematologictextbooks, the suggested decision limits for diagnosis of IDAare values less than either 12 or 20mg=L, but the scientificbasis of these limits is not well defined. In an attempt to defineproper limits for the clinical interpretation of ferritin results,the use of health-related reference values has been proposed(7), and when iron depletion was excluded by means of ironsupplementation, the cut-off limit for a low serum ferritin con-centration was estimated to be 22mg=L (7).

Those patients who have ACD and at the same timedepleted iron stores, or more precisely a clinical concernof depleted iron stores, constitute a special challenge for thedifferential diagnosis of these two types of anemia. In ACD,a characteristic finding is an increase in serum ferritin con-centration (1,4,8), and even if the condition is accompaniedby severe iron deficiency, the concentration very seldom fallsbelow 20 mg=L. Therefore, in these patients, the properdecision limit for diagnosis of iron depletion is considerablyabove the conventional lower reference limit (approximately10–15 mg=L) or the proposed health-related reference limit(22 mg=L) (7).In the differential diagnosis between ACD andcombined ACD and iron deficiency, the optimal cut-off levelfor serum ferritin concentration has been estimated to beapproximately 40–60mg=L (9). On the basis of meta-analysis,it has been even stated that a gray zone exists even between15 and 100mg=L, where iron deficiency can neither be dis-closed nor ruled out on the basis of serum ferritin concentra-tion (6). Considering these difficulties in the differentialdiagnosis of IDA and ACD, there has been a need for more dis-criminating laboratory tests of iron status. In the differentialdiagnosis of IDA and ACD, a bone marrow examination isfrequently performed to reach a definitive diagnosis regard-ing the amount of iron stores.

Transferrin is the transport protein for serum iron. Thetraditional statement is that the serum transferrin concentra-tion increases in iron deficiency and is therefore useful indiagnosis. However, when considering the differential diagno-sis of IDA and ACD, the situation is complicated because


transferrin is a negative acute phase reactant. From the prac-tical point of view, this means that patients who have ACDand an inflammatory condition have a reduction in transfer-rin synthesis due to the effects of the inflammatory cytokines.In retrospective studies, it has been shown that in IDA,transferrin concentrations are higher than in ACD but, unfor-tunately, even if the patient groups of IDA and ACD differ intheir transferrin concentrations, the majority of all the valuesstill fall within the reference range of healthy subjects(6,9,10). Therefore, serum transferrin concentration is not ofgreat value in diagnosis of ACD.

A basic phenomenon in the pathophysiology of ACD is ironredistribution into macrophage stores and reduced iron avail-ability for hemoglobin synthesis. This iron redistribution isreflected by a reduced serum iron concentration (Fig. 2). Adecrease in the serum iron concentration causes the ironsaturation of transferrin to decrease in both ACD and IDA.Although iron saturation of transferrin has become a widelyused laboratory measure of iron status, in the diagnosis ofACDand IDA, it only provides diagnostic efficiency comparableto that of the MCV value alone (6).

Figure 2 The effects of iron deficiency and inflammatory condi-tions on the markers of iron status. The " indicates an elevationin concentration while # indicates a reduction in the concentrationunder the specified condition.


SOLUBLE TRANSFERRIN RECEPTORAND TfR-F INDEX IN THE DIFFERENTIALDIAGNOSIS OF IDA AND ACD

Transferrin receptor (TfR) participates in the transport ofiron–transferrin complexes from the extracellular space intothe cell. Small amounts of TfR are expressed on the surfaceof virtually every cell but the expression is by far most abun-dant in erythroid precursors in the bone marrow. Serum TfR(sTfR) is mostly derived from bone marrow erythroblasts, andthe concentrations are elevated both by enhanced erythropoi-esis and iron deficiency (2,11–16). Soluble TfR is releasedby cleavage of membrane receptors between amino acids100–101 (Arg–Lys) just above the cell membrane. In serum,the truncated TfR exists as a complex with transferrin(17,18). High performance liquid chromatography (HPLC)size-fractionation suggests that the predominant form ofTfR in serum is a dimeric TfR in complex with transferrin(19,20). Circulating TfR provides high sensitivity and specifi-city for the diagnosis of depleted iron stores provided thathemolysis and hematological malignancies can be excludedas causes of anemia (14–16,21,22).

In several studies, TfR measurements have proven to beespecially useful in the differential diagnosis of IDA andACD (9,21,23) (Fig. 3), and the fact that TfR concentrationsare not influenced by acute phase responses provides TfR anessential benefit over the more conventional measurementsof iron status. Furthermore, in the differential diagnosis ofIDA and ACD, the interpretation of sTfR concentration iseasier than that of serum ferritin concentration, because anelevation of TfR above the upper reference limit indicates irondeficiency while in ACD the values stay within the referencelimits (9). Currently, TfR assays lack international standardi-zation and each TfR assay system has its own reference limits;these facts need to be taken into account in the interpretationof results. In the literature, the majority of articles concludethat TfR is useful in the differential diagnosis of IDA andACD, but there are also reports to the contrary. Some studieshave questioned the usefulness of TfR measurements, because


it is possible to identify patients who have replete iron storeson the basis of bone marrow iron staining despite elevatedsTfR concentrations. This has been the case, for example, inseveral studies on patients with rheumatoid arthritis andother subtypes of ACD (24–26), as well as in some studies onheterogeneous patient populations (27,28). There are two pos-sible explanations for elevated sTfR concentration when bonemarrow iron stores are present. One explanation is enhancederythropoiesis as a source of TfR. Additionally, recent studieshave suggested that at least in some cases, the elevated sTfRconcentration may reflect functional iron deficiency (29). Pro-vided that there are no signs of enhanced erythropoiesis, thisimplies that elevated sTfR concentration might reflect func-tional iron deficiency even if stainable iron is present in thebone marrow (29). The studies in which iron status has beenassessed using iron supplementation have suggested thatelevated TfR is a characteristic finding in functional (subclini-cal) iron deficiency (30). It is therefore important to rea-lize that even slightly elevated TfR concentrations have a

Figure 3 Summary of the preferred laboratory test in the differ-ential diagnosis of IDA and ACD. The basic blood count is a prere-quisite and additional tests depend on the laboratory results alongwith the clinical status.


biological background and are related either to iron deficiencyor increased rate of erythropoiesis.

Some studies indicate that the diagnostic efficiency ofserum ferritin and TfR measurements can be improved bythe calculation of the TfR-F Index, which is the sTfR=log fer-ritin ratio. The biological background for the combined use offerritin and TfR is their different behaviors in response todiminishing iron stores (Fig. 4). Serum ferritin concentrationscorrelate in a linear manner with diminishing iron stores, butthere is no decisive concentration to tell when the iron storesof a given patient have diminished so much that iron avail-ability has become a limiting factor for erythropoiesis. Thisis contrary to TfR, which responds clearly at this point withincreasing serum concentration. Several studies have shownthat the TfR-F Index is the most efficient parameter for dis-tinguishing between IDA and ACD (9,10,31). Furthermore,the TfR-F Index has been reported to be useful in detectionof functional iron deficiency irrespective of the concurrentiron status and it is also able to define the patients who can

Figure 4 The relationship between iron stores and serum concen-trations of ferritin and TfR in the course of progressive iron deple-tion. The vertical arrow indicates the point of complete depletionof iron stores.


be expected to benefit from supplemental iron. However, theTfR-F Index is a calculated parameter and the setting of theproper cut-off values needs to be done with care for each pairof assay systems. From the practical point of view, it is desir-able to have both the TfR and ferritin assays from the sameassay system manufacturer. Automated methods haverecently become available for TfR assays and, therefore, TfRnow has the potential of becoming a routine laboratory testof iron status. Thus far, the use of the TfR-F Index in clinicalroutine has been very limited, but there are emerging assaysystems which provide analysis of both ferritin and TfR onthe same assay platform or analyzer. The most ideal situationwould be to use only one analyzer to measure all theparameters needed for the diagnosis of iron deficiency.

ANALYSIS OF HEMOGLOBIN SYNTHESISAND RED BLOOD CELLS

Tests of iron status include both erythrocyte-free protopor-phyrin and zinc protoporphyrin (ZPP) measurements. At thefinal stages of heme biosynthesis, a trace of zinc instead ofiron is incorporated into protoporphyrin and when iron avail-ability is reduced ZPP formation is enhanced. ZPP measure-ment and erythrocyte-free protoporphyrin have been used inthe diagnosis of iron deficiency. However, the values are alsoelevated in all conditions that are accompanied by chronicinflammatory responses, which make them less useful forthe differential diagnosis of IDA and ACD (32,33).

Some advanced hematological analyzers provide newparameters based on the red blood cell and reticulocyte indices,and theseparametershavebeensuggested tobehelpful indiag-nosis and screening of IDA (27,34,35). These indices directlyreflect the hemoglobin content of red cells or reticulocytes andthereby allow real-time estimation of bonemarrow iron status,and theoretically they may be able to reflect rapid changes iniron availability for erythropoiesis. In patients with renal fail-ure, the management of recombinant human erythropoietin


(rhEPO) treatment is dependent on a continuous supply of ironand, in this context, the percentage of hypochromic red bloodcells has been successfully used in that evaluation of iron status(34,36,37).Ahighpercentageofhypochromic redbloodcells orareduction in hemoglobin content of reticulocytes (CHr) reflectsiron deficient erythropoiesis, which can be a result of absoluteor functional iron deficiency. The advanced RBC and reticulo-cyte indices may well be useful in identification of iron defi-ciency in screening programs, but their diagnostic usefulnessin the differential diagnosis of IDA and ACD remains to beproven.

NOVEL MARKERS OF IRON TURNOVER

Currently, laboratory tests of iron metabolism in one way orthe other deal with the amount of iron stores in the body,but these measures do not give any direct answers regardingthe ability to use iron in erythropoiesis and more specificallyin hemoglobin synthesis. The exact biological background forACD has not been clarified, but soluble components have beenhypothesized to play a role in the signaling pathways involvedin iron absorption and iron utilization. Recent studies haveproposed that a hepatic protein, hepcidin, might play a rolein the ACD. Hepcidin is a 25-amino acid peptide which linksthe inflammatory process with iron recycling and erythropoi-esis. Initially, hepcidin was defined as an antimicrobial pep-tide, but later studies revealed that it may have a centralrole in regulation of iron homeostasis (38,39). Importantly,hepcidin causes macrophages to sequester iron and it alsoprevents intestinal iron absorption. The exact mechanismsof regulation have not been dissected out yet, but importantly,hepcidin expression has been shown to be induced by proin-flammatory cytokines including IL-6 (39). In a mouse model,hepcidin expression has been shown to be sensitive to body-iron stores, erythroid demand, hypoxia, and experimentallyinduced inflammation (40). Regarding ACD, this suggeststhat hepcidin might be one of the key regulators linking theinflammatory response and the sequestration of iron into


macrophages. Very recently, hepcidin has been found to bepresent in abundant quantities in urine of ACD patients,and the concentrations are also elevated in association withiron overload (39). These findings indicate that in the nearfuture specific serum or urinary protein measurements mightbe available for the diagnosis of ACD.

SUMMARY

Anemia of chronic disease is a condition caused by chronicinfection, inflammation or cancer, and the differential diagno-sis between IDA and ACD is a common clinical problem. Thereare no direct laboratory indicators to tell that an anemia ofa given patient is ACD. Therefore, indirect measures of ironstatus have been used, but even they are partly rendered inva-lid by nonspecific changes caused by inflammation (Fig. 2). Asa part of the acute phase response, the synthesis and serumconcentration of the iron carrying protein, transferrin, isdiminished and that of the storage protein, ferritin, isincreased. In this regard, the soluble TfR measured in serumis different in that the concentrations are not affected byinflammatory reactions but the concentrations are elevatedas a result of depletion of iron stores. TfR has proved to be auseful laboratory test in the differential diagnosis of IDAand ACD (Fig. 4), and the fact that in ACD, the TfR concentra-tions are normally within the reference range of healthy sub-jects is of importance as it makes the clinical interpretation ofthe results easier. When the serum ferritin and TfR results arecombined by using the calculated TfR-F Index, the diagnosticsensitivity and specificity may be further improved, and TfRand the TfR-F Index have been reported to identify ACDpatients who have a concomitant iron deficiency. However,even the TfR assays have their limitations as the concentra-tions are not solely dependent on iron status but also influ-enced by the degree of erythropoiesis. Thus far, the aimof all the diagnostic tests in the differential diagnosis of ACDand IDA has been to evaluate iron stores. Recent advancesin basic research on the regulation of iron metabolism and


the pathophysiology of ACD may in the future provide labora-tory tests that directly and specifically indicate inflammation-specific disturbances of iron metabolism.

REFERENCES


2. Worwood M. Serum transferrin receptor assays and theirclinical application. Ann Clin Biochem 2002; 39:221–230.

3. Means RT Jr. The anemia of infection. Baillieres Clin Haematol2000; 13:151–162.

4. Konijn A. Iron metabolism in inflammation. Baillieres ClinHaematol 1994; 7:829–849.

5. Trey JE, Kushner I. The acute phase response and the hema-topoietic system: the role of cytokines. Crit Rev Oncol Hematol1995; 21:1–18.

6. GuyattGH,OxmanAD,AliM,WillanA,McIIroyW, PattersonC.Laboratory diagnosis of iron-deficiency anemia: an overview.J Gen Intern Med 1992; 7:145–153.

7. Suominen P, Irjala K. Health-related decision limits ratherthan reference limits to guide the use of ferritin measure-ments. Clin Biochem 2001; 34:345–346.

8. Worwood M. The laboratory assessment of iron status—anupdate. Clin Chim Acta 1997; 259:3–23.

9. Punnonen K, Irjala K, Rajamaki A. Serum transferrin receptorand its ratio to serum ferritin in diagnosis of iron-deficiency.Blood 1997; 89:1052–1057.

10. Suominen P, Punnonen K, Rajamaki A, Irjala K. Serum trans-ferrin receptor and transferrin receptor–ferritin index identifyhealthy subjects with subclinical iron deficits. Blood 1998;92:2934–2939.

11. Kohgo Y, Nishisato T, Kondo H, Tsushima N, Niitsu Y,Urushizaki I. Circulating transferrin receptor in humanserum. Br J Haematol 1986; 64:277–281.


12. Kohgo Y, Niitsu Y, Kondo H, Kato J, Tsushima N, Sasaki K,Hirayama M, Numata T, Nishisato T, Urushizaki I, Kato I.Serum transferrin receptor as a new index of erythropoiesis.Blood 1987; 70:1955–1958.

13. Kohgo Y, Niitsu Y, Nishisato T, Kato J, Kondo H, Sasaki K,Urushizaki I. Quantitation and characterization of serumtransferrin receptor in patients with anemias and polycythe-mias. Jpn J Med 1988; 27:64–70.

14. BegumY, Lampertz S, De Groote D, Igot D,MalaiseM, Fillet G.Soluble CD23 and other receptors (CD4, CD8, CD25, CD71) inserumof patientswith chronic lymphocytic leukemia. Leukemia1993; 7:2019–2025.

15. Flowers CH, Skikne BS, Covell AM, Cook JD. The clinicalmeasurement of serum transferrin receptor. J Lab Clin Med1989; 114:368–377.

16. Cook JD. Iron-deficiency anemia. Baillieres Clin Haematol1994; 7:787–804.

17. Shih YJ, Baynes RD, Hudson BG, Flowers CH, Skikne BS,Cook JD. Serum transferrin receptor is a truncated form oftissue receptor. J Biol Chem 1990; 265:19077–19081.

18. Shih YJ, Baynes RD, Hudson BG, Cook JD. Characterizationand quantitation of the circulating forms of serum transferrinreceptor using domain-specific antibodies. Blood 1993; 81:234–238.

19. Hikawa A, Nomata Y, Suzuki T, Ozasa H, Yamada O. Solubletransferrin receptor–transferrin complex in serum: measure-ment by latex agglutination nephelometric immunoassay. ClinChim Acta 1996; 254:159–172.

20. Kato J, Kobune M, Kohgo Y, Fujikawa K, Takimoto R,Torimoto Y, Ito Y, Bessho M, Hotta T, Hikawa A, Fujii T,Punnonen K, Niitsu Y. Ratio of transferrin (Tf) to Tf–receptorcomplex in circulation differs depending on Tf iron saturation.Clin Chem 2002; 48:181–183.

21. Punnonen K, Irjala K, Rajamaki A. Iron-deficiency anemia isassociated with high concentrations of transferrin receptor inserum. Clin Chem 1994; 40:774–776.



23. Ferguson BJ, Skikne BS, Simpson KM, Baynes RD, Cook JD.Serum transferrin receptor distinguishes the anemia of chronicdisease from iron deficiency anemia. J Lab Clin Med 1992;119:385–390.

24. Zoli A, Altomonte L, Mirone L, MagaroM, Ricerca BM, Storti S,Candido A, BizziM. Serum transferrin receptors in rheumatoidarthritis. Ann Rheum Dis 1994; 53:699–701.

25. Baumann Kurer S, Seifert B, Michel B, Ruegg R, Fehr J. Pre-diction of iron deficiency in chronic inflammatory rheumaticdisease anaemia. Br J Haematol 1995; 91:820–826.

26. Junca J, Fernandez-Aviles F, Oriol A, Navarro JT, Milla F,Sancho JM, Feliu E. The usefulness of the serum transferrinreceptor in detecting iron deficiency in the anemia of chronicdisorders. Haematologica 1998; 83:676–680.

27. Mast AE, Blinder MA, Gronowski AM, Chumley C, Scott MG.Clinical utility of the soluble transferrin receptor and compar-ison with serum ferritin in several populations. Clin Chem1998; 44:45–51.

28. Means RT Jr, Allen J, Schuster M. Serum soluble transferrinreceptor and the prediction of marrow aspirate iron results ina heterogeneous group of patients. Clin Lab Haematol 1999;21:161–167.

29. Das Gupta A, Abbi A. High serum transferrin receptor level inanemia of chronic disorder indicates coexistent iron deficiency.Am J Hematol 2003; 72:158–161.

30. Takala T, Suominen P, Lehtonen-Veromaa M, Mottonen T,Viikari J, Rajamaki A, Irjala K. Increased serum soluble trans-ferrin receptor concentration detects subclinical iron defi-ciency in healthy adolescent girls. Clin Chem Lab Med 2003;41:203–208.

31. Suominen P, Mottonen T, Rajamaki A, Irjala K. Single valuesof serum transferrin receptor and transferrin receptor ferritinindex can be used to detect true and functional iron deficiencyin rheumatoid arthritis patients with anemia. ArthritisRheum 2000; 43:1016–1020.


32. Hastka J, Lasserre JJ, Schwarzbeck A, StrauchM, Hehlmann R.Zinc protoporphyrin in anemia of chronic disorders. Blood 1993;81:1200–1204.

33. Rettmer L, Carlson TH, Origines M, Jack RM, Labbe RF. Zincprotoporphyrin=heme ratio for diagnosis of preanemic irondeficiency. Pediatrics 1999; 104:1–5.

34. Brugnara C. Use of reticulocyte cellular indices in the diagno-sis and treatment of hematological disorders. Int J Clin LabRes 1998; 28:1–11.

35. Kotisaari S, Romppanen J, Penttila I, Punnonen K. The Advia120 red blood cell and reticulocyte indices are useful in diagno-sis of iron-deficiency. Eur J Haematol 2002; 68(3):150–156.

36. Schaefer RM, Schaefer L. Hypochromic red blood cells andreticulocytes. Kidney Int 1999; 55:44–48.

37. Brugnara C, Chambers L, Malynn E, Goldberg MA,Kruskall MS. Red blood cell regeneration induced bysubcutaneous recombinant erythropoietin: iron-deficient ery-thropoiesis in iron-replete subjects. Blood 1993; 81:956–964.

38. Pigeon C, Ilyin G, Courselaud B, Leroyer P, Turlin B, Brissot P,Lo O. A new mouse liver-specific gene, encoding a proteinhomologous to human antimicrobial peptide hepcidin, is over-expressed during iron overload. J Biol Chem 2001; 276:7811–7819.


40. Nicholas G, Viatte L, Bennoun M, Kahn A, Vaulont S.Hepcidin, a new iron regulatory peptide. Blood Cells Mol Dis2002; 29:327–335.


13

Treatment of ACD: An Introduction

GUNTER WEISS

Medical University, Department of GeneralInternal Medicine, Clinical Immunology and

Infections Diseases, Innsbruck, Austria

Since anemia of chronic disease (ACD) is in most cases asecondary phenomenon driven by an activated immuneresponse initiated by an underlying disease, the treatmentof the underlying disease is thus the pivotal approach to treatACD. However, a sufficient treatment of the underlying dis-ease is not always possible, particulary in the case of patientswith malignancies, chronic infections, or autoimmune disor-ders. Thus, specific therapeutic regimen are warranted thatinclude the application of transfusions, iron, and=or humanrecombinant erythropoietin, all of which are very wellreviewed in the following chapters. Importantly, the indica-tions for the specific therapeutic regimen may vary with theunderlying disease and can be completely different, for exam-ple, between patients with mammary carcinoma and those

PART V: THERAPY

365

with Crohn’s disease. Thus, it is mandatory to specificallychoose the best available strategy for ACD therapy in respectto the underlying disease. Accordingly, little data are avail-able on therapeutic endpoints in terms of finding the besthematocrit or hemoglobin level for ACD patients. An insuffi-cient correction of anemia is associated with adverse effectstoward the patient’s quality of life or cardiopulmonal perfor-mance (1,2), while over-correction of anemia may also harborseveral problems such as thrombembolic complications. InACD patients suffering from certain types of cancer, recentdata suggest that over-correction of ACD may exert negativeeffects on the clinical course of the disease (3).

Based on recent developments concerning the pathophy-siology of ACD, new therapeutic options may emerge rangingfrom the development of antagonists of acute phase proteinssuch as hepcidin, to the use of cytokine=cytokine inhibitorssuch as IL-1 receptor antagonists or inhibitors of TNF-a activ-ity, to the clinical application of modulators of iron homeosta-sis such as iron chelators that are able to increase endogenouserythropoietin formation (4), to the therapeutic use of hor-mones or recombinant growth factors such as IL-3 or stem cellfactor that could stimulate erythroid progenitor proliferation.

REFERENCES

1. Murphy ST, Parfrey PS. The impact of anemia correction oncardiovascular disease in end stage renal disease. SeminNephrol 2000; 20:350–355.

2. Collins AJ, Ma JZ, Ebben J. Impact of hematocrit on morbidityand mortality. Semin Nephrol 2000; 20:345–349.

3. Henke M, Laszig R, Rube C, Schafer U, Haase KD, Schilcher B,Mose S, Beer KT, Burger U, Dougherty C, Frommhold H.Erythropoietin to treat head and neck cancer patients withanaemia undergoing radiotherapy: randomised, double-blind,placebo-controlled trial. Lancet 2003; 362:1255–60.

4. Salvarani C, Baricchi R, Lasagni D. Effects of desferrioxaminetherapy on chronic disease anemia associated with rheumatoidarthritis. Rheumatol Int 1996; 16:45–8.

366 Introduction

14

Human Recombinant Erythropoietin

HANSPREET KAUR and ALAN LICHTIN

Department of Hematology=Oncology,Cleveland Clinic Foundation,

Cleveland, Ohio, U.S.A.

DEEPJOT SINGH

Department of Hematology=Oncology, University Hospitals of

Cleveland, Cleveland, Ohio, U.S.A.

INTRODUCTION

Anemia of chronic disease (ACD) is commonly seen secondaryto a variety of medical conditions including rheumatoid arthri-tis, malignancies, chronic heart failure, acquired immunodefi-ciency syndrome (AIDS), and inflammatory bowel disease(IBD). The anemia of chronic disorders is characterized by aslightly shortened red cell life span, disturbed iron metabo-lism, and impaired erythropoietin-generated red cell produc-tion. The presence of a low serum iron level despite adequateiron stores indicates a profound disturbance of iron metabo-lism. The combination of a low serum iron level and transfer-rin saturation along with low total iron-binding capacity andnormal or high ferritin levels have been used to differentiate

367

ACD from iron deficiency anemia. However, the diagnosis isoften clouded by the elevation of serum ferritin in infectionand inflammatory conditions. The production of the newlydescribed hepcidin molecule is increased 100-fold in anemiaof inflammation and this may lead to increased sequestrationof iron in macrophages (1). (The differentiation of ACD fromiron deficiency anemia is explained in detail elsewhere.)

Treating the underlying disease and red blood cell (RBC)transfusions have been the traditional therapies for sympto-matic anemia until the last decade. The mainstay of success-ful therapy of anemia of chronic disease was, is, and willalways be treatment of the underlying medical condition.Since the late 1980s, clinical studies have demonstrated somepositive effects of erythropoietin on both hemoglobin valuesand patient symptoms for patients with ACD whose chronicdisease cannot be adequately ameliorated. It is essential,however, to exclude other potentially correctable causes ofanemia, including hemolysis, nutritional and hormonal defi-ciencies, hemoglobinopathies, red cell enzyme and structuraldefects, or bone marrow failure, prior to committing thepatient with ACD to erythropoietic stimulants. In thissection, we will focus on the mechanism and role of humanrecombinant erythropoietin in the treatment of ACD.

MECHANISM OF ACTION

Endogenous erythropoietin (EPO) is a glycoprotein thatstimulates the division and differentiation of committederythroid progenitors in the bone marrow, thus enhancingred blood cell production (2). In ACD, increased productionof inflammatory cytokines like tumor necrosis factor-a, inter-feron g, interleukin-6, and interleukin-1 in the bone marrowand other organs has been demonstrated. These inflamma-tory cytokines may be participants in the pathogenesis ofACD because they can impede endogenous EPO production,impair erythroid colony formation in response to EPO, thusinducing erythroid cell apoptosis, prevent the normalutilization of iron, and decrease the life span of erythrocytes

368 Kaur et al.

(3,4). Papadaki et al. (5) showed increased TNF-a productionin rheumatoid arthritis patients with ACD. Cytokine concen-trations correlated positively with the percentage of apoptoticerythroid cells. After administration of a monoclonal antibodyto TNF-a, decreased levels of TNF-a and percentage of apopto-tic cells along with an increase in erythroid progenitors andprecursors were documented.

Erythropoietin prevents the apoptotic cell death ofimmature erythroblasts by inducing the antiapoptoticproteins, Bcl-xL and Bcl-2 (6). These erythroid progenitorseventually mature to erythrocytes during erythropoiesis.Recombinant human erythropoietin (rHu-EPO) is a 165 aminoacid glycoprotein with the same biological effects as endogen-ous EPO. It is produced by recombinant DNA technology bymammalian cells into which the human erythropoietin genehas been introduced. The amino acid sequence, carbohydratecontent (40% by weight), and molecular weight (30.4 kDa) ofrecombinant and endogenous human EPO are identical (7).Administration of sufficient amounts of rHu-Epo may counter-balance the effects of proapoptotic inflammatory cytokinesthat inhibit erythropoiesis.

SERUM ERYTHROPOIETIN LEVELSAS GUIDELINES FOR THERAPY

Endogenous erythropoietin production is regulated by tissueoxygenation. Anemia and hypoxia increase erythropoietinproduction with a resultant increase in erythropoiesis.Depending upon the laboratory, reference ranges for plasmaerythropoietin levels are 4.1–22 IU=L. Erythropoietin levelsare typically inversely related to hemoglobin (and hematocrit)levels in anemias not attributed to impaired erythropoietinproduction.

During hypoxia or anemia, erythropoietin levels canincrease 100- to 1000-fold (8). In contrast, in patients withrheumatoid arthritis, inflammatory bowel disease, AIDS-associated anemia, and malignancy, production of erythro-poietin is impaired, and relatively low values of EPO

Human Recombinant Erythropoietin 369

(< 200 IU=L) for the degree of anemia are seen (9–11). In aseries of four clinical trials involving 255 patients, 60–80%of HIV-infected patients treated with zidovudine had endo-genous serum erythropoietin levels �500 IU=L (10). Patientsreceiving zidovudine at doses of �4200mg=week and withendogenous serum EPO levels �500 IU=L may respond to ery-thropoietin therapy. Decreased transfusion requirements andincreased hemoglobin and hematocrit values manifest theresponse to EPO. In general, HIV-infected patients with endo-genous EPO levels >500 IU=L neither show an increase intheir hemoglobin=hematocrit, nor do their transfusionrequirements decrease with EPO administration (12).

Later in this chapter we will discuss the role of erythro-poietin in the treatment of anemia of chronic disease relatedto HIV infection, rheumatoid arthritis, inflammatory boweldisease, and malignancy or chemotherapy. Table 1 sum-marizes the treatment protocols of the above conditions.

ERYTHROPOIETIN THERAPY FOR HIVINFECTION=HIV TREATMENT RELATED ANEMIA

Initial Therapy

The recommended starting dose of rHu-Epo in HIV=AIDSis 100 IU=kg subcutaneously (SQ) or intravenously (IV)three times per week for 8 weeks. Doses are increased in50–100U=kg increments every 4 weeks until a rise in hema-tocrit of 5%, or a hematocrit of 36%, or a maximum dose of300 IU=kg three times per week is achieved or the responseis not satisfactory in terms of reducing transfusionrequirements.

Maintenance Therapy

Once the target response is reached, patients are evaluatedevery 4–8 weeks. Doses are adjusted by 50–100 IU=kg incre-ments every 2–4 weeks. EPO is held if the hematocrit exceeds40% and treatment is resumed at a 25% dose reduction whenhematocrit drops to 36% (10,12).

370 Kaur et al.

Table

1Anem

iaof

Chronic

Disea

se:ClinicalUse

ofErythropoietin

Disea

se=condition

Optimaltrea

tmen

tsched

ule

withrecombinanterythropoietin

1.HIV

aIn

itial:100IU

=kgSQ

orIV

threetimes

awee

k�

8wee

ks

Increm

ent:50–100IU

=kgev

ery4wee

ksto

maxim

um

of300IU

=kgthreetimes

awee

kGoa

lof

trea

tmen

t:hem

atocrit

(Hct)36%

or5%

increa

sein

Hct.If

theHct

is>40%,withholdrH

u-E

PO

untilHct

is<36%.

Restart

witha25%

dosereductionandtitrateddosage

2.Rheu

matoid

arthritis

Initial:150IU

=kgtw

icewee

kly

over

12wee

kswith200mg

ofIV

iron

per

wee

kAlternative:

240IU

=kgthreetimes

awee

kfor6wee

ks

withmaintenance

doseof

240IU

=kgper

wee

k3.Autologou

sblood

don

ation

preop

Elective,

non

card

iac,

non

vascularsu

rgeryforHb

>10and

�13g=dL.250–300IU

=kgtw

icewee

kly

�3wee

kspreop

eratively

Iron

:200mgIV

atea

chdon

ation

visit

ordailyiron

supplemen

t4.In

flammatory

bow

eldisea

seIn

itial:150IU

=kgtw

ice–

thrice

wee

kly

withoralor

IViron

supplemen

tation

Maintenance:150IU

=kgper

wee

k5.Chem

otherapyor

non

myeloid

malignancy

For

hem

oglobin

�10mg=dL

andseru

mEPO

<200IU

=L

Ifhem

oglobin

is10–12g=dL,thedecisionto

trea

tis

basedon

clinicalseverity

Initial:150IU

=kgthreetimes

awee

kfor4wee

kswithiron

supplemen

tation

ifferritin

<100mg

=L

HolddoseiftheHbincrea

ses�4g=dLin

any2-w

eekperiod

Increm

ent:300IU

=kgfor4–8wee

ks(ifhem

oglobin

increa

ses<1–2g=dL)

Maintenance:titrate

rHu-E

PO

doseto

maintain

hem

oglobin

around12g=dL

aIn

gen

eral,patien

tswithseru

merythropoietin

levelsof

<500IU

=mLresp

ondto

therapywithrH

u-E

PO.


ERYTHROPOIETIN THERAPY OF ANEMIAIN PATIENTS WITH RHEUMATOID ARTHRITIS

Rheumatoid arthritis is associated with iron deficiency andACD in up to 50% of cases. Recombinant human erythro-poietin in combination with intravenous iron has been usedto treat the anemia of chronic disease related to rheumatoidarthritis. One hundred and fifty IU=kg rHu-Epo was admi-nistered twice weekly for 12 weeks along with 200mg ofintravenous iron–sucrose per week (13). A 100% responserate was seen with the average Hb concentration increasingfrom 10.7� 1.1 to 13.2� 1.0 g=dL after a mean treatmentperiod of 8.7� 2.3 weeks. Improvement in muscle strength,fatigue, and vitality scores was documented. Reduction indisease activity as evidenced by a significant improvementin the number of swollen=tender joints, erythrocyte sedi-mentation rate, Disease Activity Score, and RheumatoidArthritis Disease Activity Index was also seen duringtreatment.

Peeters et al. (14) documented significant increases inhemoglobin compared to placebo within 6 weeks of treatmentwith rHu-EPO at doses of 240 IU=kg given three times aweek. Response was sustained by weekly administration ofrHu-EPO at median doses of 240 IU=kg. Significant differ-ences in favor of the treatment were also observed in second-ary scores for disease activity including the number ofswollen joints, pain score, erythrocyte sedimentation rate,and patients’ global assessment of disease activity.

The role of rHu-EPO in stimulating an erythropoieticresponse for autologous blood donation in rheumatoid arthri-tis patients undergoing elective orthopedic procedures suchas total hip or knee arthroplasty has been studied (15). EPOwas given in doses of 600 IU=kg, IV, six times over 3 weeksand compared to placebo. The response to EPO was 624�137mL (blood harvested, mean�SD) as compared with271� 174mL (p¼ 0.02) for the placebo group. Matsuda et al.(16) reported successful preoperative autologous blood dona-tion of 400mL with rHu-EPO (12,000U per week) withoutrequiring any homologous blood transfusions.

372 Kaur et al.

TREATMENT OF ANEMIAIN INFLAMMATORY BOWEL DISEASE

Patients with IBD can have anemia that is refractory to treat-ment with iron and vitamins. The severity of anemia is animportant part of the symptom complex represented in theclinical indices of disease activity (17). Two small randomizedclinical trials (18,19) have investigated the role of EPO in theanemia of IBD refractory to iron therapy. Schreiber et al.(18)used EPO in doses of 150U=kg twice weekly along with oraliron (100mg=day) and compared these in a randomized, pro-spective double-blinded study to placebo with oral iron. Twelveweeks of combined therapy with oral iron and EPO increasedhemoglobin levels by more than 1g=dL in 82% of the patientsvs. 24% in the placebo group. Gasche et al. (19) administeredEPO in doses of 150U=kg three times weekly along with intra-venous iron and compared this to placebo in addition to intrave-nous iron therapy in a group of 40patientswithCrohns’ diseasewith anemia (hemoglobin concentration <10.5 g=dL) that hadbeen unresponsive to oral iron therapy. Patients with a recenthistory of myelosuppressive or immunosuppressive therapy,cancer, hemolysis, cobalamin or folate deficiency, hyperten-sion, thrombosis, iron overload, or those with serum creatinineconcentration >2mg=dL were excluded. Hemoglobin valuesincreased by more than 2.0 g=dL in 94% of patients given thecombination of EPO and intravenous iron. In the placebo withintravenous iron group, 66% of patients showed an increase inhemoglobin values. The response to iron alone was also mark-edly slower thanwith combined therapy. Increasedhemoglobinlevels were associated with improvement in the quality of life.Although the number of patients was small, the results areencouraging especially with respect to improvement of qualityof life parameters.

ERYTHROPOIETIN IN PATIENTS WITHMALIGNANCY OR CHEMOTHERAPY

Anemia related to malignancy can be secondary to impairedproduction related to the bone marrow replacement by cancer,


suppression of bone marrow by the direct effects of radiationtherapy and chemotherapy, or secondary to inflammatorycytokines such as TNF-a. Recent American Society of ClinicalOncology and American Society of Hematology guidelinesrecommend the use of EPO in chemotherapy related anemiawith hemoglobin concentrations of � 10g=dL (20). Use ofRBC transfusions is to be based on the clinical circumstancesand the severity of anemia. In patients with hemoglobin of10–12 g=dL, the decision of instituting erythropoietin therapyor red blood cell transfusions is left to clinical judgment. Therecommended initial dose is 150U=kg three times a week for aminimum of 4 weeks. In patients showing a poor initialresponse (< 1–2 g=dL increase in hemoglobin concentration),dose escalation can be done to 300U=kg three times a weekfor another 4–8 weeks. Once hemoglobin levels have reachedaround 12 g=dL, the EPO dosage should be titrated tomaintain that level of hemoglobin. Alternatively, it can berestarted when the hemoglobin levels fall to �10 g=dL. Analternative regimen of 40,000U weekly has been used in com-mon clinical practice although this has been supported byweaker evidence of efficacy. In the absence of increase ofhemoglobin values by a minimum of 1–2 g=dL, continuedEPO therapy after 6–8 weeks does not appear to be beneficial.

Recommended monitoring of erythroid response includesbaseline, 2–3week posttherapy initiation, and subsequentperiodic measurement of hemoglobin, iron, total iron-bindingcapacity, transferrin saturation, and ferritin levels. Iron reple-tion should be instituted when indicated (usually if the ferritinconcentration is <100mg=L). Iron therapy with EPO is impor-tant in limiting the need for EPO, improving symptoms, anddetermining failure of EPO responsiveness. The newer, long-acting form of EPO, darbepoetin is being used and has been stu-died but the level of evidence supporting its use is stillmaturing.

SIDE EFFECTS

Erythropoietin is well tolerated with the major reported sideeffect being an increase in blood pressure, especially in

374 Kaur et al.

patients with end stage renal disease (21). Rarely, seizures,thrombotic=vascular events, and hypertensive encephalopa-thy have been described; usually in the setting of excessiveincreases in red cell mass (22). Given the potential for anincreased risk of seizures during the first 90 days of therapy,blood pressure and neurologic status should be carefullymonitored. Patients should be advised to avoid potentiallyhazardous activities such as driving or operating machinery.The safety and efficacy of EPO in patients with a seizuredisorder have not been established; these patients should beclosely monitored. It is recommended that the dose of EPObe decreased if the hematocrit increase exceeds four pointsin any 2-week period.

The risk of thrombotic events, including myocardialinfarction, and mortality are significantly increased in adulthemodialysis patients with ischemic cardiac disease or con-gestive heart failure receiving epoetin alfa with a goal ofreaching a normal hematocrit (42%) as compared to a targethematocrit of 30% (22).

The increase in blood pressure is usually mild with lowdose EPO therapy and is controlled with drug therapy (23).It would be prudent to avoid excessive increases in hemoglo-bin and to monitor blood pressure initially, especially with apre-existent history of hypertension. The vehicle used in theformulation of EPO can produce burning at the injectionsite when administered subcutaneously, while intravenousadministration has been associated with a transient flushingor ‘‘flu-like’’ syndrome (11).

Treatment with epoetin alfa may result in functional orabsolute iron deficiency. Prior to and during EPO therapy,evaluation of iron stores, including transferrin saturation(serum iron divided by iron-binding capacity) and serum fer-ritin should be monitored. Transferrin saturation should beat least 20% and the serum ferritin should be at least100ng=mL. Virtually, all patients will require iron supple-mentation to increase or maintain saturation levels that willsupport erythropoiesis. Anemic patients with iron overloadmay respond well to the acute phase of EPO therapy, andmay benefit from EPO therapy and subsequent phlebotomy


once the hemoglobin rises to reduce total iron stores. All sur-gery patients should receive iron supplementation through-out the course of EPO therapy. EPO is not indicated inpatients with iron-deficiency anemia or anemias due to acuteor chronic blood loss (24).

Absolute contraindications to EPO therapy include:hypersensitivity to albumin, hamster protein, or benzyl alcoholhypersensitivity, and red cell aplasia associated with EPOtherapy.

SUMMARY

Recombinant human erythropoietin has been used effectivelyto increase hemoglobin in ACD. It must be stressed again,however, that correctable causes of anemia must be excludedprior to initiation of EPO. As health care spending continuesto rise, appropriate resource utilization has become an impor-tant issue in the daily practice of medicine. Finite medicalresources are available to all patients in the health caresystem, and appropriate allocation of these resources is theresponsibility of physicians. Though patients are concernedprimarily with their health and payers may wish to constrainhealth care expenditures, physicians and society must take aperspective that maximizes the allocation of health careresources for the benefit of all patients. An evidence-basedapproach may offer some solutions to this inherent conflict.Treatment of the underlying medical condition is moreeffective in achieving long-term cure of the anemia.

REFERENCES

1. Ganz T. Hepcidin, a key regulator of iron metabolism and med-iator of anemia of inflammation. Blood 2003; 102(3):783–788.

2. Jelkmann W. Erythropoietin: structure, control of production,and function. Physiol Rev 1992; 72(2):449–489.

3. Means RT Jr. Advances in anemia of chronic disease. Int JHematol 1999; 70:7–12.

376 Kaur et al.

4. Papadaki HA, Kritikos HD, Gemetizi C, Koutala H, Marsh JC,Boumpas DT, Eliopoulos GD. Bone marrow progenitor cellreserve and function and stromal cell function are defectivein rheumatoid arthritis: evidence for a tumor necrosis factoralpha (TNFa)-mediated effect. Blood 2002; 99:1610–1619.

5. Papadaki HA, Kritikos HD, Valatas V, Boumpas DT,Eliopoulos GD. Anemia of chronic disease in rheumatoidarthritis is associated with increased apoptosis of bone marrowerythroid cells: improvement following anti-tumor necrosisfactor-a antibody therapy. Blood 2002; 100(2):474–482.

6. Silva M, Benito A, Sanz C, Prosper F, Ekhterae D, Nunez G,Fernandez-Luna JL. Erythropoietin can induce expression ofBcl-xL through Stat5 in erythropoietin-dependent progenitorcell lines. J Biol Chem 1999; 274:22165–22169.

7. Egrie JC, Strickland TW, Lane J, Aoki K, Cohen AM,Smalling R, Trail G, Lin FK, Browne JK, Hines DK. Charac-terization and biological effects of recombinant humanerythropoietin. Immunobiology 1986; 172:213–224.

8. Graber SE, Krantz SB. Erythropoietin and the control of redcell production. Ann Rev Med 1978; 29:51–66.

9. Pincus T, Olsen NJ, Russell IJ, Wolfe F, Harris ER,Schnitzer TJ, Boccagno JA, Krantz SB. Multicenter study ofrecombinant human erythropoietin in correction of anemia inrheumatoid arthritis. Am J Med; 89:161.

10. Henry DH, Beall GN, Benson CA, Carey J, Cone LA, Eron LJ,Fiala M, Fischl MA, Sabin SJ, Gottlieb MS. Recombinanthuman erythropoietin in the treatment of anemia associatedwith human immunodeficiency virus (HIV) infection and zido-vudine therapy: overview of four clinical trials. Ann InternMed 1992; 117:739.

11. Spivak JL. Recombinant human erythropoietin and theanemia of cancer. Blood 1994; 84:997.

12. Fischl M, Galpin JE, Levine JD, Groopman JE, Henry DH,Kennedy P, Miles S, Robbins W, Starrett B, Zalusky R. Recom-binant human erythropoietin for patients with AIDS treatedwith zidovudine. NEJM 1990; 322:1488–1493.


13. Kaltwasser JP, Kessler U, Gottschalk R, Stucki G, Moller B.Effect of recombinant human erythropoietin and intravenousiron on anemia and disease activity in rheumatoid arthritis.J Rheumatol 2001; 28(11):2430–2436.

14. Peeters HR, Jongen-Lavrencic M, Vreugdenhil G, Swaak AJ.Effect of recombinant human erythropoietin on anemia anddisease activity in-patients with rheumatoid arthritis and ane-mia of chronic disease: a randomized placebo controlled doubleblind 52 weeks clinical trial. Ann Rheum Dis 1996; 55(10):739–744.

15. Goodnough LT, Marcus RE. The erythropoietic response toerythropoietin in patients with rheumatoid arthritis. J LabClin Med 1997; 130(4):354–356.

16. Matsuda S, Kondo M, Mashima T, Hoshino S, Shinohara N,Sumida S. Recombinant human erythropoietin therapy forautologous blood donation in rheumatoid arthritis patientsundergoing total hip or knee arthroplasty. Orthopedics 2001;24(1):41–44.

17. Best WR, Becktel JM, Singleton JW. Rederived values of theeight coefficients of the Crohn’s Disease Activity Index (CDAI).Gastroenterology 1979; 77:843–846.

18. Schreiber S, Howaldt S, Schnoor M, Nikolaus S, Bauditz J,Gasche C, Lochs H, Raedler A. Recombinant erythropoietinfor the treatment of anemia in inflammatory bowel disease.1996; 334(10):619–624.

19. Gasche C, Dejaco C, Waldhoer T, Tillinger W, Reinsch W,Fueger GF, Gangl A, Lochs H. Intravenous iron and erythro-poietin for anemia associated with Crohn’s disease. A rando-mized, controlled trial. Ann Intern Med 1997; 126:782–787.

20. Rizzo JD, Lichtin AE, Woolf SH, Seidenfeld J, Bennett CL,Cella D, Djulbegovic B, Goode MJ, Jakubowski AA, Lee SJ,Miller CB, Rarick MU, Regan DH, Browman GP, Gordon MS.Use of epoetin in patients with cancer: evidence-based clinicalpractice guidelines of the American Society of ClinicalOncology and the American Society of Hematology. J ClinOncol 2002; 20(19):4083–4107.

21. Maschio G. Erythropoietin and systemic hypertension.Nephrol Dialysis Transpl 1995; 10(suppl 2):74–79.

378 Kaur et al.

22. Dammacco F, Silvestris F, Castoldi GL, Grassi B, Bernasconi C,Nadali G, Perona G, De Laurenzi A, Torelli U, Ascari E, RossiFerrini PL, Caliaris-Cappio F, Pileri A, Resegotti L. The effec-tiveness and tolerability of epoietin alfa in patients with multi-ple myeloma refractory to chemotherapy. Int J Clin Lab Res1998; 28:127–134.

23. Silverberger DS, Iaina A, wexler D, Blum M. The pathologicalconsequences of anaemia. Clin Lab Haem 2001; 23:1–6.

24. Cazzola M, Mercuriali F, Bruganara C. Use of recombinanthuman erythropoietin outside the setting of uremia. Blood1997; 89(12):4248–4267.


15

Iron Therapy and the Anemiaof Chronic Disease

VICTOR R. GORDEUK

Department of Medicine, Center for SickleCell Disease, Howard University,

Washington, D.C., U.S.A.

INTRODUCTION

For most patients with the anemia of chronic disease, no spe-cific therapy is needed. Rather, treatment should be directedto the underlying disorder. The anemia typically is mild tomoderate and self-limited, and resolves with correction ofthe underlying inflammatory disorder if this is possible. Bloodtransfusion is seldom warranted. In the absence of coexistingiron deficiency or therapy with erythropoietin, neither oralnor parenteral iron is of any benefit.

381

INEFFECTIVENESS AND POTENTIALHARM OF ROUTINE IRONTHERAPY IN THE ANEMIA OFCHRONIC DISEASE

Microcytosis and hypochromia may develop in 30–50% ofpatients with the anemia of chronic disease (1), and hypofer-remia is a typical feature of acute or chronic inflammation aswell as the anemia of chronic disease (2). These featuresoverlap with the laboratory findings of iron deficiency ane-mia, and therefore it is common to see a misdiagnosis of irondeficiency and unnecessary treatment with iron in patientswith the anemia of chronic disease. Considering the trans-ferrin concentration, which tends to be high in iron defi-ciency and low in the anemia of chronic disease, and theserum ferritin concentration, which tend so be low in irondeficiency and elevated in the anemia of chronic disease,can help to arrive at the proper diagnosis and prevent unne-cessary iron therapy (see Chapter 11 in this book).

A number of considerations militate against the routineuse of iron therapy in the anemia of chronic disease. (1)Absorption of iron tends to be reduced (see Ref. 3) and ery-throid precursors have downregulation of transferrin receptorexpression (see Ref. 4) in the setting of inflammation, makingit unlikely that iron therapy will be effective in raising thehemoglobin concentration. (2) Iron is an essential growth fac-tor for all cells, including pathogenic microorganisms andmalignant cells (see Ref. 5). Any increase in the availabilityof iron in the serum, interstitial fluid, and cells because of irontherapy may therefore enhance the growth of microorganismsand malignant cells. (3) Iron-loading of macrophages, a typi-cal feature of the anemia of chronic disease, may impair Th-1-mediated immune pathways that are important in fightinginfections and malignancy (see Refs. 6, 7). An increase ofmacrophage iron stores because of iron therapy could there-fore conceivably contribute to infectious risk or proliferationof malignant cells.

382 Gordeuk

IRON DEFICIENCY

Iron deficiency is the most common nutritional deficiencyaround the world. The condition is particularly prevalentamong women of childbearing age, as well as among infantsand adolescents whose rapid growth requires an enhancediron supply to support a concomitant increase in red cellmass. Iron deficiency occurring in men or in nonmenstruatingwomen usually is the result of blood loss, most often from thegastrointestinal tract, and it is imperative to seek the specificsite and cause.

When iron requirements or the loss of iron exceed thequantity of iron absorbed, the individual experiences a stateof negative iron balance. With this negative balance, ironstores decrease progressively and synthesis of hemoglobin isimpaired after storage iron is exhausted. Lack of body ironcan therefore be divided into two categories. In iron deficiencywithout anemia, storage iron is absent but the deficit in ironis not sufficiently large to decrease the hemoglobin below thenormal level. In iron deficiency anemia, the deficit in iron is sosevere that stores are absent and the hemoglobin is franklybelow the normal range (1,8).

Causes of Iron Deficiency

Inadequate Dietary Iron for High PhysiologicRequirements

Iron deficiency is themost commonnutritional deficiency in theworld. Infants, adolescents, andwomen of childbearing age areat the highest risk of iron deficiency and may become iron defi-cient because of inadequate dietary iron to meet physiologicneeds. Infants 1–2 years of age and adolescents have increasediron requirements due to the demands of rapid growth.Womenneed more iron because of the loss of hemoglobin withmenstruation and the high iron demands of pregnancy (1,8).

Blood Loss

Other than in children and women of childbearing age, thefinding of iron deficiency almost always signifies pathological

Iron Therapy and ACD 383

blood loss of some sort. Populations with hookworm infesta-tion have an increased incidence of iron deficiency due tochronic intestinal blood loss. Individuals with gastrointestinalbleeding due to ulceration, tumors, diverticuli, polyps, andvascular malformations are prone to develop iron deficiency.Both women and men who are frequent blood donors have ahigh prevalence of iron deficiency. About 50% of hemodialysispatients develop iron deficiency due to blood lost during theprocedure and frequent diagnostic tests. The finding of irondeficiency should always prompt careful clinical considerationof potential underlying causes such as increased blood loss inmenstruating women and chronic bleeding due to esophagealreflux disease, ulcerative diseases of the stomach or duode-num, inflammatory bowel disease, venous malformations ofthe gut, or adenomas of the colon. Moreover, the patientmay have a gastrointestinal cancer (esophageal, gastric, orcolorectal are the most common), and intensive efforts shouldbe made to find a potentially curable malignancy. Hemoptysisand bleeding from a bladder tumor may lead to iron deficiencyas well (1,8).

Malabsorption

Iron deficiency, often of a severity to cause anemia, occurs inover 1=3 of childhood and adult cases of both symptomatic andasymptomatic celiac disease. Iron deficiency results frommalabsorption and GI mucosal blood loss. Some studies indi-cate that iron deficiency may be more common in adults thanchildren with celiac disease (9,10).

Iron deficiency is prone to develop in inflammatory boweldisease because of chronic blood loss from the GI mucosa,inhibition of iron absorption related to the systemic inflam-matory process, and possibly reduced absorption related toreplacement of normal mucosa in the proximal small bowelwith the disease process. In one study, 65% of patients withCrohn’s disease or ulcerative colitis and anemia respondedto parenteral iron therapy with an increase in the hemoglobinconcentration of 2 g=dL or more (11). Iron deficiency also maydevelop in patients with defective absorption of food iron due

384 Gordeuk

to tropical sprue and other tropical enteropathies, achlor-hydria, and gastric resection.

Hemoglobinuria, PulmonaryHemosiderosis

Iron deficiency may develop in the setting of intravascularhemolysis with associated hemoglobinuria–paroxysmal noc-turnal hemoglobinuria and anemia in long-distance runnersare examples. Alveolar hemorrhage as a reflection of vasculitis(e.g., Wegener’s granulomatosis, Churg-Strauss syndrome),other auto-immune disorders (e.g., Goodpasture syndrome) orvascular malformations (e.g., Osler’s disease) leads to iron-loading of pulmonary macrophages; because this iron cannotbe utilized for hemoglobin synthesis, the process can lead toiron deficiency (1,8).

Clinical Features of Iron Deficiency

These include fatigue, irritability, headaches, paresthesias,glossitis (a smooth red tongue), angular cheilitis, pallor, andkoilonychia (spooning of the nails). Pica, the craving to eatunusual substances such as ice, clay, or dirt, can be a charac-teristic feature.

Symptoms and Signs

Iron deficiency in pregnant women may be associated withlow birth weight (12) and anemia in the first year of life(13) of their children. Iron deficiency anemia in infancy maybe associated with impaired mental and motor developmentat 5 years of age (14). In adults, iron deficiency anemia is asso-ciated with reduced physical work capacity (15). Pica, a crav-ing for and the ingestion of clay, starch, or ice, is a commonfinding in iron deficiency.

Physical Findings

Physical findings of iron deficiency may include koilonychia,papillary atrophy of the tongue, angular stomatitis, esophagealweb, achlorhydria, and gastritis. Slight splenic enlargement,cause unknown, may occur in about 10% of patients (1).


Laboratory Findings

Iron deficiency in the absence of anemia is characterized bylow serum ferritin concentration (< 12 mg=L). With iron defi-ciency anemia, the red cell count, hemoglobin concentration,and hematocrit are below the normal range, review of theperipheral blood smear shows hypochromia and microcytosis,and there is a low or normal reticulocyte count. Iron defi-ciency is also typically accompanied by reductions in serumiron concentration and transferrin saturation [ratio of serumiron to total serum iron binding capacity (TIBC)]. The condi-tion is typically associated with elevations in red cell distribu-tion width, free erythrocyte protoporphyrin concentration,transferrin concentration, TIBC, and circulating transferrinreceptors (Tables 1 and 2). Typically, the automated bloodcount reveals decreased mean corpuscular volume (< 80 fl)and increased red cell distribution width (> 15%).

The serum ferritin and transferrin saturation are unreli-able tests for iron deficiency if an inflammatory condition or ahepatocellular process is present. Ferritin is a positive acutephase reactant and serum levels may be in the normal rangein a patient with both lack of iron and an inflammatoryprocess. Also, increased amounts of ferritin appear in theserum with hepatocellular damage even in the absence of irondeficiency. Typically, the transferrin saturation is less than10% (normal 20–50%) with iron deficiency anemia, but thismeasurement is also decreased in the presence of acute andchronic inflammation and it is raised with hepatocellular

Table 1 Typical Changes in Measures of Iron Status in IronDeficiency, Inflammation, and Iron Overload

ConditionSerumiron

Total ironbindingcapacity

Transferrinsaturation

Serumferritin

Serumtransferrinreceptor

Irondeficiency

Decreased Increased Decreased Decreased Increased

Inflammation Decreased Decreased Decreased Normal toincreased

Normal

Iron overload Increased Decreased Increased Increased Decreased

386 Gordeuk

damage (e.g., hepatitis due to viruses or drugs) and marrowdysfunction (e.g., due to aplasia, cancer chemotherapy or amegaloblastic process). Transferrin saturation can also beaffected by diurnal variations, typically being higher in themorning and lower in the evening.

Early in the course of iron deficiency, the anemia may benormochromic and normocytic, but, in the absence of othersystemic processes, the transferrin saturation and serum fer-ritin will be low. In many parts of the tropics, there is a highprevalence of thalassemia trait, which also causes a mildmicrocytic, hypochromic anemia and which may be difficultto distinguish from iron deficiency anemia. In an unexplainedanemia with the serum ferritin> 12 mg=L, the bone marrowcan be stained with Perl’s reagent to reveal the presence orabsence of iron; if the anemia is the result of iron deficiency,iron stores are absent. The differential diagnoses for lowtransferrin saturation and serum ferritin concentration areshown in Table 3 (1,8).

Management of Iron Deficiency Anemia

Identify and Correct the Cause

Iron deficiency as the cause of anemia is suspected onepidemiologic grounds. As discussed above, the condition isa common cause of low hemoglobin in infants, adolescents,women of childbearing age, individuals with hookworminfestation, and patients with chronic blood loss from the

Table 2 Typical Changes in the Complete Blood Count with IronDeficiency Anemia and the Anemia of Inflammation

Condition

Degreeof

anemia

Meancorpuscular

volume

Red celldistribution

width

Whitebloodcells Platelets

Iron deficiencyanemia

Mild tosevere

Decreased Increased Normal Normal toincreased

Inflammation Mild tomoderate

Normal todecreased

Normal Normal toincreased

Normal toincreased


gastrointestinal tract or other sites. Before treatment, consid-eration should be given to the cause of iron deficiency, whichmay be deduced from history and physical examination orexamination of the stool for occult blood and parasites. Ifpossible, the underlying cause should be corrected. A searchmust be made for the presence of a potentially curablemalignancy giving rise to blood loss (1,8).

Treat with Iron

Iron treatment in the setting of the anemia of chronic diseaseis discussed in the next section.

IRON THERAPY FOR PATIENTSWITH COMBINED IRON DEFICIENCYAND THE ANEMIA OF CHRONIC DISEASE

Iron therapy is indicated for patients who have the combina-tion of iron deficiency and the anemia of chronic disease. It isvery important that iron therapy is not given in isolation. Thephysician must also conduct a thorough investigation for thecause of the iron deficiency and must take steps to treat orcorrect the cause of iron deficiency that is identified. Iron defi-ciency in men or nonmenstruating women is most often theresult of gastrointestinal blood loss, and an occult malignancymay be identified and treated while curable if the first step inthe management of iron deficiency is to identify the cause of

Table 3 Differential Diagnoses of Indirect Measures of IronDeficiency

Measurement Differential diagnoses

Low serum ferritin concentration Iron deficiency

Low serum iron concentration and Iron deficiencytransferrin saturation Inflammation

Diurnal variation (evening bloodsample)

388 Gordeuk

the lack of iron. The physician must also diagnose and treatthe underlying inflammatory or malignant process associatedwith the anemia of chronic disease.

The diagnosis of iron deficiency in the setting of inflam-mation is challenging, for the serum ferritin concentrationmay be well within the normal range rather than decreased(see chapter 11 for a discussion of the diagnosis of iron defi-ciency in the setting of inflammation). In general, a serum fer-ritin concentration greater than 100 ng=mL makes it unlikelythat the patient is iron deficient unless hepatocellulardamage is present. An elevated serum transferrin receptorconcentration can be helpful in making the diagnosis of irondeficiency if it is certain that the patient does not have hemo-lysis or ineffective erythropoiesis. In a patient with both irondeficiency and inflammation, the transferrin concentrationwill not usually be elevated and it may be decreased; in con-trast, with pure iron deficiency the transferrin concentrationis often elevated. It may be necessary to stain a bone marrowaspirate for iron (the gold standard for diagnosis of irondeficiency) to definitively determine whether macrophageiron stores are present or absent (1,8).

Oral Iron Therapy

The treatment of choice for uncomplicated iron deficiency isalmost always an oral iron preparation, which effectivelyreplenishes iron stores in the vastmajority of cases. A parenteraliron preparation is rarely indicated except in certain specialsituations. For many years, the most widely recommended oraliron agent has been ferrous sulfate, which is inexpensive, wellabsorbed, and generally well tolerated. The typical adult doseof iron is 200mg of elemental iron per day, for example one300mg ferrous sulfate tablet three times a day (300mg ferroussulfate¼ 65mg elemental iron). The pediatric dose is 5mg=kgof elemental iron per day in tablet or elixir form.

Iron is best absorbed when given without food. Ironabsorption is enhanced in patients with iron deficiencyanemia, and this initially facilitates assimilation of medicinaliron, although absorption declines with correction of the ane-


mia and reaccumulation of iron stores. Oral iron therapyusually raises the hemoglobin concentration to normal levelswithin four to eight weeks at the rate of about 1 g=dL perweek. Treatment should be continued for an additional threeto six months to replenish iron stores, as indicated by a returnof the serum ferritin concentration to a normal range (50–100mg=L), at which time iron therapy should be discontinued.Iron is best absorbed if given without food.

Side effects of iron therapy include constipation, diarrhea,nausea, and abdominal pain, and if these limit compliance, themedication can be administered with food or the dose reducedby one-half. One 300mg ferrous sulfate tablet nightly at bed-time is an effective therapy, although it may be necessary tocontinue treatment for two to three times as long to achievethe same effect as with full doses. A number of other types ofiron salts have been studied and developed to identify prepara-tions that are better absorbed or are associated with fewer sideeffects, but these products generally do not offer significantimprovements and often are more expensive.

All iron tablets must be kept out of the reach of youngchildren, because iron ingestion is a common cause of poison-ing and as little as 10 ferrous sulfate tablets can prove fatalfor an infant. Carbonyl iron is an elemental iron powder thatmay be given in capsule form and has similar efficacy to ironsalts in correcting iron deficiency. Elemental iron has theadvantage of having remarkably reduced toxicity when com-pared to iron salts, and it would be reasonable to considerthe substitution of carbonyl iron for therapeutic uses (1,8,16).

Parenteral Iron Therapy

In patients who fail to respond to oral iron therapy, the mostcommon reason is poor compliance as a result of gastrointest-inal side effects, which may cause the patient to refusefurther treatment. In many cases, iron therapy can beresumed after discussion with the patient about ways tominimize symptoms by dose reduction and taking the medica-tion with food, as noted above. Rarely, patients may have atrue inability to absorb iron, most often as a result of prior

390 Gordeuk

gastrectomy or celiac disease. In addition, because of reducediron absorption related to inflammation, some patients withiron deficiency and the anemia of chronic disorders mightbenefit from parenteral iron therapy (1,8,16).

Three forms of parenteral iron are now available in theUnited States, iron dextran (INFeD) and iron sucrose pre-parations [sodium ferric gluconate complex (FERRLECIT)],and ferric hydroxide complex (VENOFER). Until recently,the only parenteral iron preparation available in the UnitedStates had been iron dextran (INFeD). Iron dextran, whethergiven intramuscularly or intravenously, does not lead to amore rapid hematologic response than oral iron and carriesthe risk of anaphylactic reactions. The presence of iron defi-ciency must be rigorously proved before embarking on parent-eral iron therapy, for its administration to patients with otherforms of anemia can lead to iron overload. Here, we willconsider two parenteral iron preparations in further detail,iron dextran and sodium ferric gluconate.

Iron Dextran

Each 1mL of iron dextran injection, USP contains the equiva-lent of 50mg of elemental iron (as an iron dextran complex).Some patients receiving parenteral iron dextran therapyexperience severe allergic reactions, including anaphylaxis,and all patients should be monitored carefully. The manufac-turer recommends that the maximum daily dose of iron dex-tran should not exceed 2.0mL of undiluted iron dextran(100mg of elemental iron). The preparation should be usedwith extreme care in patients with serious impairment of liverfunction, and it should not be given during acute infection ofthe kidney. Before administering iron dextran, the total doseneeded to both correct anemia and replace storage iron shouldbe calculated according to the following formula:

Iron to be injected (mg)

¼ ð15� patient’s hemoglobin[g/dL])

� body weight kg)�3


For adults, this total amount of iron should be dividedinto 100mg (2.0mL) increments and administered dailyeither intramuscularly or intravenously until the total calcu-lated amount is given. For children, smaller daily doses areused. Before starting therapy, a small test dose of 0.5mL(25mg) should be given to exclude the possibility of an allergicresponse. Although anaphylactic reactions after iron dextranadministration are known to occur within a few minutes orsooner, it is recommended that at least 1 hr be allowed toelapse before the remainder of the initial therapeutic dose isgiven. The physician must be immediately available duringthis observation period, and resuscitation capability (includ-ing parenteral steroids and epinephrine) must be near athand. When administered intravenously, iron dextran shouldbe given undiluted at a slow gradual rate not to exceed 50mg(1mL) per minute. When administered intramuscularly, irondextran should be injected only into the muscle mass of theupper outer quadrant of the buttock and injected deeply witha 2- or 3-in. 19- or 20-gauge needle using a Z-track techniqueto minimize the chance of subcutaneous leakage and disco-loration. Side effects include arthalgias or myalgias occurringthe day after injection; these usually respond to a mildanalgesic such as acetaminophen. Often physicians considerit impractical to administer iron dextran in divided dosesand resort instead to total dose infusion, but this is not recom-mended by the manufacturer (16). For parenteral iron admin-istration, many physicians now favor using ferric gluconatecomplex instead, as discussed below.

Sodium Ferric Gluconate Complex

In 1999, sodium ferric gluconate (iron gluconate, FERRLE-CIT) received FDA approval for iron replacement therapyby parenteral administration. This form of iron has beenavailable outside the United States, principally in Europe,for several decades. Most of the recent reported experiencehas been in combination with erythropoietin for maintenanceof an adequate blood hemoglobin level in patients receivinglong-term hemodialysis for end stage renal disease. Recent

392 Gordeuk

clinical trials in the United States have confirmed the safetyand efficacy of this approach. Iron gluconate similarly maybe considered for other applications requiring parenteraliron therapy that traditionally have been treated with irondextran. The risk for allergic reactions, including full-blownanaphylaxis, is much lower with iron gluconate than iron dex-tran. In a typical course of treatment for patients on hemodia-lysis receiving recombinant human erythropoietin therapy,sodium ferric gluconate complex in sucrose is administeredin a dose of 125 mg=dose (62.5 mg=dose in children under40kg) by slow intravenous infusion on eight consecutive days,for a total dose of 1.0 g (0.5 g in children). In other contexts, itis appropriate to calculate the total dose of iron to be given, inthe same way as outlined above for iron dextran. For example,an adult with a calculated deficit of 2.0 g who is to be treatedwith ferric gluconate complex would require 16 infusions of125 mg=dose to deliver the required amount of iron. Such aregimen is more practical for patients on hemodialysisfor whom access to the circulation is a frequent event thanit is for patients requiring iron repletion in other settings.Unfortunately, higher doses or more rapid infusion ratesmay be associated with side effects (hypotension, malaise)accompanied by high serum iron concentrations and transfer-rin saturations above 100%, raising the concern of possibleiron toxicity. Additional clinical experience should helpclarifythe role for this new product in specific clinical situations (16).

IRON THERAPY IN ANEMIA OF CHRONICDISEASE PATIENTS RECEIVINGERYTHROPOIETIN

Some patients with relatively low erythropoietin values forthe degree of anemia may benefit from administration ofrecombinant human erythropoietin, sometimes in associationwith intravenous iron. Examples are rheumatoid arthritis,inflammatory bowel disease, AIDS-associated anemia. In con-trast, iron therapy for anemia of chronic disease in the case ofmalignancy or infectious diseases appears to be generally


associated with an unfavorable course of the disease due to itspotential to promote pathogen growth, to adversely affectimmune effector pathways, or to generate highly toxic freeradicals (please refer to Chapters 9, 10 and 19 in this book).These unfavorable effects of iron therapy may also apply foranemia of chronic disease in combination with true iron defi-ciency, because tumor cells and microorganisms have sophis-ticated strategies to acquire iron.

In certain circumstances, it is conceivable that thecombination of parenteral iron with erythropoietin may beof benefit, due to the possibility that erythropoietin can stimu-late porphyrin biosynthesis and the expression of transferrinreceptors by erythroid cells and thus promote iron uptake andheme biosynthesis by these cells (17). Prospective studies areneeded to clarify if this is true in the clinic, in particular,whether or not the potential detrimental effects of increasediron availability on the course of an infectious or tumor dis-ease are outweighed by an overall benefit for the patient(please refer also to the chapters ‘‘Human recombinanterythropoietin’’ by Lichtin and ‘‘Iron and erythropoietin’’ byGoodnough in this book).

ACKNOWLEDGMENTS

This work was supported in part by NIH research grant no.UH1-HL03679–05 from the National Heart, Lung and BloodInstitute and the Office of Research on Minority Health andbyHoward University General Clinical Research Center grantno. MO1-RR10284.

REFERENCES

1. LeeGR. The anemia of chronic disorders. In: LeeGR,Bithell TC,Foerster J, Athens JW, Lukens JN, eds. Wintrobe’s ClinicalHematology. Philadelphia: Lea & Febiger, 1993:840–851.

2. Cartwright GE. The anemia of chronic disorders. SeminHematol1966; 3:351–375.

394 Gordeuk

3. Jensen NM, Brandsborg M, Boesen AM, Yde H, Dahlerup JF.Low-dose oral iron absorption test in anaemic patients withand without iron deficiency determined by bone marrow ironcontent. Eur J Haematol 1999; 63:103–111.

4. Kuiper-Kramer EP, Huisman CM, Van der Molen-Sinke J,Abbes R, vanEijk HG. The expression of transferrin receptorson erythroblasts in anemia of chronic disease, myelodysplasticsyndromes and iron deficiency. Acta Haematol 1997; 97:127–131.


6. Brock JH. Iron in infection, immunity, inflammation andneoplasia. In: Brock JH, Halliday JW, Pippard MJ, PowellLW, eds. Iron Metabolism in Health and Disease. London:WB Saunders Co. 1994:353–389.

7. Weiss G, Wachter H, Fuchs D. Linkage of cellular immunity toiron metabolism. Immunol Today 1995; 16:495–500.

8. Brittenham GM. Disorders of iron metabolism: iron deficiencyand iron overload. In: Hoffman R, et al., eds. Hematology:Basic Principles and Practice. 3rd ed. New York: ChurchillLivingstone, 2000:397–428.

9. Bonamico M, Vania A, Monti S, Ballati G, Mariani P, Pitzalis G,Benedetti C, Falconieri P, Signoretti A. Irondeficiency in childrenwithceliacdisease. JPediatrGastroenterolNutr1987;6:702–706.

10. Bottaro G, Cataldo F, Rotolo N, Spina M, Corazza GR. Theclinical pattern of subclinical=silent celiac disease: an analysisof 1026 consecutive cases. Am J Gastroenterol 1999; 94:691–696.

11. Gasche C, Waldhoer T, Feichtenschlager T, Male C, Mayer A,Mittermaier C, Petritsch W. Austrian Inflammatory BowelDiseases Study Group. Prediction of response to iron sucrosein inflammatory bowel disease-associated anemia. Am JGastroenterol 2001; 96:2382–2387.

12. Cogswell ME, Parvanta I, Ickes L, Yip R, Brittenham GM. Ironsupplementation during pregnancy, anemia, and birth weight:a randomized controlled trial. Am J Clin Nutr 2003; 78:773–781.


13. Kilbride J, Baker TG, Parapia LA, Khoury SA, Shuqaidef SW,Jerwood D. Anaemia during pregnancy as a risk factor foriron-deficiency anaemia in infancy: a case-control study inJordan. Int J Epidemiol 1999; 28:461–468.

14. Lozoff B, Jimenez E, Wolf AW. Long-term developmentaloutcome of infants with iron deficiency. N Engl J Med 1991;325:687–694.

15. Haas JD, Brownlie T. Iron deficiency and reduced work capa-city: a critical review of the research to determine a causalrelationship. J Nutr 2001; 131:676S–690S.

16. McLarenGD.GordeukVR, Irondeficiency. In:RakelRE,BopeET,eds. Conn’s Current Therapy. Philadelphia: WB Saunders Co.,2002:355–359.

17. Weiss G, Houston T, Kastner S, Grunewald K, Brock J. Regu-lation of cellular iron metabolism by erythropoietin: activationof iron regulatory protein and up-regulation of transferrin-receptor in erythroid cells. Blood 1997; 89:680–687.

396 Gordeuk

16

Blood Transfusions

ELEFTHERIOS C. VAMVAKAS

Division of Medical, Scientific and ResearchAffairs, Canadian Blood Services; and Department

of Pathology and Laboratory Medicine,University of Ottawa Faculty of Medicine,

Ottawa, Canada

With the exception of the anemia associated with chronicrenal failure, disseminated malignancy, and human immuno-deficiency virus (HIV) infection, anemia secondary to chronicsystemic diseases is not severe, and it presents with hemoglo-bin values in the range of 9–11g=dL. A hemoglobin level ofless than 8 g=dL indicates presence of one or more aggravat-ing factors, such as blood loss, acute infection, or adverse drugeffects. Thus, in anemia associated with chronic inflammation,liver disease, or endocrine failure, blood transfusion is indi-cated rarely, because the anemia is not severe. Therapeuticefforts are directed toward correcting the underlying disorder.

397

When an aggravating factor results in increasinganemia, a brief period of transfusion therapy may be useful.In general, however, most patients with anemia of chronicdisease are significantly limited by their primary disease,and transfusion for the purpose of improving the patient’sfunctional status offers little benefit. Transfusions givenunder such circumstances should be carefully evaluated fortheir objective clinical effect. The risks of transfusion shouldbe balanced against the value to the patient of whatever func-tional improvement is produced. Therefore, the anemia ofchronic disease represents a special clinical setting foradministering blood transfusions: because transfusion is notexpected to have the objective (or even life-saving) effect thatit has in other situations, all transfusion risks have to be care-fully considered. Such consideration should include emerginginfectious risks, as well as theoretical risks, and purportedadverse effects of transfusion.

The anemia of chronic renal failure can be very severe,with hemoglobin levels as low as 4 g=dL. Although patientsoften tolerate such marked anemia fairly well thank to com-pensatory adjustments, in the past as many as 25% of patientsrequiring chronic renal dialysis were transfusion-dependent.The development of recombinant human erythropoietin hasenabled definitive treatment for the anemia of chronic renalfailure. Presently, erythropoietin is administered in sufficientdoses to maintain the patient’s hemoglobin concentrationbetween 11.0 and 12.0 g=dL. In the initial phase I and II stu-dies of recombinant human erythropoietin in patients withchronic renal failure, doses of 50units=kg or more eliminatedtransfusion requirements in all patients (1). In a multicenterphase III trial, 97.4% of 333 anemic patients with renal failureachieved hematocrits of at least 35% within 18 weeks of start-ing erythropoietin therapy (2). Based on these figures, it hasbeen estimated that the use of recombinant human erythro-poietin in transfusion-dependent patients with end-stagerenal disease can eliminate 500,000 transfusions annually inthe United States [or 3% of all red blood cell (RBC) use] (3).

Recombinant human erythropoietin has also replacedRBC transfusion in selected patients with cancer,

398 Vamvakas

myelodysplasia, and HIV-related anemia. Patients with HIVinfection who have depressed erythropoietin levels benefitfrom the administration of recombinant human erythropoie-tin, but patients with elevated erythropoietin levels oftenrequire blood transfusions. The Viral Activation TransfusionStudy (VATS) enrolled 531 patients who were both HIV-and cytomegalovirus (CMV)-seropositive at the time that theyreceived their first RBC transfusion for HIV-related anemia.The patients were followed for a median period of 12 months(24 months in survivors), and they received a total of 3,864RBC units (4).

This chapter reviews the guidelines for RBC transfusiontherapy in chronic anemia, describes the established risks ofallogeneic blood transfusion (ABT), and discusses some emer-ging, theoretical, and=or purported risks of ABT. Specialemphasis is placed on the current controversy whether ABTmodulates the recipient’s immune function, predisposingto cancer recurrence, bacterial infection, and=or increasedmortality (5). The question whether ABT activates endogen-ous HIV and=or CMV infection in patients with HIV-relatedanemia of chronic disease (4,6) is also discussed.

RBC TRANSFUSION THERAPY IN CHRONICANEMIA

Multiple physiologic mechanisms exist to compensate for thedecrease in oxygen-carrying capacity associated with anemia.In patients without pulmonary, cardiac, coronary, cerebrovas-cular, or peripheral vascular disease, hemoglobin concentra-tions of approximately 7–9 g=dL are tolerated, albeit withdiminished capacity for physical activity. In patients withimpairment of these organs or tissues, higher hemoglobin con-centrations may be necessary. The goal of RBC transfusiontherapy in patients with chronic anemia is to avoid the compli-cations associated with anemia while balancing the risksattributed to transfusion. When required, RBC transfusionsshould be given on a unit-by-unit and case-by-case basis. Theexpected outcome is avoidance of the ongoing signs and symp-toms of anemia (7).

Blood Transfusions 399

A unit of whole blood consists of approximately 450mL ofblood with about 65 g of hemoglobin. ‘‘Packed’’ RBCs areprepared by centrifuging whole blood and expressing theplatelet-rich plasma into a satellite bag. The RBCs stored inCPDA-1 (citrate phosphate dextrose adenine-1) anticoagu-lant-preservative solution are concentrated to a hematocritof 80% or more, and have a volume of about 200–250mL,including up to 70mL of plasma. The RBCs stored in an addi-tive solution (e.g., Adsol) have a volume of approximately300mL, including 100mL of additive solution and some10–15mL of plasma. The RBCs are stored in the refrigeratorat 4�C, and they remain viable for 35–42 days, depending onwhether they are stored, respectively, in CPDA-1 or Adsol.Each unit of RBCs elevates an adult’s hemoglobin concentra-tion by 1 g=dL (or elevates the hematocrit by three percentagepoints). In the absence of hemolysis, the hemoglobin andhematocrit values of patients who have no erythropoiesisdecline by a similar amount per week.

Guidelines for RBC Transfusion Therapy

Several professional organizations have published guidelinesfor RBC transfusion therapy (8–11). These guidelines observethat chronic anemia is tolerated better than acute anemia (8)and that some patients with chronic anemia tolerate hemoglo-bin concentrations of less than 7 g=dL (9). In the absence ofsigns and symptoms of anemia, transfusions should be avoided,regardless of hemoglobin level. If symptoms occur, transfu-sions should be administered on a case-by-case basis (10).

Otherwise healthy people at rest have few signs or symp-toms of anemia when the hemoglobin concentration is greaterthan 7–8 g=dL. Dyspnea with exertion occurs at a hemoglobinconcentration of 7–8 g=dL, weakness at a hemoglobin of6 g=dL, dyspnea at rest at a hemoglobin of 3 g=dL, and conges-tive heart failure at a hemoglobin of 2–2.5 g=dL. In developinga transfusion plan for an individual patient, physicians must:(i) assess the symptoms caused or aggravated by the anemia,(ii) determine whether these symptoms are alleviated bytransfusion, (iii) pinpoint the minimal hemoglobin level atwhich the patient functions satisfactorily, and (iv) evaluate

400 Vamvakas

ABT’s risk-to-benefit ratio for the particular patient. Suchevaluation includes consideration of lifestyle factors, coexis-tence of other medical disorders, the likely duration ofanemia, and the patient’s prognosis (11).

The preceding guidelines reflect observed data and expertopinion. These recommendations (8–11) have been made inthe absence of results from blinded, randomized controlledtrials (RCTs). Therefore, criticisms of the guidelines haveemphasized a need to prevent symptoms of anemia (ratherthan waiting for them to occur) as well as a need to select astrategy that minimizes the complexity of the decision-makingprocess for prescribing RBC transfusions (12).

In a double-blind RCT, 118 patients received placebo ortwo different erythropoietin doses. When the effects of treat-ment were assessed, the hemoglobin concentration in the pla-cebo arm was 7.4 g=dL compared to 10.2 g=dL and 11.7 g=dLin the two arms receiving erythropoietin. Quality-of-lifeindicators did not differ between the two arms receiving ery-thropoietin, but patients treated with erythropoietin hadsignificantly improved scores for fatigue, physical symptoms,relationships, and depression compared to patients from theplacebo arm (13). Another double-blind RCT that enrolledpatients with malignancies found no statistically significantimprovement in the overall quality of life of patients receivingerythropoietin as compared to placebo, although energy levelsand ability to perform daily tasks did improve in the erythro-poietin arm. At baseline, the hematocrit was 29% in botharms, and the hematocrit increased by at least six percentagepoints in 58% of the erythropoietin-treated patients (14).

Such studies show an improvement in some subjectiveand objective quality-of-life measurements as hemoglobinconcentrations increase in chronically ill patients. Theseresults imply that a significant variance exists between theminimal hemoglobin level needed to maintain oxygen-carrying capacity (8–11) and that needed for many daily activ-ities (13,14). In light of the risks of ABT, however, the formu-lation of a plan for chronic transfusion therapy must focus onthe minimal hemoglobin concentration that achieves anacceptable level of oxygen delivery to tissues (7–11).


In 1997, the University Health System Consortium,which represents more than 100 academic medical centersin the United States, reported that 59% of its member institu-tions used a hemoglobin of 8 g=dL as the trigger for RBCtransfusion in patients with chronic anemia; 25% used ahemoglobin of 7 g=dL, and 16% used a hemoglobin of 9 g=dL.Age, gender, rapidity of onset of anemia, physiologic adapta-tions, cardiopulmonary function, history of ischemic comor-bidities, and signs and symptoms of anemia made up themajor criteria for considering the need for transfusion. Only11% of member institutions used the hemoglobin value aloneas a justification for transfusion (15).

ESTABLISHED NONINFECTIOUS RISKS OF ABT

Transfusion reactions are classified as immune vs. nonimmune.Antibody-mediated reactions are directed against RBCs, whiteblood cells (WBCs), platelets, and at least one class of immuno-globulin—IgA. The major nonimmune complications of ABTinclude circulatory overload, hypothermia, iron overload, andtransmission of infectious agents (Table 1). Transmission ofinfectious agents is considered later in this chapter.

Nonimmune Complications of ABT

Especially in patients with chronic anemia and expandedplasma volume, circulatory overload can be the most commoncomplication of ABT. Rapid infusion of blood—especially intoa small, elderly patient with heart disease—can precipitatecongestive heart failure and pulmonary edema, because ofincreased central venous pressure and increased pulmonaryblood volume. One percent of patients undergoing hip or kneearthroplasty in one study developed post-transfusion circula-tory overload (16). Patients who developed overload wereolder than those who did not develop this complication (87vs. 77 years, respectively). They had a modest intraoperativeblood loss (mean, 75mL) and a transfusion requirement ofonly 1 or 2 RBC units. Each patient was in positive fluid bal-ance (mean, 2.5L) before the transfusion that triggered the

402 Vamvakas

reaction. Autologous blood was just as hazardous as allo-geneic blood with regard to circulatory overload. In a previousstudy of all transfusion recipients, 1 in 708 patients receivingRBC transfusions developed post-transfusion circulatory over-load, and in 20% of the cases a single RBC unit was sufficientto provoke overload (17). In patients susceptible to overload,RBC transfusions should be administered slowly, and in themost concentrated form possible.

Rapid infusion of refrigerated, cold blood through a cen-tral venous line may cause hypothermia in some recipients,with the attendant risk of cardiac arrhythmia. Hypothermia,as well as potassium or citrate toxicity, is a concern primarilyin the setting of massive transfusion. When a patient receivesthe equivalent of more than one blood volume (�10 RBCunits) over 24hr, metabolic changes that occur in RBCsduring storage become a consideration. The infused RBCsupernatant is hyperkalemic and acidotic, and the infusionof potassium and acid may result in hyperkalemia or acidosis.Also, the infused citrate anticoagulant may complex theionized calcium in the recipient’s circulation, resulting in

Table 1 Established Noninfectious Complications of Transfusion

Immunologic Nonimmunologic

1. Reactions to WBCs 1. Physical complicationsFNHTRs Volume overloadTRALI HypothermiaGVHD

2. Chemical complications2. Allergic and anaphylactic reactions Iron overload (transfusion

Minor allergic reactionsAnaphylactoid reactionsAnaphylactic reactions

siderosis)Potassium toxicityCitrate toxicityDepletion of 2,3-DPG

3. Hemolytic transfusion reactionsAcute (AHTR)Delayed (DHTR)

4. PTP


hypocalcemia. Such electrolyte abnormalities are generallyself-correcting, however, and they are rarely a concern inpatients receiving transfusion therapy for chronic anemia.

Transfusion-induced iron overload is an often fatal con-sequence of chronic transfusion for refractory anemia. Chil-dren with thalassemia major are the single largest groupaffected, but a substantial number of children with congenitalanemias and adults with intensely-treated refractory anemiaare likewise at risk. Every mL of RBCs deposits 1.08mg ofiron in tissues, as RBCs age and die. Iron deposition beginsto affect endocrine, hepatic, and cardiac function when thetotal body burden has risen to more than 20 g (i.e., the equiva-lent of about 100 RBC units). Lethal cardiac complicationsoccur at about 60 g, or approximately 300 RBC units. Ironchelation therapy should be considered for all patients likelyto require intensive, chronic RBC transfusion support.

Immune Complications of ABT

Mild urticarial reactions to plasma proteins complicate 1–3%of plasma infusions, while anaphylactic shock occurs onceper 20,000–47,000units of transfused blood components (18).Anaphylactic shock as a result of ABT is sometimes associatedwith preformed IgG antibodies directed to IgA immunoglobu-lin (19). However, of 359 sera referred to the American RedCross National Reference Laboratory from patients withsuspected anaphylactic reactions, only 65 (18%) demonstratedanti-IgA antibody (20). Thus, the culprit plasma protein(s) towhich an individual with a history of anaphylactic reaction(s)has been sensitized usually remains unknown. Allergic trans-fusion reactions can be prevented by pretreatment of thepatient with an antihistamine before the transfusion. Patientswith a history of anaphylactic or severe anaphylactoid reac-tions should receive washed cellular blood components in thefuture, because such components are devoid of plasma pro-teins and cannot provoke anaphylactic reactions to plasmaproteins.

Febrile, nonhemolytic transfusion reactions (FNHTRs)may be attributable to preformed anti-WBC antibodies in

404 Vamvakas

the circulation of recipients and=or cytokines released fromWBCs into the supernatant plasma of platelet units storedat room temperature. Such reactions follow about 1% ofRBC transfusions and are much more frequent after platelettransfusions. The FNHTRs are prevented by prestorageWBC reduction that removes > 99.9% of donor WBCs, thusalso preventing release of any cytokines from WBCs duringstorage. Therefore, the introduction of universal, prestorageWBC reduction has greatly reduced the risk of FNHTRs inmany Western European countries and in Canada. In theUnited States, where universal WBC reduction has not beenimplemented, patients who have experienced one or twoFNHTRs receive WBC-reduced RBCs and platelets inthe future.

In one study, 1 in 5000 plasma-containing transfusionswas associated with transfusion-related acute lung injury(TRALI) (21). TRALI was formerly known as noncardiogenicpulmonary edema. TRALI mimics adult respiratory distresssyndrome (ARDS) but has a better prognosis, with a mortalityof only 5–10%. It is believed that TRALI is caused by donoranti-WBC antibodies that attack the recipient’s WBCs in themicrocirculation of the lungs. Thus, donors who have beenimplicated in TRALI are deferred from donating blood. In somecases, however, lipid neutrophil-priming agents and cytokinesthat accumulate in stored blood components have been impli-cated in these reactions, and it is possible that TRALI mayresult from multiple mechanisms.

Transfusion of viable lymphocytes to immunocompro-mised hosts, or transfusion of viable lymphocytes from first-degree relatives to immunocompetent hosts, can result incell-mediated graft-versus-host disease (GVHD). Because ofeither recipient immunosuppression or human leukocyteantigen (HLA) similarity between donor and recipient, trans-fused donor lymphocytes are not eliminated; instead, they arestimulated to proliferate, and they mount an immune attackagainst the recipient. Transfusion-associated GVHD is pre-vented by gamma-irradiating cellular blood components. Thisintervention removes the ability of donor lymphocytes toproliferate.


Patients who have been pregnant or transfused in thepast may have become sensitized to RBC antigens. When atransfusion recipient has a circulating antibody to an RBCantigen and is transfused with a unit of blood that carriesthat antigen, an acute hemolytic transfusion reaction (AHTR)follows within 24hr of the transfusion. Transfused donorRBCs are destroyed rapidly in the circulation of the recipient,producing fever, profound hypotension, renal shutdown, and=or disseminated intravascular coagulation.

Pretransfusion testing of the recipient and crossmatch-ing of the recipient’s serum with the donor’s RBCs ensure that(1) patients receive only ABO-compatible RBC components,and (2) patients with preformed anti-RBC antibodies do notreceive a unit that carries the corresponding antigen(s). How-ever, AHTRs still occur, with a frequency of up to 1 AHTR per20,000 transfused RBC units (22), because of clerical error,that is, because of administration of an RBC unit to a patientother than the intended recipient. Most AHTRs are due toABO incompatibility, and they can be associated with a mor-tality of up to 40%. Thus, an expert panel estimated that 1 in100,000 transfused patients may die of an AHTR (9).

A delayed hemolytic transfusion reaction (DHTR),occurring 3–21 days after RBC transfusion, is more frequentthan an AHTR. In this case, the recipient has a circulatingantibody to an RBC antigen that is not detected at the timeof pretransfusion testing because it has an extremely lowtiter. If the patient receives an RBC unit that carries the cor-responding antigen, an anamnestic immune response follows,producing a rapid rise in antibody titer. In approximately aweek, there is sufficient circulating antibody to destroy thetransfused donor RBCs, but, because hemolysis occurs slowly,hemolysis is often asymptomatic or only minimally sympto-matic. Thus, a DHTR often goes unrecognized, and itsincidence may be as high as 1 DHTR per 5,405 transfusedRBC units (23).

Post-transfusion purpura (PTP) is characterized by thedevelopment of dramatic, sudden, and self-limiting thrombo-cytopenia 5–10 days after a blood transfusion in a patientwith a history of sensitization to a platelet-specific antigen.

406 Vamvakas

Coincident with the thrombocytopenia is the development of apotent platelet-specific antibody in the patient’s serum,usually anti-PlA1. The transfusion that triggers the reactioncan consist of any blood component containing platelet anti-gens, including RBCs, plasma, and=or platelets. However,the mechanism of destruction of the patient’s own PlA1-nega-tive platelets, following the appearance of anti-PlA1 alloanti-body in the circulation, is unclear. Approximately 300 casesof PTP have been reported in the literature (24).

INFECTIOUS RISKS OF ABT

Transfusion-transmitted agents fall into one of three cate-gories: (1) agents that produce disease in transfusion recipi-ents and for which laboratory screening of all prospectiveblood donors is performed; (2) agents that produce diseasein transfusion recipients, but for which laboratory screeningof donors is not performed; and (3) agents that do not producedisease in transfusion recipients and for which laboratoryscreening of donors is not performed (Table 2). There are alsoagents, such as the agent of classic Creutzfeldt–Jakob disease(CJD), which represent a theoretical risk to transfusion reci-pients. These agents do produce disease, but it is unknownwhether they can be transmitted by transfusion. Precaution-ary policies have been implemented in many countries to pre-vent the theoretical risk of transmission of classic CJD bytransfusion. More recently, the same precautionary principlewas applied to the theoretical risk of transmission of theagent of severe acute respiratory syndrome (SARS) by trans-fusion.

Human immunodeficiency viruses 1 and 2 (HIV-1=2),hepatitis C virus (HCV), hepatitis B virus (HBV), and humanT-cell lymphotropic viruses I and II (HTLV-I=II) are the tradi-tional transfusion-transmitted viruses that produce diseasein recipients after many years. Transmission of Treponemapallidum by transfusion is theoretically possible, but exceed-ingly rare, because the phase of spirochetemia is brief, andthe spirochetes survive only for a few days during RBC


storage at 4�C. The last case of transfusion-transmitted syphi-lis in the United States occurred in 1965. Transmission ofWest Nile Virus (WNV) by transfusion had not been recog-nized previously, but, in the summer and fall of 2002, WNVemerged as the most common transfusion-transmitted virusin North America (25). Over the course of the mosquito-borneepidemic of WNV between June and October 2002 in theUnited States as a whole, the risk of transmission of WNVby transfusion was estimated to be 1 viremic donation per300,000 donations. However, at the peak of the epidemic from

Table 2 Transfusion-Transmitted Infectious Agentsa

1. Agents that produce disease and for which laboratory screeningof donors is performed:HIV-1=2HCVHBVHTLV-I=IIT. pallidumWNVb

2. Agents that produce disease but for which laboratory screeningof donors is not performed:HAVHHV-8Parvovirus B-19Trypanosoma cruziBabesia microtiPlasmodium sp.c

CMVBacteriad

3. Agents that do not produce disease and for which laboratoryscreening of donors is not performed:HGVTT virusSEN-V virus

aThe table does not list all transfusion-transmitted infectious agents.bScreening is performed only in North America.cScreening of prospective donors who report travel to malaria-endemic countries isperformed in some Western European countries.dScreening is performed in some Western European countries. Screening of plateletcomponents collected by apheresis started in North America in March, 2004.

408 Vamvakas

mid to late August, and in areas such as Detroit, Michigan orCleveland, Ohio, the risk of transmission was greater than1 viremic donation per 1000 donations. Thus, both theUnited States and Canada implemented screening of allprospective blood donors for WNV as of July 1, 2003.

Donors are not screened for the hepatitis A virus (HAV),human herpesvirus 8 (HHV-8), parvovirus B-19, or the agentsof Chagas’ disease, babesiosis, and malaria, because theseagents are so rare in North American and European donorsthat the risk of transmission by transfusion is remote. Therisk of transfusion transmission of CMV has also becomeremote in several Western European countries and inCanada, following the implementation of universal WBCreduction. The CMV resides in WBCs and is removed when>99.9% of WBCs are removed by prestorage filtration ofcellular blood components. In addition to universal WBCreduction, some countries also use CMV-seronegative donorswhen they collect products destined for transfusion to ‘‘high-risk’’ patients. The latter include fetuses, neonates, andimmunosuppressed subjects who may develop full-blownCMV disease (as opposed to asymptomatic CMV infection) fol-lowing exposure to CMV. Bacteria are presently the mostcommon infectious agents transmitted by transfusion, andthey are thus discussed in some detail.

There are also transfusion-transmitted viruses that wereinitially hypothesized to cause hepatitis, but now appearnot to cause disease in transfusion recipients. They includehepatitis G virus (HGV), TT virus, and SEN-V virus.

Traditional Transfusion-Transmitted Viruses

Although each unit of blood is screened for HIV-1=2, HCV,HBV, and HTLV-I=II, transmission of these agents by trans-fusion can still occur, because of window-period infections.The ‘‘window period’’ starts when a level of viremia sufficientto transmit infection is established in the circulation, andends when viral markers appear in peripheral blood thatare detected by tests used to screen the blood supply fortransfusion-transmitted viruses. Window-period infectionsare by far the most common cause of transmission by


transfusion. Other (rare) causes include: (1) a chronic carrierstate that is not detected by the screening assays used onblood donors, (2) viral variants that are undetectable by cur-rent assays, and (3) laboratory error. The latter possibilityis essentially eliminated when two (rather than one) screen-ing assays are performed for the same agent. The risk oftransfusion transmission because of window-period infectionscan be estimated by multiplying the incidence of new infec-tions in donors by the window period of each virus. Thismethod for estimating risk is known as the incidence=window-period model (26).

In the case of HIV, each unit of blood is screened by twoindependent laboratory tests: an assay detecting antibody toHIV-1=2 and an assay detecting circulating HIV-1 RNA.When these tests are used together, the window period ofHIV is only 11–13 days. The incidence of new HIV infectionsin Canadian blood donors is 0.53 infections per 100,000 person-years. (That is, on average, 1 donor per 200,000 donorscontracts HIV infection in Canada each year.) Therefore,the risk of HIV transmission by transfusion—that is derivedby multiplying these two figures—is only 1 viremic donationper 3,900,000–4,700,000 donations (27). Because of a higherincidence of new HIV infections in blood donors, a risk thatis approximately four times higher than the Canadian riskhas been estimated for the United States: 1 viremic donationper 900,000–1,200,000 donations (28).

In the case of HCV, each unit of blood is also screened bytwo independent laboratory tests: an assay detecting antibodyto HCV and an assay detecting circulating HCV RNA. Theresidual window period for HCV is 8–12 days, and the riskof transmission is estimated to be 1 viremic donation per2,600,000–3,100,000 donations in Canada, and 1 viremicdonation per 900,000–1,200,000 donations in the UnitedStates (27,28).

In the case of HBV, each unit of blood is currentlyscreened for the presence of hepatitis B surface antigen(HbsAg) in Canada and the United Kingdom; and for the pre-sence of both HbsAg and antibody to the HBV core antigen(anti-HBc) in the United States and many Western European

410 Vamvakas

countries. One HbsAg-negative donation per 50,000 may beviremic and testing anti-HBc-positive, because there are someHBV chronic carrier states that are not detected by the assayfor HbsAg (29). The risk of HBV transmission in the UnitesStates (secondary to window-period infections) is estimatedto be 1 viremic donation per 62,000 donations (30). The riskof HBV transmission in Canada (secondary to both window-period infections and undetectable chronic carrier states) ispresently estimated to be 1 viremic donation per 31,000 dona-tions (27). When Canada introduces anti-HBc testing of blooddonors in 2005, the Canadian risk of HBV transmission (sec-ondary to window-period infections alone) will be reduced to 1viremic donation per 82,000 donations (27). Because the win-dow period of HBV prior to the appearance of HbsAg in thecirculation is long (59 days), HBV is presently the most fre-quent transfusion-transmitted virus.

Table 3 contrasts the risk of the four classic transfusion-transmitted viruses between the early and late 1990s, as wellas between the United States and Canada. Schreiber et al. (31)used U.S. data from 1991 to 1993 to estimate the risk of trans-mission before screening of blood donors for circulating HCVand HIV-1 RNA was added in the late 1990s. Screening forviral RNA reduced the window period of HCV and HIV-1,but the window period (and thus the transmission risk) ofHBV and HTLV has remained the same, since no new testshave been introduced for these viruses. However, the intro-duction of universal WBC reduction in Canada rendered therisk of transfusion transmission of HTLV remote, becauseHTLV resides in WBCs and is removed when WBCs areremoved by prestorage filtration.

Bacterial Contamination of Blood Components

Bacteria are presently the most common infectious agents con-tracted by transfusion. The BaCon Study (32) estimated the riskof transfusion-associated bacteremia as 1 case of bacteremia per100,000units of transfused platelets or 500,000units oftransfused RBCs. Platelets are stored at room temperature,permitting any bacterial contaminants to propagate during


Table

3Estim

ation

oftheRiskof

Transm

ission

ofViralIn

fectionsThroughTransfusion

Usingthe

Inciden

ce=Window

-PeriodMod

el:Early1990svs.

Late

1990sandUnited

Statesvs.

Canada

Transfusion

-transm

itted

virus

U.S.estimate

oftherisk

oftransm

ission

reported

intheea

rly1990s(31)a

Updatedpointestimate

ofrisk

a

Point

estimate

95%

CI

United

States

(28)

Canada

(27)

HIV

1=493,000

1=202,000to

1=2,778,000

1=900,000to

1=1,200,000

1=3,900,000to

1=4,700,000

HCV

1=103,000

1=28,000to

1=288,000

1=900,000to

1=1,200,000

1=2,600,000to

1=3,100,000

HBV

1=63,000

1=31,000to

1=147,000

1=62,000(30)

1=31,000

HTLV

1=641,000

1=256,000to

1=2,000,000

1=641,000

Rem

ote

aExpressed

per

don

atedunit

ofblood

.

412 Vamvakas

storage; thus, platelets are contaminated with bacteria muchmore frequently than RBCs. However, platelets are usually con-taminated with gram-positive skin flora that enter the unit dur-ing venipuncture, in the process of whole blood collection. Mostpatients who receive a platelet unit contaminated with Staphy-lococcus epidermidis or other gram-positive skin flora eitherremain asymptomatic or develop transient fever that is notlinked to the transfusion.

In contrast, RBCs are usually contaminated by cryophilicorganisms that can proliferate in the cold, as blood is stored at4�C over 3–4 weeks. Such cryophilic organisms (e.g., Pseudo-monas sp., Serratia sp., Yersinia enterocolitica, othergram-negative enteric bacilli) originate in donors who haveasymptomatic bacteremia. During storage, the gram-negativebacilli generate endotoxin, and—when a unit containingendotoxin is transfused—septic shock follows, during orshortly after the transfusion. Such reactions are associatedwith a 25% mortality rate. Thus, in the BaCon Study (32),fatality rates were estimated as 1 death per 500,000units oftransfused platelets or 8,000,000units of transfused RBCs.

The results of the BaCon Study (32) are, most likely,gross underestimates of the problem. The study was designedto maximize positive predictive value. Possible cases of bac-teremia in patients receiving antibiotics were not counted ifthe recipients’ blood cultures remained negative. The authorscautioned that previous studies, designed to maximize sensi-tivity, had reported estimates of transfusion-associatedbacteremia that were up to 10-fold higher than the figuresgenerated by the BaCon Study. In fact, some data have sug-gested that 1 platelet unit in 2000 may be contaminated withbacteria (33). Thus, there is a general agreement that thetransfusion-associated bacteremia is an under recognized,under reported, and serious patient safety issue, and thatno good estimate of the frequency of transfusion-associatedbacteremia exists.

Several European countries have implemented screeningof donated units of platelets (and=or whole blood) for bacteria.A new standard of the American Association of Blood Banks(AABB), intended to prevent bacterial contamination of


platelets, with became effective in March 2004. It states that‘‘The blood bank or transfusion service shall have methods tolimit and detect bacterial contamination in all platelet compo-nents.’’ Thus, it is expected that, in a few years, blood dona-tions in North America and Europe will be screened for thepresence of both bacteria and viruses.

Theoretical Infectious Risks of ABT

Classic Creutzfeldt–Jakob Disease

Prions are proteinaceous infectious particles, devoid of nucleicacid and composed exclusively of conformationally changed,host-encoded prion protein. Ninety percent of cases of classicCJD are of unknown etiology and are referred to as sporadicCJD. Sporadic CJD has been occurring worldwide, with afrequency of 1 case of disease per 1,000,000 members of thepopulation per year. Less than 1% of cases of classic CJD istransmitted by contaminated neurosurgical instruments,human growth hormone, or corneal or dura mater transplantgrafts. This small number of acquired cases has raised thepossibility that classic CJD might also be transmitted bytransfusion, and individuals with risk factors for classicCJD are permanently deferred from donating blood.

Four laboratories have transmitted classic CJD tolaboratory rodents after intracerebral inoculation of humanblood from patients with classic CJD. The findings of theseexperiments have been controversial, and each success hasbeen questioned on technical grounds or has not been repro-ducible. Moreover, based on all the available data, blood fromonly 4 of 37 patients with classic CJD has been reportedto transmit disease after intracerebral inoculation intorodents (34).

Classic CJD has developed in two patients who receivedalbumin from donors who later died of unconfirmed orconfirmed CJD and in four patients who received blood trans-fusions from undocumented sources. However, without a toolto distinguish between sporadic and acquired CJD, or link theagent from a blood donor to a blood recipient, it is impossible

414 Vamvakas

to make a judgment from these case-reports about a causalrelationship between the receipt of a transfusion and thedevelopment of classic CJD (35).

The U.S. National Blood Data Resource Center is identi-fying all recipients who received blood from donors who laterdeveloped classic CJD. More than half of approximately 250investigated recipients are dead, and no death has been CJD-related (36). Also, neuropathologic examinations of brains ofdeceased, multiply transfused patients such as thalassemicsor hemophiliacs have not found a single case of classic CJD(37). Based on such data, a consensus is emerging that theclassic CJD is not transmitted by transfusion.

Variant Creutzfeldt–Jakob Disease

Variant CJD is a new disease, epidemiologically and neuro-pathologically linked to bovine spongiform encephalopathy(BSE). Worldwide as of May 2003, there had been 143 cases,almost all of them in the United Kingdom or France. BecausevCJD was a new disease, the evidence about lack of transmis-sion of classic CJD by transfusion could not be extrapolated tovCJD. In fact, three features of vCJD left open the possibilitythat this new disease might prove to be transmissible bytransfusion. These were (1) pathologic prion protein had beendemonstrated in the tonsils, spleen, lymph nodes, and appen-dix of patients with vCJD, but not classic CJD; (2) unlikeclassic CJD (which requires intracerebral inoculation fortransmission), vCJD was acquired peripherally, through theoral route; and (3) results of an experimental study hadindicated that it was possible to transmit BSE to sheep bytransfusion of whole blood collected from another sheep dur-ing the asymptomatic phase of experimental BSE infection(38,39).

The first case of transmission of vCJD by transfusion wasreported from the United Kingdom in December, 2003. A sec-ond case of transmission was reported, also from the UnitedKingdom, in August, 2004 (40). Thus, the risk of transmissionof vCJD by transfusion ceased to be theoretical: patients who


had received blood from donors who later developed vCJD hadgone on to develop vCJD themselves.

Before these cases of transmission of vCJD by transfu-sion had been reported, in the face of uncertainty about thepossibility of transmission of vCJD by transfusion, it had beendeemed appropriate to implement precautionary policies toprotect the safety of the blood supply, without any guaranteethat these measures were either necessary or sufficient.Scientifically sound measures to prevent transmission ofvCJD by transfusion await development of a screening testthat can distinguish between normally and abnormally con-formed prion proteins found in the circulation of donors. Inthe interim, because of concern that those exposed to infec-tious British beef in the 1980s and early 1990s might carrythe agent of vCJD, both the United States and Canada indefi-nitely defer all prospective donors who have, since 1980, spentmore than 3 months in the United Kingdom, more than 6months in France, or more than 5 years in other countriesof continental Europe. These deferral criteria primarilyreflect concern about exposure to (imported) British beef,and—only to a minor extent—concern about exposure toinfectious local beef, as BSE is reported in other Europeancountries.

Severe Acute Respiratory Syndrome

Suspected SARS is defined as respiratory illness of unknownetiology, occurring since February 1, 2003, and combiningfever (>38 �C), one or more respiratory symptoms, and possi-ble exposure to the SARS agent either through travel to anarea with documented community transmission of SARS orthrough close contact with a person who traveled to anaffected area or had suspected SARS. Probable SARS is sus-pected SARS that also demonstrates radiologic evidence ofpneumonia or respiratory distress syndrome. SARS is attrib-uted to a novel coronavirus that has been isolated frominfected kidney, but has not yet been isolated from theblood or serum of an infected individual. However, it has been

416 Vamvakas

possible to amplify viral RNA, using polymerase chain reac-tion technology, from the blood of a single patient. Thus, thepotential for transmission of SARS through transfusionremains unknown; it is possible that SARS may have a vire-mic phase before (and=or after) the appearance of symptoms.Based on the precautionary principle, prospective donors whoreport a history of SARS are deferred from donating bloodindefinitely in the United Kingdom or for 6 months inCanada. Moreover, donors who report travel to areas withdocumented community transmission of SARS or contact withpersons with SARS or suspected SARS are deferred for 21days (41).

PURPORTED NONINFECTIOUS RISKS OF ABT

The clinical relevance of ABT-related immunomodulation(TRIM) continues to be debated (5). The purported clinicalmanifestations of an adverse TRIM effect of ABT includeincreased recurrence of resected malignancies, increased inci-dence of postoperative bacterial infections, increased mortal-ity, and activation of CMV and HIV infections in transfused(compared with untransfused) patients. These adverse effectsof ABT may be mediated by: (1) immunologically active allo-geneic WBCs (42) present in RBC components stored for< 2 weeks, (2) proinflammatory soluble mediators releasedfrom WBC granules (43) or membranes (44) into the superna-tant fluid of RBCs and accumulating in a time-dependentmanner during storage, and=or (3) soluble HLA class I mole-cules circulating in allogeneic plasma (45). Both prestorageand poststorage WBC reduction of allogeneic RBCs canprevent a TRIM effect mediated by immunologically activeallogeneic WBCs (46). Neither prestorage nor poststorageWBC reduction of allogeneic RBCs can prevent a TRIM effectmediated by soluble HLA molecules circulating in allogeneicplasma. Only prestorage (as opposed to poststorage) WBCreduction can prevent a TRIM effect mediated by WBC-derived, soluble mediators that accumulate in the superna-tant fluid of RBCs during storage.


Immunologically Active Allogeneic WBCs asMediators of TRIM

Animal data (47,48) suggest that the TRIM effect is mediated,primarily, by transfused allogeneic WBCs. The VATS wasdesigned to test the specific hypothesis that relatively freshRBCs, presumed to contain immunologically active WBCs,activate endogenous CMV and=or HIV infection, and thuspossibly also mediate other TRIM effects that were not stu-died by the VATS investigators (4). In this study, transfusedallogeneic RBCs were stored for <2 weeks before transfusionand the two study arms were controlled for a comparableduration of storage. Enrolled HIV-seropositive patients wererandomized to receive buffy-coat-rich (standard) RBCs orRBCs WBC-reduced by prestorage filtration and then thepatients were followed with respect to viral load and survival.

There was no difference between the two study arms inthe HIV or CMV viral load after transfusion, indicating thatimmunologically active allogeneic WBCs in the transfusedRBC components had not provoked any virus activation.Furthermore, there was no difference between the two studyarms in the length of patient survival or in any of the second-ary end-points of the study. Median survival was 13.0 monthsin recipients of WBC-reduced RBCs compared to 20.5 monthsin recipients of non-WBC-reduced RBCs (p¼ 0.12). However,when adjustment was made for the effects of baseline prog-nostic factors for survival, a statistically significant (p<0.05) survival detriment was detected in association withthe use of WBC-reduced RBCs filtered before storage. Thus,the VATS results impugned the theory attributing the TRIMeffect to immunologically active allogeneic WBCs.

Eight RCTs have compared the risk of postoperativeinfection between recipients of WBC-reduced and non-WBC-reduced RBCs (49–55) or whole blood (56), generating highlycontradictory findings (Fig. 1). In these studies (49–56), theTRIM effect has varied from a 7.3-fold increase in the riskof postoperative infection (56) to no transfusion effect (49).The extreme variation in the findings of the studies precludesany attempt to integrate their results by the techniques of

418 Vamvakas

meta-analysis (5). Moreover, these studies were not designedto test the specific hypothesis that relatively fresh RBCs (con-taining immunologically active WBCs) mediate adverse TRIMeffects. Thus, these studies did not exclusively transfuse freshRBCs (stored for <2 weeks) to the treatment arm. Also, thelength of storage of the transfused RBCs was not stated inmost reports, and study arms were not controlled for compar-able duration of storage. Therefore, even if the results of theeight studies were to support the existence of an adverseTRIM effect of ABT, such an effect could not be specificallyascribed to immunologically active allogeneic WBCs.

Figure 1 RCTs comparing the risk of postoperative infectionbetween recipients of non-WBC-reduced and WBC-reduced RBCs.For each RCT, the figure shows the OR of postoperative infection inthe treatment vs. the control arm calculated based on an intention-to-treat analysis. Each OR is surrounded by its 95% CI. When the95% CI include the null value of 1, the corresponding TRIM effect isnot statistically significant (i.e., p>0.05). (From Ref. 59.)


Soluble HLA Molecules Circulating in AllogeneicPlasma as Mediators of TRIM

This author knows of only 1 RCT (53) whose design permittedinvestigators to examine the hypothesis that soluble HLAmolecules circulating in allogeneic plasma may mediate theTRIM effect of ABT. Wallis et al. (53) randomized 595 patientsundergoing open-heart surgery to receive plasma-reduced,buffy-coat-reduced, or WBC-reduced allogeneic RBCs. Thehighest risk of infection was observed in the (WBC-containing)plasma-reduced arm, where the incidence of postoperativeinfection was 17.1%, compared with 10.8% in the buffy-coat-reduced arm, and 11.3% in the WBC-reduced arm (p¼ 0.20).When the analysis was restricted to the transfused patients,there was a significant (p< 0.05) difference in the risk of infec-tion between the plasma-reduced arm and the two other arms,indicating that plasma removal did not confer any beneficialeffect with regard to the prevention of a TRIM effect of ABT.

WBC-Derived Soluble Mediators as Mediatorsof TRIM

Histamine, eosinophil cationic protein, eosinophil protein X,myeloperoxidase, and plasminogen activator inhibitor 1increase by 3–25-fold in the supernatant fluid of RBCs duringstorage (43). These substances are normally contained in intra-cellular WBC granules, and are released in a time-dependentmanner as the WBCs deteriorate. Fas-ligand is similarlyreleased from WBC membranes (44). The infusion of solubleFas-ligand may impair the function of natural killer and cyto-toxic T cells of the recipient because the infused Fas-ligandbinds the Fas molecule on immune cells, thereby preventingthe binding of the Fas molecule on immune cells to theFas-ligand on virus-infected cells. Finally, apoptotic WBCsthat accumulate during storage may also have TRIM effects(57,58). TheWBC reduction filters do not retain soluble media-tors and they do not remove WBC fragments. Thus, if solublemediators and apoptotic WBCs were responsible for someTRIM effects, WBC reduction procedures should be performedprior to storage to prevent such TRIM effects.

420 Vamvakas

RCTs Comparing Recipients of Non-WBC-Reducedvs. WBC-Reduced Allogeneic RBCs

The RCTs shown in Fig. 1 (49–56) have not demonstrated abenefit from prestorage (compared with poststorage) WBCreduction (59). Eight RCTs have compared the risk of post-operative infection between recipients of buffy-coat-rich orbuffy-coat-reduced allogeneic RBCs or whole blood vs. recipi-ents of allogeneic RBCs or whole blood WBC-reduced before orafter storage. Three studies administered buffy-coat-richRBCs to the treatment arm, including whole blood (56), RBCs(52), or plasma-reduced RBCs from which the buffy coat hadnot been removed (53). Six studies administered buffy-coat-reduced RBCs to the treatment arm (49–51,53–58). TheRCT of Wallis et al. (53) enrolled two treatment arms, receiv-ing plasma-reduced or buffy-coat-reduced allogeneic RBCs.Four studies administered allogeneic RBCs (50–52) or wholeblood (56) WBC-reduced after storage to the control arm andfive studies (49,51,53–55) administered buffy-coat-reducedallogeneic RBCs WBC-reduced before storage. The RCT ofvan deWatering et al. (51) enrolled two control arms, receivingbuffy-coat-reduced allogeneic RBCs WBC-reduced before orafter storage.

Across five RCTs (49,51,53–55) comparing the risk of post-operative infection between recipients of buffy-coat-reducedallogeneic RBCs WBC-reduced prior to storage (control arm)and recipients of buffy-coat-reduced allogeneic RBCs (treat-ment arm), the summary odds ratio (OR) of postoperativeinfection in the treatment vs. the control arm was not statisti-cally significant (summary OR¼ 1.19; 95% confidence interval[CI], 0.87–1.63; p>0.25) (Fig. 2). The studies shown in Fig. 2are medically homogeneous, because they administered thesame RBC product to both the treatment arm and the controlarm. These studies are also statistically homogeneous, becausethe noted variation in results (Fig. 2) is sufficiently modest tobe attributed to chance. Thus, it is legitimate to combine theresults of the studies by the techniques of meta-analysis,and the synthesis of these studies (59) indicates that—basedon the cumulative experience from 2,455 randomized


subjects—a 20% or greater increase in the risk of postoperativeinfection in association with the receipt of buffy-coat-reducedallogeneic RBCs can be ruled out with greater than 75% confi-dence. Thus, further comparisons of patients receivingbuffy-coat-reduced vs. prestorage-filtered allogeneic RBCs areunlikely to demonstrate an adverse TRIM effect, and such ran-domized comparisons should not be undertaken in the future.

Figure 2 Summary OR of postoperative infection derived fromRCTs administering buffy-coat-reduced allogeneic RBCs vs. buffy-coat-reduced allogeneic RBCs WBC-reduced prior to storage. Theincluded RCTs are homogeneous both medically and statistically(59). For each RCT, the figure shows the OR of postoperative infec-tion in the treatment vs. the control arm calculated based on anintention-to-treat analysis. Each OR is surrounded by its 95% CI.The 95% CI of the summary OR includes the null value of 1, andthus the summary TRIM effect is not statistically significant. (FromRef. 59.)

422 Vamvakas

Among the available RCTs that used postoperative infec-tion as the outcome variable (49–56), no other group of studieswas homogeneous both statistically and medically (59). FourRCTs administered allogeneic RBCs or whole blood WBC-reduced after storage to the control arm (50–52,56). Acrossthese four RCTs, the summary TRIM effect was statisticallysignificant (p< 0.05), and it consisted of a two-fold increasein the risk of infection in the treatment vs. the control arm(summary OR¼ 2.25; 95% CI, 1.20–4.25) (Fig. 3). Two of thesefour RCTs administered buffy-coat-rich allogeneic RBCs (52)or whole blood (56) to the treatment arm.

Across three RCTs administering buffy-coat-rich allo-geneic RBCs (52,53) or whole blood (56) to the treatmentarm, and allogeneic RBCs or whole blood WBC-reduced before

Figure 3 Summary OR of postoperative infection derived fromRCTs administering allogeneic RBCs or whole blood WBC-reducedafter storage to the control arm. Patients in the treatment armreceived buffy-coat-reduced allogeneic RBCs (50,51), allogeneicRBCs (52), or allogeneic whole blood (56). Thus, these studies aremedically heterogeneous, but they are statistically homogeneous(59). For each RCT, the figure shows the OR of postoperative infec-tion in the treatment vs. the control arm calculated based on anintention-to-treat analysis. Each OR is surrounded by its 95% CI.


(53) or after (52,56) storage to the control arm (Fig. 4), there wasalso a greater risk of postoperative infection in the treatment(compared with the control) arm (summary OR¼ 1.77; 95%CI, 1.02–3.09; p< 0.05). Thus, the meta-analysis (59) pointedto a TRIM effect mediated by buffy-coat-rich allogeneic RBCsor whole blood, but also prevented by allogeneic RBCs or wholeblood that had been WBC-reducedafter storage. The latter find-ing contradicts the hypothesis attributing the TRIM effect toWBC-derived soluble mediators that accumulate during sto-rage, because such mediators would not have been removed—by means of poststorage filtration—from the RBC componentsadministered to the control arm of these studies.

Figure 4 Summary OR of postoperative infection derived fromRCTs administering buffy-coat-rich allogeneic RBCs (52), buffy-coat-rich but plasma-reduced allogeneic RBCs (53), or buffy-coat-rich allogeneic whole blood (56) to the treatment arm. Patients inthe control arm received allogeneic RBCs or whole blood WBC-reduced before or after storage. Thus, these studies are medicallyheterogeneous, but they are statistically homogeneous (59). For eachRCT, the figure shows the OR of postoperative infection in the treat-ment vs. the control arm calculated based on an intention-to-treatanalysis. Each OR is surrounded by its 95% CI. (From Ref. 59.)

424 Vamvakas

TRIM and Increased Mortality

An association between WBC-containing ABT and increasedpostoperative mortality in open-heart surgery was reportedfrom the RCTs of van de Watering et al. (51) and Bilginet al. (55), generating a hypothesis that ABT may predisposeto multiple-organ failure that may in turn predispose toincreased mortality (46). Recently, the introduction of univer-sal, prestorageWBC reduction in Canada allowedHebert et al.(60) and Fergusson et al. (61) to undertake large-scale obser-vational studies comparing the frequency of various adverseoutcomes between adult (60) or neonatal (61) patients trans-fused before or after the introduction of universal WBC reduc-tion. Among adults, and compared with the period before theintroduction of universal WBC reduction, the adjusted OR ofdeath was apparently reduced thanks to the implementationof this intervention (OR¼ 0.87; 95%CI, 0.75–0.99; p¼ 0.04);the adjusted OR of serious nosocomial infection did notdecrease, however (OR¼ 0.97; 95% CI, 0.87–1.09) (60). Amongneonates, no reduction in the risk of death or bacteremia coin-cided with the implementation of universal WBC reduction.The adjusted ORs of bronchopulmonary dysplasia and retino-pathy of prematurity were apparently reduced thanks to theimplementation of this intervention (61).

Because the two Canadian studies (60,61) were observa-tional and retrospective, they could not attribute anyimproved outcome(s) observed postuniversal WBC reductionto the introduction of this intervention per se (62). Changesin patient case-mix or other medical treatments coincidingwith the introduction of universal WBC reduction might havecaused the differences recorded by Hebert et al. (60) andFergusson et al. (61). Moreover, other observational, before-and-after studies that made similar comparisons betweenpatients transfused before or after the introduction of univer-sal WBC reduction in France (63) or England (64) reached dif-ferent conclusions, and the findings of Hebert et al. (60) andFergusson et al. (61) contrast sharply with the results of theRCT of Dzik et al. (65) that assessed the benefits of universalWBC reduction in a prospective and experimental manner.


More recently, the RCT of Bracey et al. (66) also found no differ-ence in the risk of postoperative mortality or infection betweenopen-heart surgery patients transfused with non-WBC-reducedRBCs or WBC-reduced RBCs filtered before storage.

TRIM and Cancer Recurrence

An association between ABT and cancer recurrence wasreported from a number of observational studies (67). However,the available RCTs (49,68,69) that compared the risk of cancerrecurrence between recipients of buffy-coat-reduced allogeneicRBCs or whole blood and recipients of autologous or WBC-reduced allogeneic RBCs or whole blood showed no reductionin the risk of cancer recurrence in association with the receiptof autologous or WBC-reduced allogeneic blood (5,70).

SUMMARY

Although some patientswith chronic renal failure, disseminatedmalignancy, or HIV-related anemia may become transfusion-dependent, ABT is indicated rarely in anemia secondary tochronic systemic disease—mostly when an aggravating factorresults in increasing anemia. In such situations, ABT shouldbe administered only after careful consideration of the clinicalbenefit expected from the transfusion and its risks. The risksof ABT should be neither overstated nor understated, and aneffort should be made to ascertain the objective benefit that apatient derives from RBC transfusion. Because most patientswith anemia of chronic disease are significantly limited by theirprimary disease, transfusion for improving the patient’s func-tional status may offer little benefit, rendering a considerationof the risks of ABT most important.

Allogeneic blood transfusion carries various proven,purported, or theoretical risks that can be infectious or nonin-fectious. The most likely complications of ABT in patientswith anemia of chronic disease are volume overload andminor allergic reactions to plasma proteins (i.e., itchy hives).Bacterial contamination of the transfused blood componentsis probably the greatest infectious risk, because the risk of

426 Vamvakas

the traditional transfusion-transmitted viruses (e.g., the HIVor hepatitis viruses) has become extraordinarily small. Therisk of transmission of an emerging pathogen through trans-fusion is omnipresent, however, and this risk was demon-strated in North America in the summer and fall of 2002when WNV was reported to be transmitted by transfusionfor the first time (25). Theoretical infectious risks of ABT alsoexist, and they may turn out to be significant if they areshown to be real.

The ABT-related TRIMmay also be shown to be a seriouscomplication of ABT in the future. For the time being, however,results from available RCTs do not support any of the threehypotheses attributing a purported adverse TRIM effect ofABT to: (1) immunologically active allogeneic WBCs, (2)WBC-derived soluble mediators, or (3) soluble HLA moleculescirculating in allogeneic plasma. Moreover, the implementa-tion of universal WBC reduction in many Western Europeancountries and in Canada abrogated concerns about possibleTRIM effects mediated by allogeneic WBCs. The possibility ofTRIM effects mediated by allogeneic plasma still exists, butevidence that supports plasma-mediated immunomodulationis scarce (44,45,71).

The existence of an adverse TRIM effect of ABT has beenhard to demonstrate in observational studies, because of theconfounding effect of severity of illness. Patients with moresevere (or advanced) disease require transfusion more often,and they also experience a worse outcome than other patientsbecause of their more severe (or advanced) disease. Thus,ABT is bound to have an adverse effect on a patient’s prog-nosis in observational studies because severity of illness can-not be objectively quantified or adjusted for by statisticaltechniques (72). In fact, the literature is replete with observa-tional studies showing an association between ABT and worse(even much worse) clinical outcomes (73).

Because of the confounding effect of severity of illness,whether ABT per se has an adverse effect on a patient’s prog-nosis can be ascertained only by carefully conducted RCTs.Although more research is needed to answer this question,the hitherto available data from RCTs do not support such an


assertion. Therefore, in decidingwhether to transfuse apatientwith anemia of chronic disease, attention should be paid toestablished transfusion risks, as opposed to such purportedrisks as transfusion-associated immunomodulation.

REFERENCES

1. Eschbach JW, Egrie JC, Downing MR, Browne JK,Adamson JW. Correction of the anemia of end-stage renaldisease with recombinant human erythropoietin: results of acombined phase I and II clinical trial. N Engl J Med 1987; 316:73–78.

2. Eschbach JW, Abdulhadi MH, Browne JK, Delano BG, DowningMR, Egrie JC, Evans RW, Friedman EA, Graber SE, Haley NR,Adamson JW. Recombinant human erythropoietin in anemicpatients with end-stage renal disease: results of a phase IIImulticenter clinical trial. Ann Intern Med 1989; 111:992–1000.

3. Adamson JW. Cytokine biology: implications for transfusionmedicine. Cancer 1991; 67:2708–2711.

4. Collier AC, Kalish LA, Busch MP, Gernsheimer T, AssmannSF, Lane TA, Asmuth DM, LedermanMM,Murphy EL, KumarP, Kelley M, Flanigan TP, McMahon DK, Sacks HS, KennedyMS, Holland PV. Leukocyte-reduced red blood cell transfusionsin patients with anemia and human immunodeficiency virusinfection: the viral activation transfusion study: a randomizedcontrolled trial. JAMA 2001; 285:1592–1601.

5. Vamvakas EC, Blajchman MA. Deleterious clinical effects oftransfusion-associated immunomodulation: fact or fiction?Blood 2001; 97:1180–1195.

6. Vamvakas EC. Blood transfusion and virus activation. In:Vamvakas EC, Blajchman MA, eds. ImmunomodulatoryEffects of Blood Transfusion. Bethesda: AABB Press, 1999:213–235.

7. Menitove JE. Red cell transfusion therapy in chronic anemia.In: Mintz PE, ed. Transfusion Therapy: Clinical Principlesand Practice. Bethesda: AABB Press, 1999:1–12.

428 Vamvakas

8. American Society of Anesthesiologists Task Force on BloodComponent Therapy. Practice guidelines for blood componenttherapy. Anesthesiology 1996; 84:732–747.

9. Office of Medical Applications of Research, National Institutesof Health. Preoperative red blood cell transfusion. JAMA 1988;260:2700–2703.

10. Welch HG, Meehan KR, Goodnough LT. Prudent strategies forelective red blood cell transfusion. Ann Intern Med 1992;116:393–402.

11. Expert Working Group. Guidelines for red blood cell andplasma transfusion for adults and children. Can Med Assoc J1997; 156:S1–S24.

12. Sherrard DJ. Perioperative indications for red blood cell trans-fusion—has the pendulum swung too far? Mayo Clin Proc1993; 68:512–514.

13. Canadian Erythropoietin Study Group. Association betweenrecombinant human erythropoietin and quality of life andexercise capacity of patients receiving hemodialysis. Br MedJ 1990; 300:573–578.

14. Case DC, Bukowski RM, Carey RW, Fishkin EH, Henry DH,Jacobson RJ, Jones SE, Keller AM, Kugler JW, Nichols CR.Recombinant human erythropoietin therapy for anemic cancerpatients on combination chemotherapy. J Natl Cancer Inst1993; 5:801–806.

15. Cummings JP. Technology Assessment: Red Cell TransfusionGuidelines. Oakbrook: UHC Services Corporation, 1977.

16. Audet AM, Andrejewski C, Popovsky MA. Red blood cell trans-fusion practices in patients undergoing orthopedic surgery. Amulti-institutional analysis. Orthopedics 1998; 21:851–864.

17. PopovskyMA, Taswell HF. Circulatory overload: an underdiag-nosed consequence of transfusion [abstr]. Transfusion 1985;25:469.

18. Pineda AA, Taswell HF. Transfusion reactions associated withanti-IgA: report of four cases and review of the literature.Transfusion 1975; 15:10–15.


19. Vyas GN, Perkins HA, Fudenberg HH. Anaphylactoid transfu-sion reactions associated with anti-IgA. Lancet 1968; ii:312–315.

20. Eckrich RJ, Malory DM, Sandler SG. Laboratory tests toexclude IgA deficiency in the investigation of suspected anti-IgA transfusion reactions. Transfusion 1993; 33:1–5.

21. Popovsky MA, Moore SB. Diagnostic and pathogenetic consid-erations in transfusion-related acute lung injury. Transfusion1985; 25:573–577.

22. Moore SB, Taswell HF, Pineda AA, Sonnenberg CL. Delayedhemolytic transfusion reactions. Evidence of the need for animproved pretransfusion compatibility test. Am J Clin Pathol1980; 74:94–97.

23. Vamvakas EC, Pineda AA, Reisner R, Santrach P, Moore SB.The differentiation of delayed hemolytic and delayed serologictransfusion reactions: incidence and predictors of hemolysis.Transfusion 1995; 35:26–32.

24. Lau P, Sholtis CM, Aster RH. Post-transfusion purpura: anenigma of alloimmunization. Am J Hematol 1980; 9:331–336.

25. Biggerstaff BJ, Peterson LR. Estimated risk of transmission ofWest Nile Virus through blood transfusion in the U.S., 2002.Transfusion 2003; 43:1007–10017.

26. Kleinman SH, Busch MP, Korelitz JJ, Schreiber GB. The inci-dence=window period model and its use to assess the risk oftransfusion-transmitted HIV and HCV infection. Transf MedRev 1997; 11:155–172.

27. Kleinman S, Chan P, Robillard P. Risks associated with trans-fusion of cellular blood components in Canada. Transf MedRev 2003; 17:120–162.

28. Kleinman SH, Busch MP. The risks of transfusion-transmittedinfection: direct estimation and mathematical modeling.Baillere’s Clin Haematol 2000; 13:631–649.

29. Allain JP, Hewitt PE, Tedder RS, Williamson LM. Evidencethat anti-HBc but not HBV DNA testing may prevent someHBV transmission by transfusion. Br J Haem 1999; 107:186–195.

430 Vamvakas

30. Glynn S. Residual risks for hepatitis and HIV and unreporteddeferrable risk in U.S. blood donors. Proceedings of theCanadian Blood Services Consensus Conference: ‘‘Blood-BorneHIV and Hepatitis: Optimizing the Donor Selection Process’’,Ottawa, Ont., Canada, Nov 7–9, 2001. Transf Med Rev2003;17:1–30.

31. Schreiber GB, Busch MP, Kleinman SH, Korelitz JJ. The riskof transfusion-transmitted viral infections. N Engl J Med1996; 334:1685–1690.

32. Kuehnert MJ, Roth VR, Haley NR, Gregory KR, Elder KV,Schreiber GB, Arduino MJ, Holt SC, Carson LA, BanerjeeSN, Jarvis WR. Transfusion-transmitted bacterial infectionin the United States, 1998 through 2000. Transfusion 2001;41:1493–1499.

33. Leiby DA, Kerr KL, Campos JM, Dodd RY. A retrospectiveanalysis of microbial contaminants in outdated random-donorplatelets from multiple sites. Transfusion 1997; 37:259–263.

34. Brown P. Can Creutzfeldt–Jakob disease be transmitted bytransfusion? Curr Opin Hematol 1995; 2:472–477.

35. Ricketts MN, Cashman NR, Stratton EE, Elsaadany S. IsCreutzfeld–Jakob disease transmitted by blood? Emerg InfectDis 1997; 3:155–163.

36. CJD investigational lookback study. AABBNews 1999; 21[4]:7.

37. Evatt B, AustinH, Banhart E, Schonberger L, Sharer L, Jones R,DeArmond S. Surveillance for Creutzfeldt–Jakob diseaseamong persons with hemophilia. Transfusion 1998; 38:817–820.

38. Murphy M. New variant Creutzfeldt–Jakob disease (nvCJD):the risk of transmission by blood transfusion and the potentialbenefit of leukocyte-reduction of blood components. TransfMed Rev 1999; 13:75–83.

39. Kleinman S. New variant Creutzfeldt–Jakob disease andwhite cell reduction: risk assessment and decision-making inthe absence of data. Transfusion 1999; 39:920–924.

40. Peden AH, Head MW, Ritchie DL, Bell JE, Ironside JW. Pre-clinical vCJD after blood transfusion in a PRNP codon 129 het-erozygous patient. Lancet 2004, 364:527–529.


41. Food and Drug Administration. Guidance for industry: recom-mendations for the assessment of donor suitability and bloodproduct safety in cases of suspected severe acute respiratorysyndrome (SARS) or exposure to SARS. (Final guidance, April2003). Rockville, MD: Centers for Biologics Evaluation andResearch, U.S. Department of Health and Human Services.

42. Bordin JO, Heddle NM, Blajchman MA. Biologic effects ofleukocytes present in transfused cellular blood products. Blood1994; 84:1705–1721.

43. Nielsen HJ, Reimert CM, Pedersen AN, Brunner N,Edvardsen L, Dybkjaer E, Kehlet H, Skov PS. Time-dependent,spontaneous release of white cell- and platelet-derived bioac-tive substances from stored human blood. Transfusion 1996;36:960–965.

44. Ghio M, Contini P, Mazzei C, Brenci S, Barberis G, Filaci G,Indiveri F, Puppo F. Soluble HLA Class I, HLA Class II, andFas ligand in blood components: a possible key to explain theimmunomodulatory effects of allogeneic blood transfusion.Blood 1999; 93:1770–1777.

45. Magee CC, Sayegh MH. Peptide-mediated immunosuppres-sion. Curr Opin Immunol 1997; 9:669–675.

46. Vamvakas EC. Possible mechanisms of allogeneic blood trans-fusion-associated postoperative infection. Transf Med Rev2002; 16:144–160.

47. Blajchman MA, Bardossy L, Carmen R, Sastry A, Singal DP.Allogeneic blood transfusion-induced enhancement of tumorgrowth: two animal models showing amelioration of leukode-pletion and passive transfer using spleen cells. Blood 1993;81:1880–1882.

48. Kao KJ. Induction of humoral immune tolerance to major his-tocompatibility complex antigens by transfusions of UV-Birradiated leukocytes. Blood 1996; 88:4375–4382.

49. Houbiers JGA, Brand A, van de Watering LMG, Hermans J,Verwey PJM, Bijnen AB, Pahlplatz P, Eeftinck Schattenkerk M,Wobbes TH, de Vries JE, Klementschitsch P, van de MaasAHM, van de Velde CJH. Randomized controlled trial comparingtransfusion of leukocyte-depleted or buffy-coat-depleted blood insurgery for colorectal cancer. Lancet 1994; 344:573–578.

432 Vamvakas

50. Jensen LS, Kissmeyer-Nielsen P, Wolff B, Qvist N. Rando-mized comparison of leucocyte-depleted versus buffy-coat-poorblood transfusion and complications after colorectal surgery.Lancet 1996; 34:841–845.

51. van de Watering LMG, Hermans J, Houbiers JGA, van deBroek PJ, Bouter H, Boer F, Harvey MS, Huysmans HA,Brand A. Beneficial effect of leukocyte depletion of transfusedblood on post-operative complications in patients undergoingcardiac surgery: a randomized clinical trial. Circulation1998; 97:562–568.

52. Tartter PI, Mohandas K, Azar P, Endres J, Kaplan J, SpivackM. Randomized trial comparing packed red cell blood transfu-sion with and without leukocyte depletion for gastrointestinalsurgery. Am J Surg 1988; 176:462–466.

53. Wallis JP, Chapman CE, Orr KE, Clark SC, Forty JR. Effects ofwhite cell reduction of transfused RBCs on postoperative infec-tion rates in cardiac surgery. Transfusion 2002; 42:1127–1134.

54. Titlestad IL, Ebbesen LS, Ainsworth AP, Lillevang ST, QvistN, Georgsen J. Leukocyte-depletion of blood components doesnot significantly reduce the risk of infectious complications:results of a double-blind, randomized study. Int J ColorectalDis 2001; 16:147–153.

55. Bilgin YM, van de Watering LMG, Lorinser JE, VersteeghMIM, Eijsman L, Oers M, Brand A. Double-blind, randomizedcontrolled trial on the effect of leukocyte-depleted erythrocytetransfusions in cardic valve surgery. Circulation 2004;109:2755–2760.

56. Jensen LS, Andersen AJ, Christiansen PM, Hokland P,Juhl CO, Madsen G, Mortensen J, Moller-Nielsen C, Hanberg-Sorensen F, Hokland M. Postoperative infection and naturalkiller cell function following blood transfusion in patientsundergoing elective colorectal surgery. Br J Surg 1992; 79:513–516.

57. Innerhofer P, Luz G, Spotl L, Hobisch-Hagen P, Schoberers-berger W, Fischer M, Nussbaumer W, Lochs A, Irschick E.Immunologic changes following transfusion of autologous orallogeneic buffy-coat-poor versus leukocyte-depleted blood inpatients undergoing arthroplasty. Proliferative T-cell


responses and T-helper=T-suppressor cell balance. Transfusion1999; 39:1089–1096.

58. Innerhofer P, Tilz G, Fuchs D, Luz G, Hobish-Hagen P,Schoberersberger W, Nussbaumer W, Lochs A, Irschick E.Immunologic changes following transfusion of autologous orallogeneic buffy-coat-poor versus leukocyte-depleted bloodtransfusions in patients undergoing arthroplasty. II. Activa-tion of T-cells, macrophages and cell-mediated lympholysis.Transfusion 2000; 40:821–827.

59. Vamvakas EC. Meta-analysis of randomized controlled trialsinvestigating the risk of postoperative infection in associationwith white-blood-cell-containing allogeneic blood transfusion:the effects of the type of transfused red blood cell productand surgical setting. Transf Med Rev 2002; 16:304–314.

60. Hebert PC, Fergusson D, Blajchman MA, Wells GA, Kmetic A,Coyle D, Heddle N, Germain M, Goldman M, Toye B, Schweit-zer I, van Walraven C, Devine D, Sher GD. Clinical outcomesfollowing institution of the Canadian universal leukoreductionprogram for red blood cell transfusions. JAMA 2003; 289:1941–1949.

61. Fergusson D, Hebert PC, Lee SK, Walker CR, Barrington KJ,Joseph L, Blajchman M, Shapiro S. Clinical outcomes follow-ing institution of universal leukoreduction of blood transfu-sions for premature infants. JAMA 2003; 289:1950–1956.

62. Corwin HL, AuBuchon JP. Is leukoreduction of blood compo-nents for everyone? JAMA 2003; 289:1993–1995.

63. Baron JF, Gourdin M, Bertrand M, Mercadier A, Delort J,Kieffer E, Coriat P. The effect of universal leukodepletion ofpacked red blood cells on postoperative infections in high-riskpatients undergoing abdominal aortic surgery. Anesth Analg2002; 94:529–537.

64. Llewelyn CA, Taylor RS, Todd AA, Stevens W, Murphy MF,Williamson LM. The effect of Universal leukoreduction onpostoperative infections and length of stay in elective orthope-dic and cardiac surgery. Transfusion 2004; 44:489–500.

65. Dzik WH, Anderson JK, O’Neill EM, Assmann SF, Kalish LA,Stowell CP. A prospective, randomized clinical trial of univer-sal WBC reduction. Transfusion 2002; 42:1114–1122.

434 Vamvakas

66. Bracey AW, Radovancevik R, Nussimeier NA, LaRocco M,Vaughn WK, Cooper JR. Leukocyte-reduced blood in open-heart surgery patients: effects on outcome. Transfusion 2002;42(suppl):5S.

67. Vamvakas E. Perioperative blood transfusion and cancer recur-rence: meta-analysis for explanation. Transfusion 1995; 35:760–768.

68. Busch ORC, Hop WCJ, van Papendrecht MAWH, Marquet RI,Jeekel J. Blood transfusions and prognosis in colorectalcancer. N Engl J Med 1993; 328:1372–1376.

69. Heiss MN, Mempel W, Delanoff C, Jausch K-W, Gabka C,Mempel M, Dietrich H-J, Eissner H-J, Schildberg F-W. Bloodtransfusion-modulated tumor recurrence: first results of arandomized study of autologous versus allogeneic bloodtransfusion in colorectal cancer surgery. J Clin Oncol 1994; 12:1859–1867.

70. McAlister FA, Clark HD, Wells PS, Laupacis A. Perioperativeallogeneic blood transfusion does not cause adverse sequelaein patients with cancer: a meta-analysis of unconfoundedstudies. Br J Surg 1998; 85:171–178.

71. Vamvakas E, Carven JH. Exposure to allogeneic plasma andrisk of postoperative pneumonia and=or wound infection incoronary artery bypass graft surgery. Transfusion 2002; 42:107–113.

72. Vamvakas E, Moore SB. Blood transfusion and postoperativeseptic complications. Transfusion 1994; 34:714–727.

73. Blumberg N, Heal JM. Effects of transfusion on immune func-tion: cancer recurrence and infection. Arch Pathol Lab Med1994; 118:371–379.


17

Iron and Erythropoietin

LAWRENCE T. GOODNOUGH

Departments of Pathology and Medicine,Stanford University, Palo Alto,

California, U.S.A.

INTRODUCTION

Anemia has traditionally been identified as an abnormallaboratory value, with a focus on whether or not it shouldbe treated, rather than perceived as a serious clinical condi-tion for which treatment is mandated (1). For example, a deci-sion to manage anemia with blood transfusion is based on therelative risks and benefits, in which the estimated risks of ablood transfusion are quantifiable (2,3) and can be communi-cated to patients. But the magnitude of risks associated withan (untreated) anemia has been poorly understood, andhas not been effectively conveyed to patients. In contrast totreatment strategies for other diseases, guidelines for thegeneral management of anemia have only occasionally been

437

developed (4–6) and are usually limited to patients in particu-lar clinical settings such as the anemia of chronic kidneydisease (7,7a), anemia associated with malignancy or cancertreatment (8), or surgery (9,10).

PREVALENCE OF ANEMIA OFCHRONIC DISEASE

The prevalence of anemia in patients associated with disease(anemia of chronic disease, ACD) has been studied in somedepth. ACD has been identified in 30–70% of patients withchronic liver disease (11), 27% of patients with rheumatoidarthritis (12), and 28–55% (dependent upon the extent of dis-ease) with HIV infection (13). Anemia associated with canceror cancer treatment varies widely with stage of disease and=ortreatment. In small cell carcinoma of the lung, anemia wasidentified in 30–60% of patients, depending on treatment sta-tus (14). In an analysis of Medicare patients, the percentage ofpatients with a diagnosis of anemia in addition to a primarydiagnosis was 20% for AIDS, 17% for rheumatoid arthritis,21% for inflammatory bowel disease, 27% for chemotherapy,27% for radiation therapy, and 30% for congestive heartfailure (CHF) (1). These studies suggest that anemia occurscommonly in conjunction with a chronic illness. If anemiahad no independent consequences, then these statistics repre-sent just abnormal laboratory values. However, a growingbody of evidence underscores the independent impact ofanemia on a variety of clinical indicators, and suggests thattreatment of anemia is effective in improving patientoutcomes.

ANEMIA AND ADVERSE OUTCOMES

Mortality

Higher mortality rates are almost always observed in patientswith anemia. Does anemia merely reflect a sicker patient,or does anemia itself have an incremental impact? Thisrelationship has been explored most fully in patients

438 Goodnough

undergoing chronic hemodialysis. In a retrospective review(15) of nearly 20,000 hemodialysis patients, levels ofHgb�8.0 g=dL were associated with a twofold increase in oddsof death when compared with Hgb ranges of 10.0–11.0 g=dL. Asimilar study (16) of nearly 100,000 hemodialysis patients con-firmed that Hct> 30% was associated with a lower mortality.Compared to patients with Hct> 30%, the overall relative riskof death was 33–51% higher for the group with Hct< 7%, and12–20% higher for the group with Hct 27–30%, with and with-out adjustments for severity of disease. Moreover, patientswho began the study interval with Hct< 30% and were treatedto achieve Hct above 30% after one-year follow-up had an oddsratio of death that was no different from patients who beganand finished the study interval with Hct> 30%. Subsequentanalyses have determined that sustained Hct levels between33% and 36% were associated with the lowest risk of death(17,18). These studies provided the first evidence that man-agement of anemia, independent of other risk factors,improves mortality.

Data on the impact of anemia on survival are also avail-able in patients with medical conditions other than chronickidney disease. A growing body of evidence suggests thatanemia influences mortality in patients with CHF or withischemic heart disease. Almost 20 years ago, hemoglobin(Hgb) was identified as one of eight significant factors predict-ing prognosis and response to treatment in patients with CHF(19). Several large observational studies noted an associationbetween anemia and increased mortality for patients withcardiovascular disease and suggested that such patients donot tolerate anemia as well as patients with other diagnoses(20–22).

Several large studies have identified a strong correlationbetween level of Hgb and survival in patients with CHF(23–25). A prospective, randomized trial in patients withCHF explored whether the treatment of anemia influencedoutcomes. Patients enrolled in the study had moderate-to-severe CHF (New York Heart Association (NYHA) Class III–IV), an LVEF of� 40% despite maximally tolerated doses ofCHF medications, and Hgb levels that were persistently

Iron and Erythropoietin 439

between 10.0 and 11.5 g=dL (26). Over a mean of 8.2� 2.6months, the Hgb levels in the treatment cohort increased from10.3 to 12.9 g=dL, compared to 10.9 to 10.8 g=dL in the controlgroup. Twenty-five percent of the control group died duringthe study interval, compared to none in the treatment group.

A recent retrospective observational analysis (27) of78,974 elderly patients hospitalized with acute myocardialinfarction (MI) in the United States found that blood transfu-sions in patients with admission hematocrits less than 33%were associated with significantly lower 30-day mortality.The positive transfusion effect disappeared, however, forpatients with hematocrits greater than 33% upon admission.For these patients, transfusions were associated with a higherodds ratio of mortality, in which underlying comorbiditiesmay have overridden any positive transfusion effect. Bothadvanced age (28,29) and the presence of flow limiting coron-ary stenosis (30) markedly impair the cardiac compensatoryresponse to anemia, even without the added insult of acutemyocardial injury. These, among other limits on physiologicreserve, may explain why levels of hemoglobin tolerated byyounger patients would not be tolerated by the elderly, andwhy elderly patients with acute myocardial infarction repre-sent a group at extremely high risk for mortality, despiteinfarct sizes similar to younger patients (31). In the absenceof prospective data or other data to the contrary, a substantialnumber of lives may be saved when patients who present withacute myocardial infarction are maintained at hematocritlevels >33% (32).

Morbidity

In addition to its impact upon mortality, anemia also signifi-cantly influences patient morbidity. In patients with end-stage renal disease (ESRD), multiple studies support thisassertion. One study showed that in patients on hemodialysis,the risk of hospitalization declines with improvement inhematocrit (Hct), with a 16–22% lower hospitalization ratefor patients with Hct levels between 36% and 39% comparedto those withHct levels between 33% and 36% (33). Prospective

440 Goodnough

clinical trials in ESRD patients have demonstrated a relation-ship betweenHct levels, left ventricular dilatation (34), and leftventricular hypertrophy (LVH) (35–38). Management of theanemia led to improvement in LVH, thus supporting the exis-tence of a cause–effect relationship between anemia and LVHin this setting (39). In other studies, patients with ESRD whounderwent treatment for anemia had increased cognitive func-tion, increased exercise tolerance, increased quality of life, andreduced disease progression (40–43).

In patients with chronic kidney disease and CHF, treat-ment of anemia improves many of the abnormalities seen inCHF: reducing LVH (44–46), preventing left ventricular dila-tation (47), and increasing the left ventricular ejection frac-tion (LVEF) (45–47), the stroke volume, and the cardiacoutput (48). Despite this association between anemia andCHF, anemia is not mentioned in guidelines for the diagnosisand treatment of CHF (49). As discussed in the section onMortality, a prospective, randomized trial studied the treat-ment of anemia in patients with moderate-to-severe CHF(NYHA Class III–IV) who had an LVEF of �40%. Patientswho received treatment had a 42.1% improvement in NYHAClass, compared to the control cohort who had a decrease of11.4% (26). The number of hospitalized days, need for diuretictherapy, and renal function impairment were all significantlyhigher in the control group compared to the treatment group.

An intriguing association has also been observed betweenanemia and disease progression among patients undergoingradiotherapy (RT), particularly in those with cervical carci-noma and squamous cell carcinoma of the head and neck. Inwomen with cervical cancer, two-thirds are anemic at base-line, and 82% are anemic during RT (50). Correlationsbetween anemia, tumor tissue oxygenation, local recurrence,and survival have been demonstrated (51,52). The Gynecol-ogy Oncology Group (GOG) is currently conducting a prospec-tive, randomized clinical trial on the effect of anemiamanagement on rate of disease recurrence and on survivalin patients with cervical carcinoma. In patients with headand neck cancer, 75% of patients undergoing combined che-motherapy and radiotherapy become anemic (Hgb< 8 g=dL)


(53) and anemia has been associated with worse local regionalcontrol and survival (54,55). On the basis of these observa-tions, the Radiation Therapy Oncology Group (RTOG) hasinitiated a prospective, randomized study of the effect of ane-mia management on rate of disease recurrence and survivalin patients with head and neck cancer undergoing combinedmodality therapy.

Quality of Life

In addition to reductions in the quantity of life, patients withanemia have significantly impaired quality of life. Multiplestudies have examined this issue in patients with kidney fail-ure on dialysis. For example, in one study, patients on hemo-dialysis underwent interventions to increase the Hct from anaverage of 30.9% to 38.4% (and Hgb from 10.2 to 12.5 g=dL).Treatment of the anemia was associated with improvementsin quality of life based upon both Karnofsky Score andSickness Impact Profile (56).

For cancer patients undergoing chemotherapy, several(57–60) large, community-based prospective (unblinded andnoncontrolled) studies of patients with solid, nonmyeloidmalignancies found significant improvements in quality of lifewith anemia management over a four-month interval. Incre-mental analysis determined that significantQOLchangeswereassociated with Hgb ranges of 8–14 g=dL, and the largestimprovement in QOL occurred between Hgb 11 and 12g=dL(61). This relationship was maintained after controlling fortumor type and status, transfusions, and extent of chemother-apy or radiotherapy. These findings have been corroborated bytwo prospective, randomized and double-blinded studies(62,63), both of which showed significant improvement inQOL with anemia management. Evidence-based guidelinesdeveloped by members of The Canadian Cancer and AnemiaGuidelines Development Group concluded that erythropoietintherapy produced clinically relevant improvements in QOL inpatientswith cancer (64). TheAmerican Society ofHematology(ASH) and The American Society of Clinical Oncology (ASCO)have jointly developed guidelines regarding the managementof anemia in patients with cancer (8).

442 Goodnough

MANAGEMENT OF ANEMIA

Despite the compelling evidence that anemia is an indepen-dent contributor to poor outcomes, anemia is frequently over-looked and untreated. In a study of 200,000 patients enrolledin a Health Maintenance Organization between 1994 and1997 (65), 23% of patients with chronic kidney disease(defined as gender-specific, elevated creatinine concentra-tions) had Hct levels <30%, of whom only 30% were receivingtreatment for anemia.

Anemia has been identified as a risk factor for the devel-opment of left ventricular hypertrophy (LVH) in CKD (37–39)and LVH has been identified as a risk factor for mortality inthis setting (66). Yet, the mean Hct in patients starting dialy-sis between April 1, 1995 and December 31, 1999 was only29.3%, and a suboptimal percentage of dialysis patientsreceive erythropoietin therapy (67). With the relationshipbetween anemia and morbidity and mortality in patientsundergoing hemodialysis established, the National KidneyFoundation’s Kidney Disease Outcomes Quality Initiatives(K=DOQI) Guidelines recommend that these patients bemaintained at Hgb levels between 11 and 12g=dL (7,7a).Yet, in a study of patients on peritoneal dialysis, 11% hadHgb levels below 10 g=dL (68).

In patients with cancer, one review reported an extre-mely high incidence of anemia (14). Depending upon thedefinition used, the incidence ranged from 80% (WHO Grade3–4, Hgb between 8.0 and 6.5g=dL) to 100% (WHO Grade 1–2,Hgb 11.0–9.5 g=dL. Yet, fewer than 30% of patients withHgb< 10g=dL were treated for their anemia.

ERYTHROPOIETIN, IRON, ANDERYTHROPOIESIS

Emerging understanding regarding the relationship betweenerythropoietin, iron, and erythropoiesis in patients under-going recombinant human erythropoietin therapy (EPO) hasimplications for management of the anemia of chronic disease.


During EPO therapy, iron-restricted erythropoiesis is evidenteven in the presence of storage iron and oral iron supplemen-tation. For example, intravenous iron therapy in renal dialysispatients undergoing EPO therapy can produce hematologicalresponses in patients with serum ferritin levels up to400mg=L, indicating that traditional biochemical markers ofstorage iron in patients with anemia of chronic disease havelimitations in assessment of iron status. Newermeasurementsof erythrocyte and reticulocyte indices using automated coun-ters show promise in the evaluation of iron-restricted erythro-poiesis. The availability of safer intravenous iron preparationsallows for carefully controlled studies of their value in patientswith chronic disease.

Clinical settings have served as ‘‘natural experiments’’that have furthered our understanding of the relationshipbetween erythropoietin, iron, and the erythropoietic responseto anemia in man. Nearly 20 years ago, Finch (69) summar-ized the knowledge gained primarily from studies of normalindividuals, patients with hereditary hemolytic anemias,and patients with hemochromatosis. Under conditions of basalerythropoiesis in normal subjects, plasma iron turnover (as anindex of marrow erythropoietic response) is little affected,whether transferrin saturation ranges from very low to veryhigh levels. In contrast, the erythropoietic response in indivi-duals with congenital hemolytic anemia, in whom erythropoi-esis is chronically raised up to sixfold over basal levels (70), isaffected (and limited) by serum iron levels and by transferrinsaturation (71). Patients with hemochromatosis who under-went serial phlebotomy were observed to mount erythropoieticresponses of up to eightfold over basal rates, attributed to themaintenance of very high serum iron and transferrin satura-tion levels in these patients (72), whereas normal individualswere shown to have difficulty providing sufficient iron to sup-port rates of erythropoiesis greater than three times basalrates (73). These observations led Finch (74) to identify a ‘‘rela-tive iron deficiency’’ state, defined as circumstances whenincreased erythron iron requirements exceed the available sup-ply of iron. Insights gained over the last 20 years regarding therelationship between erythropoietin, iron, and erythropoiesis

444 Goodnough

in patients with anemia (75), alongwith implications for clinicalmanagement of patients with anemia, will be reviewed.

ERYTHROPOIETIN RESPONSE TO ANEMIA

The practice of autologous blood donation in patientsscheduled for elective surgery is a natural experiment in indu-cing anemia. Patients undergoing autologous blood phlebot-omy may donate a unit (450 � 45mL) of blood as often astwice weekly, until 72 hr before surgery (76). Under routineconditions, patients usually donate once weekly (77). Oral ironsupplements are routinely prescribed. The associated anemiais accompanied by a response in endogenous erythropoietinlevels that, while increased significantly over basal levels,remain within the range of normal (4–26m=L) (78). The ery-thropoietic response that occurs under these conditions ismodest (75,76). A summary of selected prospective, controlledtrials (79–84) of patients undergoing phlebotomy is presentedin Table 1. Calculated estimates of red blood cell (RBC) volumeexpansion (erythropoiesis in excess of basal rates) were deter-mined (85). Two hundred and twenty to 351mL [11–19% RBCexpansion (79,80) or the equivalent to one to 1.75 blood units(86)] are produced in excess of basal erythropoiesis, definingthe efficacy of this blood conservation practice.

For patients subjected to more aggressive (up to two unitsweekly) phlebotomy, the endogenous erythropoietin response ismore substantial (81–84). In one clinical trial (82), a linear-logarithmic relationship was demonstrated between change inhemoglobin level and erythropoietin response (87), predictedpreviously by phlebotomy experiments in normal subjects(88). Erythropoietin-mediated erythropoiesis in this setting is397–568mL [19–26% RBC expansion (81–84) or the equivalentof two to three blood units (86)].

Clinical trials have demonstrated a dose–response rela-tionship between erythropoietin and red blood cell expansion(84). A study of ‘‘very low’’ dose EPO therapy in autologousblood donors found that 400u=kg administered over a two-week interval resulted in clinically significant erythropoiesis


Table

1Endog

enou

sErythropoietin-M

ediatedErythropoiesis

Patien

ts(n)

Baseline

RBC

(mL)

Req

uested=

don

ated(units)

RBC

(mL)

Iron

therapy

Referen

ces

Don

ated

Produced

Expansion

(%)

Standard

phlebotom

y108

1884

32.7

522

351

19

PO

79

22

1936

32.8

590

220

11

Non

e80

45

1991

32.9

621

331

17

PO

80

41

1918

32.9

603

315

16

PO

þIV

80

Aggressivephlebotom

y30

2075

�3

3.0

540

397

19

Non

e81

30

2024

�3

3.1

558

473

23

PO

81

30

2057

�3

2.9

522

436

21

IV81

24

2157

64.1

683

568

26

PO

82,83

23

2257

64.6

757

440

19

PO

84

Data

expressed

asmea

ns.

PO

¼Oral;IV

¼intraven

ous.

Mod

ified

from

Ref.75.

446 Goodnough

(89). Table 2 details red cell volume expansion in 134 patientstreated with EPO therapy during aggressive blood phlebot-omy (82–84,90,91), ranging from 358 to 1764mL (28–79%RBC expansion) over 25–35 days, or the equivalent of two tonine blood units (86). The range in response (erythropoiesis)to dose (erythropoietin) is not related to patient gender orage (92,93), suggesting that patient-specific factors such asaccompanying chronic disease, iron-restricted erythropoiesis,or other factors that normally cause the wide distribution ofthe hemoglobin level, account for the variability in erythro-poietic response to erythropoietin.

Studies in patients donating blood who also have anemiaof chronic disease [osteo-arthritis (94–96) or rheumatoidarthritis (97,98)] are summarized in Table 3. Red cell volumeexpansion ranged from 157 to 353mL (11–24%) for endogen-ous erythropoietin-mediated erythropoiesis and 268 to 673mL(21–40%) with erythropoietin therapy. These erythropoieticresponses are indistinguishable from patients with anemiafrom blood loss alone, noted in Tables 1 and 2. An additionalstudy of 17 patients with inflammatory bowel disease treatedwith EPO and oral iron therapy demonstrated a similar res-ponse, with an estimated 20% increase in red cell volumewhen compared to placebo-treated patients (99).

IRON-RESTRICTED ERYTHROPOIESIS

Erythropoiesis in response to aggressive autologousphlebotomy via endogenous erythropoietin has been estimatedto increase by up to threefold (84,100). No apparent relation-ship exists between basal iron stores and thismagnitude of ery-thropoiesis, suggesting that under conditions of moderateerythropoiesis, serum iron and transferrin saturation for ery-thron requirements are adequatelymaintained by storage iron(81–84). Little or no benefit to oral iron supplementation wasfound in two studies (81,101), whereas a third study (12) foundsome benefit (Table 1). Intravenous iron supplementation wasnot found to be of value in enhancing erythropoiesis underthese conditions (80,81).


Table

2Erythropoiesis

DuringBlood

LossandErythropoietin

(EPO)Therapy

Patien

ts(n=sex)

Baseline

RBC

(mL)

TotalEPO

dose(u=kg)

Units

don

ated

RBC

(mL)

Iron

therapy

Referen

ces

Don

ated

Produced

Expansion

(%)

(10=F)

1285

900SQ

3.4

435

358

28

IV90

(24)

1949

900IV

5.2

864

621

32

PO

84

(10=F)

1293

1800SQ

4.3

526

474

37

IV90

(26)

2032

1800IV

5.5

917

644

32

PO

84

(11=F)

1796

3600IV

4.9

809

701

39

PO

82,83

(12=M)

2296

3600IV

5.9

1097

1102

48

PO

82,83

(23)

2049

3600IV

5.4

970

911

45

PO

82,83

(18)

2019

3600IV

5.6

972

856

42

PO

84

(1=M)

2241

4200IV

81600

1764

79

Hem

a-

chromatosis

91

Data

expressed

asmea

ns.

PO

¼oral;IV

¼intraven

ous.

Mod

ified

from

Ref.75.

Table

3Erythropoietin

(EPO)andErythropoiesis

inPatien

tswithAnem

iaaof

Chronic

Disea

se

Patien

ts(n)

Units

don

ated

RBC

(mL)

Iron

Rx

Referen

ces

Produced

Expansion

(%)

I.Osteoarthritis

1.Placebo

62.6

157

11

PO

94

Placebo

33.3

220

18

PO

þIV

EPO

(1800u=kgIV

)b10

3.7

268

21

PO

EPO

(1800u=kgIV

)9

5.2

560

43

PO

þIV

EPO

(3600u=kgIV

)8

4.0

289

22

PO

EPO

(3600u=kgIV

)12

5.0

515

40

PO

þIV

2.Placebo

77

3.0

353

24

PO

95

EPO

(3600u=kgIV

)75

4.5

673

44

PO

3.Placebo

26

Non

e4

0.3

PO

96c

Placebo

26

Non

e18

1IV

EPO

(1200u=kgSQ)

26

Non

e219

14

PO

EPO

(1200u=kgSQ)

26

Non

e220

15

IVII.Rheu

matoid

arthritis

Placebo

62.3

271

25

PO

97

EPO

(3600u=kgIV

)4

4.8

624

37

PO

EPO

(1800u=kgIV

)11

2.6

291

27

IV98

EPO

(800u=kgSQ)

11

2.5

337

27

IV

Data

expressed

asmea

ns.

Mod

ified

from

Ref.75.

aWithmea

surable

storageiron

.bEPO

istotaldosageof

EPO

administered.

c Perisurgicaltherapywithou

tautologou

sphlebotom

y.

With enhanced erythropoiesis during EPO therapy,iron-restricted erythropoiesis occurs even in patients withmeasurable storage iron (Fig. 1). Despite an eightfold increasein gastrointestinal iron absorption (102), serum ferritin andtransferrin saturation levels decline up to 50% with EPOtherapy (103). A fourfold increase in erythropoietic activityis accompanied by declining reticulocyte counts and theappearance of hypochromic red cells by the second week oferythropoietin therapy (91,104). In a study of escalating(400%) EPO dose administered to patients undergoingaggressive phlebotomy, the marrow erythropoietic indexincreased from 2.9-fold (with endogenous erythropoietin sti-mulation) to 3.6-fold over basal rates of erythropoiesis, repre-senting only a 58% increase in erythropoiesis (Fig. 2). Thesuperior erythropoietic response in a patient with hemochro-matosis further suggests iron-restricted erythropoiesis inpatients treated with EPO (91) (Table 2).

Anemia of chronic disease is multifactorial. In rheuma-toid arthritis (RA), 40% of patients can be identified to be irondeficient, whereas 60% of patients have a reticuloendothelia(RE) iron-replete anemia of chronic disease (105). Tumor

Figure 1 The relationship between initial storage iron (mg) andred blood cell volume expansion (mL=kg) in patients undergoingaggressive phlebotomy with erythropoietin (EPO) therapy. Linearregression analysis demonstrated a significant correlation (r¼ 0.6,p ¼ 0.02). (Reprinted with permission from Ref. 100.)

450 Goodnough

necrosis factor alpha (TNF a) has been demonstrated to havea key role in the cytokine-mediated impairment of red bloodcell precursors via increased apoptotic cell death (106). Treat-ment with chimeric monoclonal antibody (infliximab) directedagainst TNF a results in an increase in hemoglobin values(107). These results support the presence of RE iron in anemiaof chronic disease as a form of iron-banking rather than iron-trapping or blockage (105). Yet iron studies of anemia of chronicdisease in RA show that red cell precursors are primed to max-imize transferrin iron uptake (108). Erythropoietin therapy iseffective at mobilizing iron stores in patients with the anemiaof chronic disease. A randomized, double blind, placebo-con-trolled trial of EPO therapy in patients with RA resulted inan immediate and significant increase in hemoglobin, alongwith clinical evidence of improvement in disease activity (109).

The success of EPO therapy in correcting the anemia ofchronic renal failure has led to substantial clinical experiencein iron metabolism and erythropoiesis in this setting(110,111). Hypo-responsiveness to EPO therapy is a common

Figure 2 The erythropoietic response, as reflected in the bonemarrow erythropoietic index, in four cohorts of autologous donorstreated with placebo or escalating doses of EPO (m=kg) therapy.Erythropoietic response (mL=kg=day) was estimated for each treat-ment group, according to the formula: Bone marrow erythropoieticindex ¼ [(RBC expansion) þ (Baseline RBC production)] ¼ BaselineRBC production. (Reprinted with permission from Ref. 75.)


phenomenon (112,113) due to a variety of comorbid conditions,particularly aluminum toxicity and iron deficiency. Anemicpatients undergoing dialysis may show suboptimal responseto oral iron therapy for several reasons. During EPO therapy,absorption of iron increases up to fivefold (114). However, exter-nal iron losses, including hemodialysis and blood testing, exceedgastrointestinal iron absorption (111). Poor compliance due togastrointestinal symptoms is problematic, and significantlyreduced iron absorptionmay occurwith some newer iron formu-lations (115). Iron-restricted erythropoiesis is evident by clinicalresponses to ascorbate supplementation, thought to facilitatethe release of iron from reticuloendothelial stores and increasedironutilizationbyerythrons (113), aswell as the success of intra-venous iron therapy in reducing EPO dosage (111).

Intravenous iron administration has become standardtherapy in renal dialysis patients undergoing EPO therapy(116). Patients treated with intravenous iron (100mg twiceweekly) achieved a 46% reduction in EPO dosage required tomaintain hematocrit levels between 30% and 34%, comparedto patients supplemented with oral iron (111). In a study(117) of chronic renal failure (nondialysis) patients, two-thirdsof patients who were unresponsive to oral iron responded toweekly intravenous iron therapy. Improved erythropoiesisoccurred despite initial serum ferritin levels as high as 400mg=L (118), indicating that biochemical markers of storage ironin these patients are not helpful in evaluating iron-restrictederythropoiesis.

The effect of intravenous iron therapy in other patientswith the anemia of chronic disease undergoing EPO therapyis shown in Table 3. Patients with osteoarthritis and measur-able storage iron doubled their red cell expansion, from arange of 21% to 22% (with oral iron) to 40% to 43% with intra-venous iron (94). Intravenous iron therapy in iron-deficientpatients with inflammatory bowel disease also results inimproved responses to EPO therapy (119), compared to asimilar patient group who received oral iron supplementation(99). The clinical response to intravenous iron may beattributed to the effect of EPO therapy on iron mobilizationfrom the reticuloendothelial system (RES) into red cell

452 Goodnough

precursors (120). The risk=benefit profile of intravenous ironis controversial in anemic renal dialysis patients (7,7a,121)as well as in patients with anemia of chronic disease (122).

Intravenous iron can allow up to a fivefold erythropoieticresponse to significant blood loss anemia in normal indivi-duals (71,123). A greater rate of hemoglobin production isprobably not possible unless red marrow expands into yellowmarrow space, as is seen in hereditary anemias (70,123). Onelimitation to intravenous iron therapy in patients not under-going EPO therapy may be that much of the administerediron is transported into the RES as storage iron, where it isless readily available for erythropoiesis (124). For iron-deficient patients, 50% of intravenous iron is incorporatedinto hemoglobin within 3–4 weeks (125), whereas for patientswith anemia of chronic disease or renal failure, intravenousiron is less rapidly mobilized from the RES (126).

The value of intravenous iron administration in patientsreceiving EPO therapy outside the setting of renal dialysis iscurrently not established. In one clinical trial in patients pre-operatively (94), significantly greater erythropoietic responseswere seen with intravenous iron therapy compared to patientssupplemented with oral iron only (Table 3). Another studyfound no difference in red cell production between oral ironand intravenous iron therapy in patients before orthopedicsurgery (96).

A third study found that intravenous iron supplementa-tion was not accompanied by an erythropoietic response corre-sponding to increasing doses of EPO therapy; a twofoldincrease in EPO dose was associated with only a 32% increasein red cell production (90), similar to the dose–responserelationship utilizing oral iron supplementation (84). Intrave-nous iron administered to normal subjects treated with EPOabolished the marked reduction in serum ferritin andincreased the reticulocyte hemoglobin content (a measure ing=L of the hemoglobin contained in all reticulocytes); how-ever, the total number of reticulocytes generated over eightdays after therapy was not affected (127). A multicenter, ran-domized study of cancer patients receiving chemotherapy whowere treated with EPO, compared no iron supplementation to


oral or intravenous iron therapy; the cohort receiving intrave-nous iron had superior hematopoietic responses (68%, definedas an increased Hgb to 12 g=dL or a 2g=dL increase) comparedto those receiving oral iron (36%) or no iron (25%) supplemen-tation (128). The current status of intravenous iron therapy inpatients with anemia is summarized in Table 4.

LABORATORY EVALUATION OF IRONMETABOLISM

Biochemical Parameters

The diagnosis of iron deficiency is traditionally based on acombination of parameters, including iron metabolism andhematological indices. Technical and biological issues limitthe usefulness of these assays in the clinical setting (129–131), and the value of iron, transferrin, and transferrin satura-tion is limited to uncomplicated iron deficiency. Transferrinsaturation falls below 16% only when iron stores areexhausted, in contrast to EPO therapy-induced erythropoiesis,

Table 4 Intravenous Iron Therapy for Anemia

Beneficial No benefit Investigational

I. Anemia of renalfailure, with orwithout EPO therapy(106,112,117)

I. Autologousblood donation inpatients with orwithout irondeficiency (80,81)

I. Blood loss, irondeficiency, and EPOtherapy (95,100)

II. Patients withongoing blood loss(71,73,103,123)

II. Anemia of chronicdisease and EPOtherapy (97,99,179)

III. Jehovah’s Witnesspatients with irondeficiency(125) and=orblood loss (124)

III. Perisurgical anemia,with or without EPOtherapy (96,103)

Absolute iron deficiency is defined as ferritin <200 ug=L and=or iron saturation<20%, or relative iron deficiency (ferritin <400 ug=L in dialysis patients receivingerythropoietin therapy, Ref. 49) or the presence of > 10% hypochromic erythrocytesand=or reticulocytes.Modified from Ref. 75.

454 Goodnough

in which iron saturation falls even in the presence of storageiron (132). Detection of iron-restricted erythropoiesis duringEPO therapy (132–134) therefore poses additional challenge.

Ferritin is widely used as a marker of iron storage(135,136), with a cutoff of 15mg=L indicating absent ironstores in normal individuals (137). However, one study foundthat 25% of women with no stainable bone marrow iron hadserum ferritin levels above the 15 mg=L cutoff (138). A levelof 30mg=L (139) to 40mg=L (140) for anemic patients is there-fore desirable in order to provide optimal diagnostic efficiency(positive predictive values of 92–98%, respectively), evenwithout clinical evidence of infection or inflammation.Subjects treated with EPO exhibit a rapid decrease in ferritinto levels 50–75% below baseline (141). Ferritin also decreasedrapidly even after intravenous iron was administrated innormal subjects treated with EPO (127). Under these condi-tions, ferritin most likely reflects the iron content of a smaller,more labile pool in equilibrium with both the erythropoieticcompartment and storage iron.

Many patients have underlying disorders with ‘‘inappro-priately high’’ serum ferritin levels. Ferritin levels are elevatedin conditions such as hyperthyroidism, inflammation=infec-infection, hepatocellular disease, malignancies, alcohol con-sumption, and oral contraceptives (142). Two-thirds of renaldialysis patients respond to intravenous iron therapy, withmean ferritins of 94mg=L and mean transferrin saturations of22% that are no different than the patients who are not respon-sive to intravenous iron (118). This has led to suggested guide-lines (7,7a) for renal dialysis patients, in which a ferritin of lessthan 200mg=L alone, or less than 400mg=L with a transferrinsaturation <20%, are used to determine the need for intrave-nous iron therapy; only at transferrin saturations greater than50% or ferritin levels in excess of 800 mg=L, are these patientsconsidered unlikely to benefit from iron therapy.

In patients with the anemia of cancer, up to 50% are unre-sponsive to EPO therapy (143). After two weeks of EPO ther-apy, ferritin levels greater than 400mg=L correctly predictedlack of response in 88% of the cases, while levels<400mg=Lcorrectly predicted response in 75% of cases (143). However,


several studies have failed to show a role for ferritin in eitherpredicting response to EPO therapy or in identifying functionaliron deficiency in patients with cancer-related anemia (144–147). Clinical studies to determine the level of ferritin thatwould predict response to intravenous iron in oncology patientsundergoing EPO therapy are needed.

The soluble transferrin receptor (TfR) is derived primar-ily from red cell precursor normoblasts (148) and provides anestimate of the erythroid compartment mass. Both enhancederythropoiesis and iron deficiency elevate TfR (149,150).Serum ferritin is the most sensitive and specific index of ironstatus when there are residual iron stores, whereas TfR ismost sensitive in the presence of iron-restricted erythropoi-esis (140,149).

In a study of 43 healthy, nonanemic adult women, 17(40%) had significant changes in TfR in response to oral irontherapy, indicating the presence of subclinical iron deficiency(151). Moreover, 25% of patients undergoing routine ferritintests who were also studied for TfR measurements were cate-gorized as iron deficient by TfR (>2.8 mg=L) but not by ferritin(>12 mg=L) (140). These values could represent iron-repleteindividuals with increased erythropoiesis, or iron-deficientpatients with an acute phase increase of the ferritin value.The clinical utility of the TfR may therefore be of value inthe subset of patients in whom iron deficiency is suspectedbut whose ferritin values are normal or raised (140), seencommonly in the anemia of chronic disease; a number of stu-dies (139,140,149,152) have shown TfR to be of value in differ-entiating iron deficiency anemia (where TfR is usuallyincreased) from the anemia of chronic disease (where TfR isusually normal).

The value of TfR in predicting the response to EPO ther-apy and the adequacy of iron availability is modest. Lowerbaseline or low-normal TfR levels may predict the initialresponse to EPO therapy in patients on dialysis (153). Otherstudies, however, have shown little predictive value for thisassay in patients receiving EPO, since serum TfR valuesabove normal are observed both in iron deficiency and duringerythropoietin-induced expansion of erythropoietic activity

456 Goodnough

(154). Further studies are required to delineate the clinicalusefulness of TfR measurements in patients undergoingEPO therapy.

Hematologic Parameters

Since reticulocytes are normally released from the marrow18–36 hr before their final maturation into erythrocytes, theyprovide a real-time assessment of the functional state oferythropoiesis. However, in the early phases of stimulatederythropoiesis, changes in absolute reticulocyte counts reflectthe release from the marrow of immature reticulocytes ratherthan true expansion of erythropoiesis (123,126,147). Baseline(145) and early treatment factors (155) are not clinically use-ful for predicting response to EPO therapy in anemic cancerpatients. It has been suggested that response to EPO therapycould be assessed by measuring hemoglobin and reticulocytecounts after 4 weeks of therapy: a change in hemoglobin levelgreater than 1.0 g=dL and=or a change in absolute reticulo-cyte count more than 40� 109=L could indicate that thepatient is a ‘‘responder’’ to EPO therapy (145,156).

Flow cytometric analysis of reticulocytes allows precisemeasurements of reticulocyte cell volume (MCVr), hemoglo-bin concentration (CHCMr), and hemoglobin content (CHr)(157). In normal subjects, EPO therapy induces an increasein MCVr and a decrease in CHCMr (158). Normal subjectstreated with erythropoietin with baseline serum ferritin>100mg=L have almost no production of hypochromic reticu-locytes. Detection of iron-restricted erythropoiesis takes placeat an earlier stage if reticulocyte parameters are used ratherthan red cell indices (141,159,160). CHr has been studied indialysis patients. CHr demonstrated 100% sensitivity and80% specificity and was a more accurate predictor of responseto iron therapy than serum ferritin, transferrin saturation, orpercentage hypochromic erythrocytes (161). Another studyshowed that a baseline CHr< 28 pg had 78% sensitivity and71% specificity to detect iron-restricted erythropoiesis, com-pared with 50% and 39% for traditional biochemical measures(162). In dialysis patients treated with EPO, CHr increases


during intravenous iron therapy, indicating value as an earlyindicator of iron-restricted erythropoiesis (163), even withnormal serum ferritin or transferrin saturation (164).

Measurements of total reticulocyte hemoglobin, an inte-grated index which is derived from the absolute reticulocytecount and the CHr (165) showed that reticulocyte-hemoglobinwas much higher in subjects treated with intravenous iron(127). Moreover, administration of intravenous iron alongwith EPO therapy in cardiac surgery patients abolished theproduction of hypochromic reticulocytes, and CHr remainedwithin the normal range (166). A recent study concluded thatCHr was the strongest predictor of iron deficiency in children(167), and should be considered an alternative to standardiron studies for diagnosis of iron deficiency.

Erythropoietin Assay

A classification of anemias has been proposed around the con-cept of an adequate or an inadequate erythropoietin responseto the degree of anemia, using patients with iron deficiency orchronic hemolytic anemia as reference population (168). Thecorrelation between the percentage of patients showing an‘‘inadequate’’ erythropoietin response to anemia and the per-centage of patients responding to EPO therapy (according tothe author’s criteria) can be illustrated (Fig. 3) for several dis-eases, with a range in response in myelodysplastic syndromes(169), multiple myeloma (170), and rheumatoid arthritis (171).

There are several problems with the use of erythropoietinlevels in the management of individual patients. The interpre-tation of an erythropoietin level must take into account thedegree of anemia at the time of measurement; commercialassay results do not take this into consideration, so thatclinicians need familiarity withmathematical corrections suchas observed=predicted ratios (168). A retrospective analysis ofEPO therapy in anemic cancer patients not receiving che-motherapy (172) found that pretreatment erythropoietinlevels of <200mu=mL were correlated with red cell responseto EPO therapy; however, subsequent analyses have foundthat erythropoietin levels are not predictive for response in

458 Goodnough

cancer patients receiving chemotherapy (173,174). Sincealmost all anemic cancer patients and renal dialysis patientshave erythropoietin levels that are inadequate for the degreeof anemia (168), measuring erythropoietin levels are not use-ful in these settings. Furthermore, guidelines (8) recommendthat EPO therapy be instituted only when hemoglobin levelsfall below 10 g=L, a level at which interpretation of erythro-poietin level is not valid (168). The erythropoietin assay maybe most useful as a determinant of response to therapy in cer-tain settings such as patients with myelodysplasia (169).

Figure 3 Correlation between the percentage of cases showing‘‘inadequate EPO response to anemia’’: and the percentage of casesresponding to EPO therapy (according to the author’s criteria). Thenumbers are derived directly or calculated from reported data: ARF,anemia of renal failure; RA, anemia of rheumatoid arthritis; HIV,anemia in HIV-infected patients; MM, anemia in multiple myeloma;Cancer, anemia of cancer; MDS=MMM, anemia in myelodysplasticsyndromes and myelofibrosis with myeloid metaplasia. (Reprintedwith permission from Ref. 168.)


IRON THERAPY STRATEGIES

Intravenous iron therapy has been closely scrutinized for risksand adverse events. Imferon (Iron dextran BP) was previouslyapproved for parenteral use (175). This product was associatedwith a 0.6% risk of anaphylactoid reactions and a 1.7% risk ofsevere, serum sickness-like reactions characterized by fever,arthragias, and myalgias (176). An incidence of delayed reac-tions of up to 30% and severe reactions of 5.3% were subse-quently described, attributed to changes in manufacturingprocesses (177); this product was withdrawn from use.

InFed (Iron dextran USP, Schein Pharm Corp, FlorhamPark, NJ) is currently approved for parenteral (intramuscularor intravenous) use in the United States. InFed administeredintravenously during dialysis is associated with significantadverse reactions in 4.7% of patients, of which 0.7% are ser-ious or life threatening, and another 1.7% are characterizedas anaphylactoid (178). The prevalence of these reactions doesnot differ among patients receiving low-dose (100mg) orhigher-dose (250–500mg) infusions (179). A recent reviewreported 196 allergic=anaphylaxis cases with use of irondextran in the United States between 1976 and 1996, of which31 (15.8%) were fatal (180).

Safety aspects of parenteral iron dextran, ferric gluco-nate, and iron saccharate have been scrutinized (181–183).Iron saccharate is available in Europe but not in the UnitedStates. Ferric gluconate has been available in Europe formore than 20 years and was approved for use in the UnitedStates in 1999 for intravenous administration (Ferrlecit,Schein Pharm Corp, Florham Park, NJ) in renal dialysispatients. Dosage is limited to 125mg infused over 1 hr at eachadministration. The rate of allergic reactions (3.3 episodes permillion doses) appears lower than iron dextran (8.7 episodesper million doses) and the safety profile is substantiallybetter; among 74 severe adverse events reported from 1976to 1996, there were no deaths (184).

Adverse events associated with ferric gluconate includehypotension, rash, chest, or abdominal pain, with an inci-dence of 1.3% for serious reactions (185,186). Intravenous

460 Goodnough

iron therapy can cause a clinical syndrome (nausea, facialreddening, and hypotension) which may be attributed acuteiron toxicity due to oversaturation (>100%) of transferring(187) or due to nonspecific drug toxicity (188). The increasederythropoietic effect (4.5–5.5 times basal) of intravenous irondextran (with an estimated half-life of 60hr) is transient andlasts 7–10 days, after which the remaining iron is sequesteredin the RES and erythropoiesis returns to basal rates (71). Ironmeasurements and intravenous iron therapy are thereforemore optimal at two-week intervals.

A dose–response relationship between EPO therapy anderythropoiesis that is affected favorably by intravenous ironhas important implications for EPO dosage and cost (189). Cur-rent total recommended EPO dosage for patients scheduled forelective surgery (190) range from 1800u=kg (191) to 4200u=kg(192), which for a 70-kg patientwould cost $1300 to $3000 (193).As in renal dialysis patients (111), intravenous ironmay reduceEPO dosage in patients by improving iron-restricted erythro-poiesis (128), even in patients with demonstrable iron stores;additional clinical trials to study this are desirable (194,195).

CURRENT ISSUES IN ERYTHROPOIETINTHERAPY

Safety

Thrombotic events were described in an initial uncontrolledtrial of epoietin alfa therapy in patients undergoing dialysis(196). The observation of thrombotic events with epoietinalfa administration in this setting required a subsequentrandomized placebo-controlled trial. Diastolic blood pres-sures showed a mild but significant elevation in epoietinalfa-treated patients maintained at ‘‘higher’’ levels of hemo-globin (115–130 g=L) compared with ‘‘lower’’ (95–110g=L)levels or placebo-treated patients; venous access clottingwas similarly increased (197). Studies of epoietin alfa ther-apy in the setting of uremia suggested that effects on plate-let number, platelet aggregation, blood coagulation, and=orfibrinolysis could influence the risk of thrombosis duringepoietin alfa therapy in patients who are uremic (198–200).


The shortening of the bleeding time in these patients treatedwith epoietin alfa, however, is related to increasedhematocrit (201,202); any thrombotic effect may, in part,be related to acute increases in hematocrit affecting bloodrheology in patients at risk (203).

Subsequently, a randomized trial was conducted in hemo-dialysis patients with clinical evidence of CHF or ischemicheart disease to study the risks and benefits of normalizinghematocrit (to achieve andmaintain a hematocrit of 42%) com-pared to maintenance of hematocrit at 30% (204). The primaryendpoint was length of time to death or a first nonfatal myo-cardial infarction. The study was halted after 29 months with183 deaths and 19 first nonfatal myocardial infarctions in thenormal hematocrit cohort compared to 150 and 14, respec-tively, in the low hematocrit group; while the differences werenot statistically significant, they were sufficient to precludeany possibility that the study would reveal a benefit for thenormal hematocrit cohort. Of note, the mortality rates decre-ased with increasing hematocrit in both groups.

Thrombotic events have not been associated with epoie-tin alfa therapy in carefully controlled trials of patients sched-uled for surgery. The safety of epoetin alfa therapy in patientsundergoing noncardiac surgery has been demonstrated by theequal distribution of concomitant adverse events betweenpatients treated with epoietin alfa or placebo in over 1000surgical patients participating in clinical trials. The overallprevalence of thrombotic events in 10 (2.8%) of 365 evaluablepatients in three clinical trials (205–207) undergoingpreoperative autologous blood donation, with or withoutepoietin alfa therapy, is similar to rates of thrombotic compli-cations reported in patients undergoing orthopedic surgery.The occurrence of myocardial infarction in the setting ofABD has also been described in patients undergoing radicalprostatectomy (208). Careful studies of hemostasis, fibrinoly-sis, and rheology in autologous blood donors have failed toidentify (pro) thrombotic changes (209,210). In view of thethrombotic events reported during the preoperative blooddonation interval in both (placebo and epoietin alfa) patientcohorts, volume replacement in patients undergoing aggres-

462 Goodnough

sive (twice weekly) phlebotomy in any patient known to havecardiovascular risks seems prudent.

An unresolved question is the safety of epoietin alfatherapy in patients with cardiac or vascular disease. In aEuropean trial (211), the investigators found no differencesin mortality, thrombotic events, or serious adverse events in76 patients undergoing cardiac surgery between the epoietinbeta-treated and placebo cohorts, nor any differences inhemostatic parameters in their patients during the 14-daypreoperative interval in which increased hemotocrits (from42� 3% to 48� 3%) were demonstrated (210). In fact, theinvestigators were able to demonstrate that epoietin beta-treated patients had an improved extractable oxygen perio-peratively, when compared with placebo-treated patients.This was also associated with a lower incidence of lactic acido-sis in the epoietin beta-treated patients (212). A U.S. study(213) also observed no differences in adverse events betweenepoietin alfa- and placebo-treated patients undergoing coron-ary artery bypass surgery, and concluded that epoietin alfatherapy was well tolerated. However, in this study, in unevendistribution of these events between the placebo and theepoietin alfa-treated groups could not be ruled out to anydegree of certainty. For example, even if the true mortalityrate was 0% in the placebo group and 6% in the combinedepoetin alfa groups, the probability is only 23% (power of0.229) that the resulting data would produce a statisticallysignificant p value of <0.05 (214).

What is the current role of erythropoietin therapy inpatients with cardiac or vascular disease, particularly forthe United States, in which cardiac and vascular surgeriesare excluded as an indication for its use perisurgically? Thisapproach remains a valuable tool for patients with specialrequirements, such as Jehovah’s Witness patients for whomblood transfusion is not an option (194). Until additionalsafety data are forthcoming, the off-label use of erythropoietintherapy in patients with cardiac or vascular disease can onlybe recommended in the context of clinical trials. Emergingdata on the use of erythropoietin therapy in noncardiac proce-dures, such as elderly men undergoing radical prostatectomy


(215), may provide additional evidence that perioperative ele-vations of hematocrit, even in patients at risk for ischemicheart disease, are well tolerated. Potential concerns that a‘‘too rapid’’ rise in hematocrit may be harmful are notsupported by experience in two clinical settings: patientsreceiving blood transfusions who have immediate and substan-tial increases in hematocrit (which is actually the desired effect),and iron-deficient patients receiving total dose iron infusionswho were reported to have 2g=dL hemoglobin increases withinone-week time interval, with no adverse consequences (216).

Of more recent concern is the increasing identification ofpatients since 1998 who have developed pure red cell aplasia(PRCA) while undergoing erythropoietin therapy (comparedto a total of three cases prior to 1998) (217). This complicationhas been associated with the demonstration of neutralizingantibodies to erythropoietin (218,219). One hundred eightyeight cases of PRCA were documented from 1998 through200. Data submitted to the Food and Drug Administrationsuggest important differences among brands of epoietin;(175) of the reported cases involve patients receiving Eprex,which is manufactured and distributed to patients outsidethe United States. This product has undergone a significantmanufacturing change since 1998 with the removal of humanserum albumin as a stabilizer. PRCA has been reportedpredominantly among dialysis patients who have taken thedrug subcutaneously but not intravenously. These observa-tions have led to speculation that the immunogenicity of therecombinant product has been enhanced through possiblecombinations of changes in the manufacturing, handling,and=or administration of the recombinant product. For thisreason, the route of Eprex administration in patients withchronic renal failure has now been recommended to be intra-venous. After addressing issues related to storage, handling,and administration of EPREX, an 83% reduction in incidenceof PRCA has been observed (217).

Central Nervous System Effects

An intriguing demonstration of an effect of erythropoietinthat is in addition to its central role in erythropoiesis, erythro-

464 Goodnough

poietin crosses the blood–brain barrier and exerts a neuro-protective effect in animal models of experimental braininjury (220). Similar to its regulation in the peripheral circu-lation, erythropoietin within the central nervous system isinducible by hypoxia. Systemically administered epo therapyhas been shown to function as a neuroprotective agent in ani-mal models of focal brain ischemia, concussive brain injury,experimental autoimmune encephalitis (eae), and kainate-induced seizures (220). The manner in which erythropoietinserves as a neuroprotectant is unclear. One hypothesis is thaterythropoietin could rescue cells from death through modula-tion of aptoptosis, a role well defined in erythropoiesis andsince extended to neuronal-like cells in vitro. Important clin-ical implications are the potential benefit of improving cogni-tion in the elderly and in protecting cognition in patientsreceiving chemotherapy, who have been demonstrated tohave impaired cognitive dysfunction (221).

Darbopoietin Alfa

Another important development is the development andcharacterization of a novel erythropoiesis stimulating protein(darbopoietin alfa). Darbopoietin alfa is a genetically engi-neered molecule that is biochemically distinct from recom-binant human erythropoietin (epoietin alfa), containingadditional carbohydrate and sialyc acid moieties, which pro-long its serum half-life and thus increase its in vivo biologicactivity (222). In pharmacokinetic studies in patients withrenal disease, darbopoietin alfa was shown to have a threefoldlonger half-life than epoietin alfa after intravenous adminis-tration (25.3 vs. 8.5 hr, respectively). Subcutaneous adminis-tration extended the half-life of darbopoietin alfa to 48.8 hr.Darbopoietin alfa is now approved in many countries world-wide for treatment of anemia associated with chronic renalfailure.

Darbopoietin alfa has also undergone clinical trials inpatients with cancer (223–225). In a combined analysis trial,the feasibility of reduced frequency of administration withdarbopoietin alfa compared to epoietin alfa was demon-


strated, comparing weekly darbopoietin alfa vs. thrice-weeklyepoietin alfa (according to current labeling indications); andcomparing every other week and even every third week dar-bopoietin alfa vs. weekly epoietin alfa (according to currentprevailing practice in oncology). This trial (223) also demon-strated a dose-dependent relationship between darbopoietinalfa and multiple measures of efficacy, including proportionof patients responding (defined as an increase in Hgb> 2 g=L)to darbopoietin alfa. In a worldwide (non-USA) trial of dar-bopoietin alfa in cancer patients not receiving chemotherapy,up to 83% of patients responded to once-weekly subcutaneousadministration of darbopoietin alfa NESP (224). Darbopoietinalfa was well tolerated in all trials, and there has been no evi-dence of antibody formation in over 126,000 patients with68,000 patient-years exposure as of 12=02. Darbopoietin alfais now approved for treatment of anemia in patients with non-myeloid malignancies receiving chemotherapy in the UnitedStates, European Union, Australia, Switzerland, and Israel.

CONCLUSION

The development of new laboratory methods to assess iron-restricted erythropoiesis, along with clinical trials of patientsundergoing EPO therapy, have furthered our understandingof the relationship between erythropoiesis and iron metabo-lism. Reticulocyte parameters hold promise in the evaluationof iron-restricted erythropoiesis, but more studies are neededin order to define their role in patients undergoing ironand=or EPO therapy. The erythropoietin assay and transfer-rin receptor assay are valuable tools for clinical research,but their roles in routine clinical practice remain undefined.The availability of a safer intravenous iron preparation allowsnot only its routine use in bloodless medicine, but also repre-sents an opportunity to study its value in patients with ane-mia of chronic disease undergoing EPO therapy. Given thelow prevalence but potential side effects, the future use ofintravenous iron needs to be defined by controlled clinicaltrials.

466 Goodnough

REFERENCES

1. Goodnough LT, Dubois R, Nissenson A. Anemia: not just aninnocent bystander? Arch Int Med. 2003; 163:1400–1404;Correction Arch Int Med 2003; 163:1820.

2. Goodnough LT, Shander A, Brecher ME. Transfusion medi-cine: looking to the future. Lancet 2003; 361:161–169.

3. Goodnough LT, Brecher ME, Kanter MH, Aubuchon JP.Transfusion medicine, part I: blood transfusion. N Engl JMed 1999; 340:438–444.

4. Welch HG, Meehan KR, Goodnough LT. Prudent strategiesfor red blood cell transfusion. Ann Int Med 1992; 116:393–402.

5. American College of Physicians. Practice strategies for electivered blood cell transfusion. Ann Int Med 1992; 116:403–406.

6. Expert Working Group. Guidelines for red blood cell andplasma transfusions for adults and children. Can Med AssocJ 1997; 156(suppl 11):S1–S25.

7. National Kidney Foundation. K=DOQI. Clinical practiceguidelines for the treatment of anemia of chronic kidneydisease. Am J Kidney Dis 1997; 30:S192–S237.

7a. National Kidney Foundation. K=DOQI. Clinical practiceguidelines for the treatment of anemia of chronic kidneydisease. Am J Kidney Dis 2001; 37:S182–S238.

8. Rizzo JD, Lichtin AE, Woolf SH, et al. Use of epoietin inpatients with cancer: evidence-based clinical practice guide-lines of the American based clinical practice guidelines ofthe American Society of Clinical Oncology and the AmericanSociety of Hematology. J Clin Oncol 2002; 20:4083–4107.

9. Consensus conference: perioperative red blood transfusion.JAMA 1988; 260:2700–2703.

10. Practice guidelines for blood component therapy: a report bythe American Society of Anesthesiologists Task Force on BloodComponent Therapy. Anesthesiology 1996; 84:732–747.e 63.

11. Siciliano M, Tomasella D, Milani A, Ricerca BM, Storti S,Rossi L. Reduced serum levels of immunoreactive erythro-


poietin in patients with cirrhosis and chronic anemia.Hepatology 1995; 22:1132–1135.

12. Baer AN, Dessypris EN, Krantz SB. The pathogenesis ofanemia in rheumatoid arthritis: a clinical and laboratoryanalysis. Sem Arthritis Rheum 1990; 19:209–223.

13. Zon LI, Arkin C, Groopman JE. Haematologic manifestationsof the human immune deficiency virus (HIV). Sem Hematol1988; 25:208–212.

14. Groupman JE, Itri LM. Chemotherapy-induced anemia inadults: incidence and treatment. J Nat Cancer Inst 1999;91:1616–1634.

15. Ma J, Ebben J, Xia H, et al. Hct level and associated morta-lityin hemodialysis patients. J Am Soc Nephrol 1999; 10:610–619.

16. Collins A, Xia H, Ebben J, et al. Change in Hct and risk ofmortality. J Am Soc Nephrol 1998; 9:204A.

17. Collins AJ, Li S, St. Peter W, et al. Death, hospitalization,and economic associations among incident hemodialysispatients with hematocrit values of 36 to 39%. J Am SocNephrol 2001; 12:2465–2473.

18. FurulandH, Linde T, Ahlmen J, Christensson A, Stromborn Y,Danielson BG. A randomized controlled trial of haemoglobinnormalization with epoietin alfa in pre-dialysis and dialysispatients. Nephrol Dial Transplant 2003; 18:353–361.

19. Maeda K, Tanaka Y, Tsukano Y, Fujita K, Tanaka C,Takeuchi K, et al. Multivariate analysis using a linear discri-minant function for predicting the prognosis of congestiveheart failure. Jap Circ J 1982; 46:137–142.

20. Carson JL, Duff A, Poses RM, et al. Effect of anemia andcardiovascular disease on surgical mortality and morbidity.Lancet 1996; 348:1055–1060.

21. Hebert PC, Wells G, Sweeddale M, et al. Does transfusionpractice affect mortality in critically ill patients? Am J RespirCrit Care Med 1997; 155:1618–1623.

22. Carson JL, et al. Mortality and morbidity in patients withvery low postoperative Hb levels who decline blood transfu-sion. Transfusion 2002; 42:812.

468 Goodnough

23. Horwich TB, et al. Anemia is associated with worse symp-toms, greater impairment in functional capacity and a signif-icant increase in mortality in patients with advanced heartfailure. J Am Coll Card 2002; 39:1780.

24. KosiborodM, SmithGL, RadfordMJ, Foody JM,KrumholzHM.The prognostic importance of anemia in patients with heartfailure. Am J Med 2003; 114:112–119.

25. Ezekowitz JA, McAlister FA, Armstrong PW. Anemia is com-mon in heart failure and is associated with poor outcomes.Insights from a cohort of 12,065 patients with new-onsetheart failure. Circulation 2003; 107:223–225.

26. SilverbergOS,WexlerD, ShepsD, BlumM,KerenG, Baruch R,et al. The effect of correction of mild anemia in severe, resistantcongestive heart failure using subcutaneous erythropoietin andintravenous iron: a randomized controlled study. J Am CollCardiol 2001; 37:1776–1779.

27. Wu WC, Rathmore SS, Wang Y, Radford MJ, Krumholz HM.Effectiveness of blood transfusion in elderly patientswith acutemyocardial infarction. N Eng J Med 2001; 345:1230–1236.

28. Duke M, Abelman WH. The hemodynamic response tochronic anemia. Circulation 1969; 39:503–515.

29. Ogawa T, Spina RJ, Martin WH III, et al. Effects of aging, sexand physical training on cardiovascular responses to exercise.Circulation 1992; 86:494–503.

30. Levy PS, Song-Jung K, Eckel PK, Chavez R, Ismail EF,Gould SA, Salem MR, Crystal GJ. Limit to cardiac compensa-tion during acute isovolemic hemodilution: influence of coron-ary stenosis. Am J Physiol 1993; 265(Heart Circ Physiol 34):H340–H349.

31. Maggioni AA, Meseri A, Fresco C, Franzosi MG, Mauri F,Santoro E, Tognoni G. The Investigators of the GruppoItaliano por lo Studio della Sopravvivenza nell’InfartooMiocardico (GISSI-2). Age-related increase in mortalityamong patients with first myocardial infarctions treated withthrombolysis. N Engl J Med 1993; 329:1442–1448.

32. Goodnough LT, Bach RG. Anemia, transfusion, and mortality.N Engl J Med 2001; 345:1272–1273.


33. Besarab A, Bolton WK, Browne JK, Egrie JC, Nissenson AR,Okamoto DM, et al. The effects of normal as compared withlow hematocrit values in patients with cardiac disease whoare receiving hemodialysis and epoetin. N Engl J Med 1998;339:584–590.

34. Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC,Barre PE. The impact of anemia on cardiomyopathy, morbid-ity, and mortality in end stage renal disease. Am J KidneyDis 1996; 28:53–61.

35. Levin A, Thompson CR, Ethier J, Carlisle EJF, Tobe S,Mendelssohn D, et al. Left ventricular mass index increasein early renal disease: impact of decline in hemoglobin. AmJ Kidney Dis 1999; 34:125–134.

36. Levin A, Singer J, Thompson CR, Ross H, Lewis M. Prevalentleft ventricular hypertrophy in the predialysis population:identifying opportunities for intervention. Am J Kidney Dis1996; 27:347–354.

37. Silverberg JS, Rahal DP, Patton R, Sniderman AD. Roleof anemia in the pathogenesis of left ventricular hypertrophyin end-stage renal disease. Am J Cardiol 1989; 64:222–224.

38. Rigatto C, Foley R, Jeffrey J, Negrijn C, Tribula C, Parfrey P.Electrocardiographic left ventricular hypertrophy in renaltransplant recipients: prognostic valve and impact of bloodpressure and anemia. J Am Soc Nephol 2003; 14:462–468.

39. Canella G, LaCanna G, Sandrini M, Gaggiotti M, Nordio G,Movilli E, et al. Nephrol Dial Transplant 1991; 6:31–37.

40. Collins A, Ma JZ, Ebben J, et al. Impact of hematocrit onmorbidity and mortality. Sem Nephrol 2000; 20:345–349.

41. Pickett JL, Theberge DC, Brown WS, et al. Normalizinghematocrit in dialysis patients improves brain function. AmJ Kidney Dis 1999; 33:1122–1130.

42. Valderrabano F, Jofre R, Lopez-Gomez JM. Quality of life inend-stage renal disease patients. Am J Kidney Dis 2001;38:443–464.

43. Deicher R, Horl WH. Anemia is a risk factor for the progres-sion of chronic kidney disease. Curr Opin Nephol Hypertens2003; 12:139–142.

470 Goodnough

44. Low I, Grutzmacher P, Bergmann M, Schoeppe W. Echocar-diographic findings in patients in maintenance hemodialysissubstituted with recombinant human erythropoietin. ClinNephrol 1989; l31:26–30.

45. Low-Friedrich I, Grutzmacher P, Marz W, Bergmann M,Schoeppe W. Therapy with recombinant human erythropoie-tin reduces cardiac size and improves heart function inchronic hemodialysis patients. Am J Nephrol 1991; 11:54–60.

46. Goldberg N, Lundin AP, Delano B, Friedman EA, Stein RA.Changes in left ventricular size, wall thickness, and functionin anemic patients treated with recombinant human erythro-poietin. Am Heart J 1992; 124:424–427.

47. Foley RN, Parfrey PS, Morgan J, et al. A randomized con-trolled trial of complete vs. partial correction of anemia inhemodialysis patients with asymptomatic concentric LVhypertrophy or LV dilatation [abstr]. J Am Soc Znephrol1998; 9:208.

48. Linde T, Wikstrom B, Andersson LG, Danielson BG. Renalanaemia treatment with recombinant human erythropoietinincreases cardiac output in patients with ischemic heartdisease. Scand J Urol Nephrol 1996; 30:115–120.

49. Packer M, Cohn JN. Consensus recommendations for themanagement of chronic heart failure. Am J Cardiol 1999;83:1–38A.

50. Harrison LB, Shasha D, White C, Ramdeen B. Radiotherapy-associated anemia: the scope of the problem. Oncologist 2000;5(suppl 2):1–7.

51. Bush RS. The significance of anemia in clinical radiationtherapy. Int J Rad Oncol Biol Phys 1986; 12:2047–2050.

52. Grogan M, Thomas GM, Melamed I, Wong FLW, Pearcey RG,Joseph PK, et al. The importance of hemoglobin levels duringradiotherapy for carcinoma of the cervix. Cancer 1999;86:1528–1536.

53. Lee JJ, Scott C, Komaki R, Fossella FV. Concurrent chemor-adiation therapy with oral etoposide and cisplatinum forlocally advanced inoperable non-small cell lung cancer:RTOG Protocol 91–06. Clin Oncol 1996; 8:167–175.


54. Fein DA, LeeWR, Hanlon L, Ridge JA, Langer CJ, CurranWJ,Coia LR. Pretreatment hemoglobin level influences localcontrol and survival of T1–T2 squamous cell carcinomas ofthe glottic larynx. J Clin Oncol 1995; 13(8):2077–2083.

55. Lee WR, Berkey B, Marcial V, Fu KK, Cooper S, Vikrum B,Coia LR, Rotman M, Ortiz H. Anemia is associated withdecreased survival and increased locoregional failure inpatients with locally advanced head and neck carcinoma: asecondary analysis of RTOG 85–27. Int J Radiat Oncol BiolPhys 1998; 42(5):1069–1075.

56. Moreno F, Sanz-Guajardo D, Lopez-Gomez JM, Jofre R,Valderrabano F. Increasing the hematocrit has a beneficialeffect on quality of life and is safe in selected hemodialysispatients. Spanish Cooperative Renal Patients Quality of LifeStudy Group of the Spanish Society of Nephrology. J Am SocNephrol 2000; 11:335–342.

57. Glaspy J, Bukowski R, Steinberg D, Taylor C, Tchekmedyian S,Vadhan-Raj S. Impact of therapy with epoetin alfa onclinical outcomes in patients with nonmyeloid malignanciesduring cancer chemotherapy in community oncology practice.J Clin Oncol 1997; 15:1218–1234.

58. Demetri GD, Kris M, Wade J, Degos L, Cela D. Quality-of-lifebenefit in chemotherapy patients treated with epoetin alfa isindependent of disease response or tumor type: results from aprospective community oncology study. J Clin Oncol 1998;10:3412–3425.

59. Cella D, Zagari MJ, Vandoros C, Gagnon DD, Hurtz HJ,Nortier JW. Epoietin alfa treatment results in clinically sig-nificant improvements in quality of life in anemic cancerpatients when referenced to the general population. J ClinOncol 2003; 21:366–373.

60. Oshaughnessy JA. Effects of epoietin alfa on cognitive func-tion, mood, asthenia, and quality of life in women with breastcancer undergoing adjuvant chemotherapy. Clin BreastCancer 2002; 3(suppl 3):S116–S120.

61. Cleveland CS, Demetri GD, Glaspy J, Cella DF, Portenoy K,Cremieux PY, Itri LM. Identifying hemoglobin level for opti-

472 Goodnough

mal quality of life: results on an incremental analysis. ProcAm Soc Clin Oncol 1999; 18:574a.

62. Abels R. Erythropoietin for anemia in cancer patients. Eur JCancer 1993; 29(suppl 2):52–58.

63. Littlewood TJ, Bajetta E, Nortier JWR, Vercammen E,Rapoport B. Effects of epoetin alfa on hematologic parametersand quality of life in cancer patients receiving nonplatinumchemotherapy: results of a randomized, double-blind,placebo-controlled trial. J Clin Oncol 2001; 19:2865–2874.

64. Turner R, Anglin P, Burkes R, et al. Epoietin alfa in cancerpatients: evidence-based guidelines. J Pain Sympt Manage2001; 22:954–965.

65. Nissenson AR, Collins AJ, Hurley J, Peterson H, Pereira BJG,Steinberg EP. Opportunities for improving the care ofpatients with chronic renal insufficiency: current practice pat-terns. J Am Soc Nephrol 2001; 12:1713–1720.

66. Parfrey PS, Griffiths SM, Harnett JD, Taylor R, King A,Hand J. Outcome of congestive heart failure, dilated cardiomyo-pathy, hypertrophic hyperkinetic disease, and ischemic heartdisease in dialysis patients. Am J Nephrol 1990; 10:213–221.

67. Cody J, Daly C, Campbell M, et al. Frequency of administra-tion of recombinant human erythropoietin for anaemia of endstage renal disease in dialysis patients (Cochrane Review).Cochrane Database Syst Rev 2002; 4:CD003895.

68. Flanigan MJ, Rocco MV, Prowant B, Frederick PR, Franken-field DL. Clinical performance measures: the changing statusof peritoneal dialysis. Kidney Int 2001; 60:21377–21384.

69. Finch CA. Erythropoiesis, erythropoietin, and iron. Blood1982; 60:1241–1246.

70. Crosby WH. The metabolism of hemoglobin and bile pigmentin hemolytic disease. Am J Med 1955; 18:112–122.

71. Hillman RS, Henderson PA. Control of marrow production bythe level of iron supply. J Clin Invest 1969; 48:454–460.

72. Crosby WH. Treatment of haemochromatosis by energeticphlebotomy. One patient’s response to the letting of 55 litersof blood in 11 months. Br J Haemat 1958; 4:82–88.


73. Coleman PH, Stevens AR, Dodge HT, Finch CA. Rate ofregeneration after blood loss. Arch Int Med 1953; 92:341–348.

74. Finch CA. Perspectives in iron metabolism. N Engl J Med1982; 25:1520–1528.


76. Menitove J, ed. Standards for Blood Banks and TransfusionServices. 19th ed. Bethesda, MD: American Association ofBlood Banks, 1999:76.

77. Goodnough LT, Brittenham G. Limitations of the erythro-poietic response to serial phlebotomy: implications for autolo-gous blood predeposit programs. J Lab Clin Med 1990;115:28–35.

78. Kickler TS, Spivak JL. Effect of repeated whole blooddonations on serum immunoreactive erythropoietin levels inautologous donors. JAMA 1988; 260:65–67.

79. Kasper SM, Gerlich W, Buzello W. Preoperative red cellproduction in patients undergoing weekly autologous blooddonation. Transfusion 1997; 37:1058–1062.

80. Kasper SM, Lazansky H, Stark C, Klimek M, Laubinger R,Borner Y. Efficacy of oral iron supplementation is notenhanced by additional intravenous iron during autologousblood donation. Transfusion 1998; 38:764–770.

81. Weisbach V, Skoda P, Rippel R, Layer G, Glaser A, Zingsem J,Zimmerman R, Esckstein R. Oral or intravenous iron asan adjunct to autologous blood donation in elective surgery:a randomized, controlled study. Transfusion 1999; 39:465–472.

82. Goodnough LT, Rudnick S, Price TH, Ballas SK, Collins M,Crowley JP, Kosmin M, Kruskall MS, Lenes BA, Menitove JE,Silbertein CE, Smith KJ, Wallas Ch, Abels R, Von Tress M.Increased preoperative collection of autologous blood withrecombinant human erythropoietin therapy. N Engl J Med1989; 321:1163–1168.

83. Goodnough LT, Price TH, Rudnick S, Soegiarso RW. Preo-perative red blood cell production in patients undergoing

474 Goodnough

aggressive autologous blood phlebotomy with and withouterythropoietin therapy. Transfusion 1992; 32:441–445.

84. Goodnough LT, Price TH, Friedman KD, Johnston M,Ciavarella D, Khan N, Sacher R, Vogler WR, Wissel M,Abels RI. A phase III trial of recombinant human erythro-poietin therapy in non-anemic orthopedic patients subjectedto aggressive autologous blood phlebotomy: dose, response,toxicity, efficacy. Transfusion 1994; 34:66–71.

85. Brecher MA, Monk TG, Goodnough LT. A standardizedmethod for calculating blood loss. Transfusion 1997; 37:1070–1074.

86. Goodnough LT, Bravo J, Hsueh Y, Keating L, BrittenhamGM. Red blood cell volume in autologous and homologousblood units. Implications for risk benefit assessment for auto-logous blood ‘‘crossover’’ and directed blood transfusion.Transfusion 1989; 29:821–822.

87. GoodnoughLT, Price TH, ParvinCA, FriedmanKD,VoglerWR,Khan N, Sacher R, Johnston M, Wissel M, Ciavarella D.Erythropoietin response to anaemia is not altered by surgeryor recombinant human EPO therapy. Br J Haem 1994; 87:695–699.

88. Adamson JW. The erythropoietin=hematocrit relationship innormal and polycythemic man: implications of marrow regu-lation. Blood 1968; 32:597–609.

89. Sans T, Bofill C, Joven J, Cliville X, Simo JM, Llobet X, Pero A,Galbany J. Effectiveness of very low doses of subcutaneousrecombinant human erythropoietin in facilitating autologousblood donation before orthopedic surgery. Transfusion 1996;36:822–826.

90. Mercuriali F, Inghilleri, Biffi E, Colotti MT, Vinci A, Oriani G.Epoietin alfa in low hematocrit patients to facilitate autolo-gous blood donation in total hip replacement: a randomized,double-blind, placebo-controlled, dose-ranging study. ActaHaematol 1998; 100:69–76.

91. Brugnara C, Chambers LA, Malynn E, Goldberg MA, Krus-kall MS. Red blood cell regeneration induced by subcutaneousrecombinant erythropoietin: iron-deficient erythropoiesis iniron-replete subjects. Blood 1993; 81:956–964.


92. Goodnough LT, Verbrugge D, Marcus RE, Goldberg V.The effect of patient size and dose of recombinant humanerythropoietin therapy on red blood cell expansion. J Am CollPhys 1994; 179:171–176.

93. Goodnough LT, Price TH, Parvin CA. The endogenous ery-thropoietin response and the erythropoietic response to bloodloss anemia: the effects of age and gender. J Lab Clin Med1995; 126:57–64.

94. Mercuriali F: Zanella A, Barosi G, Inghilleri G, Biffi E, VinciA, Colotti MT. Use of erythropoietin to increase the volume ofautologous blood donated by orthopedic patients. Transfusion1993; 33:55–60.

95. Price TH,GoodnoughLT,VoglerWR, SacherRA,HellmanRM,JohnstonMFM,BolgianoDC,AbelsRI.Theeffectof recombinanthuman erythropoietin on the efficacy of autologous blood dona-tioninpatientswith lowhematocrits:amulticenter, randomized,double-blind, controlled trial. Transfusion 1996; 36:29–36.

96. Olijhoek G, Megens JGN, Musto P, et al. Role of oral versusIV iron supplementation in the erythroepoietic response torHuEPO: a randomized, placebo-controlled trial. Transfusion2001; 41:957–963.

97. Goodnough LT, Marcus RE. The erythropoietic response toerythropoietin therapy in patients with rheumatoid arthritis.J Lab Clin Med 1997; 130:381–386.

98. Mercuriali F, Inghilleri G, Biffi E, Colotti MT, Vinci A,Sinigaglia L, Gualtieri G. Comparison between intravenousand subcutaneous recombinant human erythropoietin (epoie-tin alfa) administration in presurgical autologous blood dona-tion in anemic rheumatoid arthritis patients undergoingmajor orthopedic surgey. Vox Sang 1997; 72:93–100.

99. Schreiber S, Howaldt S, Schnoor M, Nikolauss, Bavditz J,Gasche G, Lochs H, Raedler A. Recombinant erythropoietinfor the treatment of anemia in inflammatory bowel disease.N Engl J Med. 1996; 334:619–623.

100. Goodnough LT,Marcus RE. Erythropoiesis in iron-depleted andiron-replete patients stimulated with erythropoietin: the rele-vance of storage iron. Vox Sang 1998; 75:128–133.

476 Goodnough

101. Biesma DH, Kraaijenhagen RJ, Poortman J, Marx JJM,Van De Wiel A. The effect of oral iron supplementation onerythropoiesis in autologous blood donors. Transfusion1992; 32:162–165.

102. Skikne BS, Cook JD. Effect of enhanced erythropoiesis oniron absorption. J Lab Clin Med 1992; 120:746–751.

103. Heiss MM, Tarabichi A, Delanoff C, Allgayer H, Jauch KW,Hernandez-Richter T, Mempel W, Beck KG, Schildberg FW,Messmer K. Perisurgical erythropoietin application in ane-mic patients with colorectal cancer: a double-blind rando-mized study. Surgery 1996; 119:52–57.

104. Biesma DH, Van De Wiel A, Beguin Y, Kraaijenhagen RJ,Marx JJM. Erythropoietic activity and iron metabolism inautologous blood donors during recombinant human erythro-poietin theapy. Eur J Clin Invest 1994; 24:426–432.

105. Fitzsimons EJ, Sturrock RD. The chronic anaemia of rheu-matoid arthritis: iron banking or blocking? Lancet 2002; 360:1713–1714.

106. Papadaki HA, Kritikos HD, Gemetzi C, et al. Bone marrowprogenitor cell reserve and function and stromal cell functionare defective in rheumatooid arthritis: evidence for a tumornecrosis factoralpha-mediatedeffect.Blood2002;99:1610–1619.

107. Papadaki HA, Kritikos HD, Valatas V, Boumpas DT,Eliopoulos GD. Anemia of chronic disease in rheumatoidarthritis associated with increased aptoptosis of bone marrowerythroid cells: improvement following anti-tumor factor-aantibody therapy. Blood 2002; 100:474–482.

108. Fitzsimmons EJ, Houston T, Munro R, Sturrock RD,Speekenbrink ABJ, Brock JH. Erythroblast iron metabolismand serum soluble transferrin receptor values in the anemiaof chronic disease. Arthritis Rheum 2002; 47:166–171.

109. Peeters HRM, Jongen-Lavrencic M, Vreugdenhil G, SwaakAJG. Effect of recombinant human erythropoietin on anemiaand disease activity in patients on anemia and diseaseactivity in patients with rheumatoid arthritis and anemiaof chronic disease: a randomized placebo controlled doubleblind 52 weeks clinical trial. Ann Rheum Dis 1996; 55:735–744.


110. Eschbach J, Egrie J, Downing M, Browne JK, Adamson JW.Correction of anemia of end-stage renal disease with recombi-nant human erythropoietin. N Engl J Med 1987; 316:73–78.

111. Fishbane S, Frei GL, Maesaka J. Reduction in recombinanthuman erythropoietin doses by the use of chronic intravenousiron supplementation. Am J Kidney Dis 1995; 26:41.

112. Means RT. Commentary: an anemia of chronic disease, afterall? J Invest Med 1999; 47:203–207.

113. Tarng DC, Hyang TP, Chen TW, Yang WC. Erythropoietinhyporesponsiveness: from iron deficiency to iron overload.Kidney Int 1999; 55:S107–S118.

114. Skikne BS, Cook JD. Effect of enhanced erythropoiesis oniron absorption. J Lab Clin Med 1992; 120:746–751.

115. Wingard RL, Parker RA, Ismail N, Hakim RM. Efficiency oforal iron therapy in patients receiving recombinant humanerythropoietin. Am J Kidney Dis 1995; 25:433–439.

116. Adamson JW, Eschbach JW. Erythropoietin for end-stagerenal disease. N Engl J Med 1998; 339:625–627.

117. Silverberg DS, Iaina A, Peer G, Kaplan E, Levi BA, Frank N,Steinbruch S, Blum M. Intravenous iron supplementation forthe treatment of the anema of moderate to severe chronicrenal failure patients not receiving dialysis. Am J KidneyDis 1996; 27:234–238.

118. Taylor JE, Peat N, Porter C, Morgan AG. Regular, low-doseintravenous iron therapy improves response to erythropoie-tin in haemodialysis patients. Nephrol Dial Transplant1996; 11:1079–1083.

119. Gasche C, Dejaco C, Waldhoer T, Tilinger W, Reinisch W,Fueger GF, Gangl A, Lochs T. Intravenous iron and erythro-poietin for anemia associated with Crohn’s disease. Ann IntMed 1997; 126:782–787.

120. Weiss G, Houston T, Kastner S, Johrer K, Grunewald K,Brock JH. Regulation of cellular iron metabolism by erythro-poietin: activation of iron-regulatory protein and up regula-tion transferrin receptor expression in erythroid cells. Blood1997; 89:680–687.

478 Goodnough

121. Eschbach JW, Adamson JW. Iron overload in renal failurepatients: changes since the introduction of erythropoietintherapy. Kidney Int 1999; 55:S35–S43.

122. Weiss G. Iron and anemia of chronic disease. Kidney Int1999; 55:S12–S17.

123. Hamstra RD, Block MH. Erythropoiesis in response to bloodloss in man. J Appl Physiol 1969; 27:503–507.

124. Beutler E. The utilization of saccharated Fe59 oxide in red cellformation. J Lab Clin Med 1958; 51:415–419.

125. Wood JK, Milner FA, Pathak UN. The metabolism of irondextran given as a total dose infusion to iron deficient Jamai-can subjects. Br J Haematol 1968; 14:119–129.

126. Beamish MR, Davies AG, Eakins JD, Jacobs A, Trevett D.The measurement of reticuloendothelial iron release usingiron-dextran. Br J Haematol 1971; 21:617–622.

127. Major A, Mathez-Loic F, Rohling R, Gautschi K, Brugnara C.The effect of intravenous iron on the reticulocyte response torecombinant human EPO. Br J Haematol 1997; 98:292–294.

128. Auerbach M, Barker R, Bahrain H, et al. Intravenous ironoptimizes the response to erythropoietin in patients withanemia of cancer and cancer chemotherapy: results of a mul-ticenter, open-label, randomized trial. Blood 2001; 98:799a.

129. Cavill I. Disorders of iron metabolism. Diagnostic methods.Clin Haematol 1982; 11:259.273–277.

130. Eckfeldt JH, Witte DL. Serum iron: would analyticalimprovement enhance patient outcome? Clin Chem 1994;40:505–507.

131. Tietz NW, Rinker AD, Morrison SR. When is a serum ironreally a serum iron? The status of serum iron measurements.Clin Chem 1994; 40:546–551.

132. Cazzola M, Pootrakul P, Bergamaschi G, Huebers HA,Eng M, Finch CA. Adequacy of supply for erythropoiesis: invivo observations in humans. J Lab Clin Med 1987; 110:734–739.

133. Goodnough LT, Price TH, Rudnik S. Iron-restricted erythro-poiesis as a limitation to autologous blood donation in the


erythropoietin-stimulated bone marrow. J Lab Clin Med1991; 118:289–296.

134. Hastka J, Lasserre JJ, Schwarzbeck A, Reiter A, HehlmannR. Laboratory tests of iron status: correlation or commonsense? Clin Chem 1996; 42:718–724.

135. Cook JD, Lipschitz DA, Miles LEM, Finch CA. Serum ferritinas a measure of iron stores in normal subjects. Am J ClinNutr 1974; 27:681–687.

136. Lipshitz DA, Cook JD, Finch CA. A clinical evaluation ofserum ferritin as an index of iron stores. N Engl J Med1994; 290:1213–1922.

137. Olivieri NF, Brittenham GM, Matsui D, Berkautch M, Blen-dis LM, Cameron RG, McCelland RA, Liu PD. Iron-chelationtherapy with oral deferiprone in patients with thalassemiamajor. N Engl J Med 1995; 332:918–922.

138. Hallberg L, Bengtsson C, Lapidus L, Lindstedt G, LundbergPA, Hulten L. Screening for iron deficiency: an analysis basedon bone-marrow examinations and serum ferritin determina-tions in a population sample of women. Br J Haematol 1993;85:787–798.

139. Punnonen K, Irjala K, Rajamaki A. Serum transferrinreceptor and its ratio to serum ferritin in the diagnosis of irondeficiency. Blood 1997; 89:1052–1057.

140. Mast AE, Blinder MA, Gronowski AM, Chumley C, Scott MG.Clinical utility of the soluble transferrin receptor and com-parison with serum ferritin in several populations. ClinChem 1998; 44:45–51.

141. Brugnara C, Colella GM, Cremins JC, Langley RC,Schneider TJ, Rutherford CJ, Goldberg MA. Effects of subcu-taneous recombinant human erythropoietin in normal sub-jects: development of decreased reticulocyte hemoglobincontent and iron-deficient erythropoiesis. J Lab Clin Med1994; 123:660–664.

142. Leggett BA, Brown NN, Bryant SJ, Duplock L, Powell LW,Halliday JW. Factors affecting the concentrations of ferritinin serum in a healthy Australian population. Clin Chem1990; 36:1350–1798.

480 Goodnough

143. Ludwig H, Fritz E, Leitgeb C, Pecherstorfer M, Samonigg H,Schuster J. Prediction of response to erythropoietin treat-ment in chronic anemia of cancer. Blood 1994; 84:1056–1063.

144. Cazzola M, Messinger D, Battistel V, Bron D, Cimino R,Emller Zugler L, Esser SU, Greil R, Grossi A, Jager G.Recombinant human erythropoietin in the anemia associatedwith multiple myeloma or non-Hodgkin lymphoma:dose find-ing and identification of predictors of response. Blood 1995;86:4446–4453.

145. Henry D, Abels R, Larholt K. Prediction of response to recom-binant human erythropoietin (Epoietin-a) therapy in cancerpatients. Blood 1995; 85:1676–1678.

146. Kasper C, Terhaar A, Fossa A, Welt A, Seeber S,Nowrousian MR. Recombinant human erythropoietin in thetreatment of cancer-related anemia. Eur J Haematol 1997;58:251–256.

147. Cazzola M, Ponchio L, Pedrotti C, Farina C, Cerani P,Lucottic C, Novella A, Rovati A, Bergamaschi G, Beguin Y.Prediction of response to recombinant human erythropoietin(rHuEpo) in anemia of malignancy. Haematologica 1996;81:434–441.

148. Flowers CH, Skikne BS, Covell AM, Cook JD. The clinicalmeasurement of serum transferrin receptor. J Lab ClinMed 1989; 114:368–377.

149. Dasgupta AD, Abbi A. High serum transferrin receptor levelin anemia of chronic disorders indicates coexistent irondeficiency. Am J Hematol 2003; 72:158–161.


151. Suominen P, Punnonen K, Rajamaki A, Irjala K. Serumtransferrin receptor and transferrin receptor-ferritin indexidentify healthy subjects with subclinical iron deficits. Blood1998; 92:2934–2938.

152. Ferguson BJ, Skikne BS, Simpson KM, Baynes RD, Cook JD.Serum transferrin receptor distinguishes the anemia ofchronic disease from iron deficiency anemia. J Lab ClinMed 1992; 19:385–390.


153. Beguin Y, Loo M, R’Zik S, Sautois B, Lejeune F, Rorive G,Fillet G. Early prediction of response to recombinant humanerythropoietin in patients with the anemia of renal failure byserum transferrin receptor and fibrinogen. Blood 1993;82:2010–2016.

154. Cook JD, Skikne BS, Baynes RD. Serum transferrin receptor.Annu Rev Med 1993; 44:63–74.

155. Littlewood TJ, Zagari M, Pallister C, Perkins A. Baseline andearly treatment factors are not clinically useful for predictingindividual response to erythropoietin in anemic cancerpatients. Oncologist 2003; 8:99–107.


157. d’Onofrio G, Chirillo R, Zini G, Caenaro G, Tommasi M,Micciulli G. Simultaneous measurement of reticulocyte andred blood cell indices in healthy subjects and patients withmicrocytic and macrocytic anemia. Blood 1995; 85:818–823.

158. Beguin Y, Loo M, R’Zik S, Sautois B, Lejeune F, Rorive G.Quantitative assessment of erythropoiesis in haemodialysispatients demonstrates gradual expansion of erythroblastsduring constant treatment with recombinant human EPO.Br J Haematol 1995; 89:17–23.

159. Brugnara C. Use of reticulocyte cellular indices in the diagno-sis and treatment of hematological disorders. Int J Clin LabRes 1998; 28:1–11.

160. Mast AE, Blinder MA, Lu Q, Flax S, Dietzen DJ. Clinicalutility of the reticulocyte hemoglobin content in the diagnosisof iron deficiency. Blood 2002; 99:1489–1991.

161. Fishbane S, Galgano C, Langley RC Jr, Canfield W, MaesakaJK. Reticulocyte hemoglobin content in the evaluation ofiron status of hemodialysis patients. Kidney Int 1997; 52:217–222.

162. Mittman N, Sreedhara R, Mushnick R, Chattopadlyay S,Zelmonovic P, Vasagli M, Avram MM. Reticulocyte hemoglo-bin content predicts functional iron deficiency in hemodialy-

482 Goodnough

sis patients receiving rHuEPO. Am J Kidney Dis 1997; 30:912–913.

163. Bhandari S, Norfolk D, Brownjohn A, Turney J. Evaluation ofRBC ferritin and reticulocyte measurements in monitoringresponse to intravenous iron therapy. Am J Kidney Dis1997; 30:814–821.

164. Bhandari S, Turney JH, Brownjohn AM, Norfolk D. Reticulo-cyte indices in patients with end stage renal disease on hemo-dialysis. J Nephrol 1998; 11:78–82.

165. Davis BH, Bigelow N, Ball ED, Mills L, Cornwell GG III.Utility of flow cytometric reticulocyte quantification as a pre-dictor of engraftment in autologous bone marrow transplan-tation. Am J Hematol 1989; 32:81.

166. Sowade O, Sowade B, Brilla K, Franke W, Stephan P,Gross J, Scigalla P, Warnke H. Kinetics of reticulocytematurity fractions and indices and iron status during therapywith epoetin beta in cardiac surgery patients. Am J Hematol1997; 55:89–96.

167. Brugnara C, Zurakowski D, DiCanzio J, Boyd T, Platt O.Reticulocyte hemoglobin content to diagnose iron deficiencyin children. JAMA 1999; 281:2225–2247.

168. Barosi G. Inadequate erythropoietin response to anemia:definition and clinical relevance. Ann Hematol 1994;68:215–223.

169. Thompson JA, Gilliland DG, Prchal J, Bennett JM, Larholt K,Nelson RA, Rose EH, Dugan MH. Effects of recombinanthuman erythropoietin combined with ranulocyte=macrophage cologny-stimulating factor in the treatment ofpatients with myelodyplastic syndrome. Blood 2000; 95:1175–1179.

170. Ludwig H, Fritze, Kotzmann H, Hocker P, Gisslinger H,Barnas U. Erythropoieitin treatment of anemia associatedwith multiple myeloma. N Engl J Med 1990; 322:1693–1699.

171. Pincus T, Olson NJ, Russel IJ, Wolfe F, Harris ER, SchnitzerTJ, Boccagno JA, Krantz SB. Multicenter study of recombi-nant human erythropoietin in correction of anemia in patientswith rheumatoid arthritis. Am J Med 1990; 89:161–168.


172. Abels R, Larholt K, Krantz K, Bryant E. Recombinant humanerythropoietin for the treatment of the anemia of cancer. In:Murphy MJ Jr, ed. Blood Cell Growth Factors: Their Presentand Future Use in Hematology and Oncology. Dayton, OH:Alpha Med, 1991:121–141.

173. Glaspy J, Bukowski R, Steinberg D, Taylor C, Tchekme-dyians S, Vadhan Raj S. Impact of therapy with epoietin alfaon clinical outcomes in patients with nonmyeloid malignan-cies during cancer chemotherapy in community oncologypractice. J Clin Oncol 1997; 15:1218–1234.

174. Glaspy B, Demetri G, Cremieux P. Does baseline endogenouserythropoietin level impact hemoglobin (Hb) response torHuEPO in cancer patients? Blood 1999; 94:15B.

175. Auerbach M, Witt D, Toler W, Fierstein M, Lerner RG,Ballard H. Clinical use of the total dose intravenous of irondextran. J Lab Clin Med 1988; 111:566–569.

176. Hamstra RD, Block MH, Schocket AL. Intravenous irondextran in clinical medicine. JAMA 1980; 243:1726–1729.

177. Woodman J, Shaw RJ, Shipman AJ, Edwards AM. A surveil-lance program on a long-established product: Imferon (IronDextran BP). Pharmaceut Med 1987; 1:289–296.

178. Fishbane S, Ungureanu VD, Maeska JK, Kaupke CJ, Lim V,Wish S. The safety of intravenous iron dextran in hemodialy-sis patients. Am J Kidney Dis 1996; 28:529–534.

179. Auerbach M, Winchester J, Wahab A, Richards K, McGinleyM, Hall F, Anderson J, Briefel G. A randomized trial of threeiron dextran infusion methods for anemia in erythropoietin-treated dialysis patients. Am J Kidney Dis 1998; 31:81–86.

180. Faich G, Strobos J. Sodium ferric gluconate complex insucrose: safer intravenous iron therapy than iron dextran.Am J Kidney Dis 1999; 33:464–470.

181. Sunder-Plassmann G, Horl WH. Safety aspects of parenteraliron in patients with end stage renal disease. Drug Safety1997; 17:241–250.

182. Silverberg DS, Blum M, Peer G, Kaplan E, Iaina A. Intrave-nous ferric saccharate as an iron supplementation in dialysispatients. Nephron 1996; 72:413–417.

484 Goodnough

183. Sunder-Plassmann G, Horl WH. Importance of iron supplyfor erythropoietin therapy. Nephrol Dial Transplant 1995;10:2070–2076.

184. Nissenson AR, Lindsay RM, Swan S, Sellgman P, Strobos J.Sodium ferric gluconate complex in sucrose is safe and effec-tive in hemodialysis patients: North American Trial. Am JKidney Dis 1999; 33:471–482.

185. Calvar C, Mata D, Alonso C, Ramos B, Lopez de Novales E.Intravenous administration of iron gluconate during haemo-dialysis. Nephrol Dial Transplant 1997; 12:574–575.

186. Pascual J, Teruel JL, Liano F, Sureda A, Ortuna J. Seriousadverse reactions after intravenous ferric gluconate. NephrolDial Transplant 1992; 7:271–272.

187. Zanen AL, Adriaansen HJ, Van Bommel EFH, Posthuma R,Th. de Jong GM. ‘‘Over saturation’’ of transferrin after intra-venous ferric glucomate (Ferrlecit) in haemodialysis patients.Nephrol Dial Transplant 1996; 11:820–824.

188. Seligman PA, Schleicher RB. Comparison of methods used tomeasure serum iron in the presence of iron gluconate or irondextran. Clin Chem 1999; 45:898–900.

189. Henderson PA, Hillman RS. Characteristics of iron dextranutilization in man. Blood 1969; 34:357–375.

190. Goodnough LT, Monk TG, Andriole GL. Erythropoietintherapy. N Engl J Med 1997; 336:933–938.

191. Messmer K. Consensus statement: using epoietin alfa todecrease the risk of allogeneic blood transfusion in the surgi-cal setting. Sem Hematol 1996; 33:73–80.

192. Goldberg MA, McCutchen JW, Jove M, DiCesare P,Friedman RJ, Poss R, Guilfoyle M, Freid D, Young DC. Asafety and efficacy comparison study of two dosing regimentsof epoietin alfa in patients undergoing major orthopedicsurgery. Am J Orth 1996; 25:544–597.

193. Doolittle RF. Biotechnology—the enormous cost of success.N Engl J Med 1991; 324:360–362.


194. Goodnough LT, Spencer R, Shander A. Bloodless Medicine:Clinical care without allogeneic blood transfusion 2003;43:668–676.

195. Goodnough LT, Shander A, Brecher ME. Transfusion Medi-cine: looking to the future. Lancet 2003; 361:161–169.

196. Eschbach JW, Egrie JC, Downing MR, et al. Correction of theanemia of endstage renal disease with recombinant humanerythropoietin: results of a combined phase I and II clinicaltrial. N Engl J Med 1987; 316:73–78.

197. Muirhead N. Recombinant human erythropoietin (epoietinalpha) in anemic patients on hemodialysis: Canada. In:Erslev AJ, ed. Erythropoietin. Baltimore, MD: Johns HopkinsUniversity Press, 1991:241–268.

198. Kaupke CJ, Butler GC, Vaziri ND. Effect of recombinanthuman erythropoietin on platelet production in dialysispatients. J Am Soc Nephrol 1993; 10:1672–1675.

199. Taylor JE, Henderson IS, Stewart WK, et al. Erythropoietinand spontaneous platelet aggregation in heamodialysispatients. Lancet 1991; 338:1361–1362.

200. Taylor JE, Belch JJF, McLaren M, et al. Effect of erythropoie-tin therapy and blood withdrawal on blood coagulation andfibrinolysis in hemodialysis patients. Kidney Int 1993;44:182–190.

201. Gordge MP, Leaker B, Patel A, et al. Recombinant humanerythropoietin shortens the uremic bleeding time withoutcausing intravascular haemostatic activation. Thromb Res1990; 57:171–182.

202. Vigano G, Benigni A, Mendogni P, et al. Recombinant humanerythropoietin to correct uremic bleeding. Am J Kidney Dis1991; 1:44–49.

203. Raine AEG. Hypertension, blood viscosity, and cardiovascu-lar morbidity in renal failure: implications for erythropoietintherapy. Lancet 1988; I:97–100.

204. Besarab A, Bolton WK, Browne JK, et al. The effects of nor-mal as compared with low hematocrit values in patients withcardiac disease who are receiving hemodialysis and epoietin.N Engl J Med 1998; 339:584–590.

486 Goodnough

205. Goodnough LT, Rudick S, Price TH, et al. Increased collectionof autologous blood preoperatively with recombinant humanerythropoietin therapy. N Engl J Med 1989; 321:1163–1167.

206. Goodnough LT, Price TH, EPO Study Group. A phase III trialof recombinant human erythropoietin in non-anemic ortho-paedic patients subjected to aggressive autologous blood phle-botomy: dose, response, toxicity, and efficacy. Transfusion1994; 34:66–71.

207. Price TH, Goodnough LT, Vogler W, et al. The effect of recom-binant erythropoietin therapy administration on the efficacyof autologous blood donation in patients with low hemato-crits. Transfusion 1996; 36:29–36.

208. Goodnough LT, Monk TG. Evolving concepts in autologousblood procurement: case reports, of perisurgical anemia compli-cated by myocardial infarction. Am J Med 1996; 101:12a–33s.

209. Biesma DH, Bronkhorst PJH, de Groot PG, et al. The effect ofrecombinant human erythropoietin on hemostasis, fibrinoly-sis, and blood rheology in autologous blood donors. J Lab ClinMed 1994; 124:42–47.

210. Sowade O, Ziemer S, Sowade B, et al. The effect of preopera-tive recombinant human erythropoietin therapy on plateletsand hemostasis in patients undergoing cardiac surgery. JLab Clin Med 1997; 129:376–383.

211. Sowade O, Warnke H, Scigalla P, et al. Avoidance of allo-geneic blood transfusions by treatment with epoetin beta(recombinant human erythropoietin) in patients undergoingopen heart surgery. Blood 1997; 89:411–418.

212. Sowade O, Gross J, Sowade B, et al. Evaluation of oxygenavailability with oxygen status algorithm in patientsundergoing open heart surgery treated with epoetin beta.J Lab Clin Med 1997; 129:97–105.

213. D’Ambra MN, Gray RJ, Hillman R, et al. Effect of recombi-nant human erythropoietin on transfusion risk in coronarybypass patients. Ann Thorac Surg 1997; 64:1686–1693.

214. Goodnough LT, Despotis GJ, Parvin CA. Erythropoietin ther-apy in patients undergoing cardiac operations. Ann ThoracSurg 1997; 64:1579–1580.


215. Monk TG, Goodnough LT, Brecher ME, et al. A prospective,randomized trial of three blood conservation strategies forradical prostatectomy. Anesthesiology 1999; 91:24–33.

216. Kaisi M, Ngwalle EWK, Runyoro DE, Rogers J. Evaluation oftolerance of and response to iron dextran administered bytotal dose infusion to pregnant women with iron deficiencyanemia. Int J Gynecol Obstet 1988; 26:235–243.

217. Bennett C, Luminari S, Nissenson A, et al. Pure red cell apla-sia and epoietin therapy Nengl J Med 2004; 351:1403–1408.

218. Lasadevall N, Nataf J, Viron B, et al. Pure red-cell aplasia andantierythropoietin antibodies in patients treated with recom-binant erythropoietin. N Engl J Med 2002; 346:469–475.

219. Gershon SK, Luksenburg H, Cote TR, Braun MM. Pure red-cell aplasia and recombinant erythropoietin. N Engl J Med2002; 346:1584–1585.

220. Brines ML, Ghezzi P, Keenan S, et al. Erythropoietin crossesthe blood-brain barrier to protect against experimental braininjury. PNAS 2000; 97:10526–10531.

221. Brezden. Cognitive function in breast cancer patients receiv-ing adjuvant chemotherapy. J Clin Onc 2000; 18:2695–2701.

222. Egrie JC, Browne JK. Development and characterization ofnovel erythropoiesis stimulating protein (NESP). Br J Cancer2001; 84(suppl 1):3–10.

223. Glaspy J, Jadeja JS, Justice G, et al. A dose-finding and safetystudy of novel erythropoiesis stimulating protein (NESP) forthe treatment of anaemia in patients receiving multicyclechemotherapy. Br J Cancer 2001; 84(suppl 1):17–23.

224. Smith RE, Jaiyesimi IA, Meza LA, Tchekmedylan NS, ChanD, Griffith H, et al. Novel erythropoiesis stimulating protein(NESP) for the treatment of anaemia of chronic disease asso-ciated with cancer. Br J Cancer 2001; 84(suppl 1):24–30.

225. Mirtsching B, Chary V, Vadhan-Raj S, et al. Every 2weekdarbopoietin alfa is comparable to rHuEPO in treatingchemotherapy-induced anemia. Oncology 2002; 16:7–12.

488 Goodnough

18

Positive Effects of Correctionof Anemia in Malignant Diseases

GUDRUN POHL and HEINZ LUDWIG

Department of Medicine I and MedicalOncology, Wilhelminenspital, Vienna, Austria

INTRODUCTION

Definition of Anemia

Anemia is the most common hematological manifestation ofmalignant diseases, and it may significantly impair organand tissue function and the general condition of cancerpatients. It is defined as an inadequate circulating numberof red blood cells or level of hemoglobin and may arise as aresult of the underlying disease, chemotherapy, or radiationtherapy (1). Normal limits for hemoglobin values are12.0–16.0 g=dL for women and 14.0–18 g=dL for men. TheNational Cancer Institute (NCI) and the World Health

PART VI: CONTROVERSIES IN ACD THERAPY

489

Organization (WHO) toxicity criteria slightly differ in theirclassification of lesser grades of anemia, but are the same intheir classification of more severe grades of anemia (Table 1).

Mechanisms of Anemia in Cancer Patients

Several mechanisms contribute to the development of anemiain cancer patients (Table 2). Anemia may arise as a conse-quence of treatment with cytotoxic chemotherapy and radio-therapy or as chronic anemia of cancer. This type of anemiais a consequence of the presence of the malignant diseaseitself and is referred to as the anemia of chronic disease(ACD). The syndrome of ACD is characterized by a hypore-generative, normocytic, normochromic anemia associated

Table 1 The National Cancer Institute (NCI) and the WorldHealth Organization (WHO) Toxicity Criteria for Anemia

Severity WHO NCI

Grade 0 (within normal limits) >11.0 g=dL Within normal limitsGrade 1 (mild) 9.5–10.9 g=dL 10.0 g=dL to normal limitsGrade 2 (moderate) 8.0–9.4 g=dL 8.0–10.0 g=dLGrade 3 (serious=severe) 6.5–7.9 g=dL 6.5–7.9 g=dLGrade 4 (life threatening) < 6.5 g=dL < 6.5 g=dL

Table 2 Mechanisms of Anemia in Cancer Patients

Hemodilution from impaired diuresisBleedingHypersplenismAutoimmune hemolysisInfectionsNutritional deficiencies (iron, folate, vitamin B12)Renal impairmentAbnormalities of bone marrow functionCytokine activation (e.g., IL-1, TNF-a, TGF-b)ChemotherapyRadiotherapyTumor itself


490 Pohl and Ludwig

with reduced serum iron and transferrin saturation butelevated (or normal) ferritin levels. Anemia of chronic dis-ease in cancer patients is thought to be a cytokine-mediateddisorder. Tumor interaction with the immune system leads tooverproduction of inflammatory cytokines such as interleukin-1and tumor necrosis factor-a. These cytokines can impair ery-throid colony formation in response to erythropoietin (EPO),decrease the life span of erythrocytes, impede EPO produc-tion, and prevent the normal utilization of iron (2,3).

Other causes of anemia in cancer are mechanisms oftennot related to the underlying malignancy such as vitaminB12, folate, and iron deficiency, as well as endocrine and renaldisorders or, more frequently, complications of the malignantdisease. Among these, blood loss, hemolysis, splenomegaly,DIC, hypervolumia, clonogenic disorders of hematopoiesis,replacement of hematopoietic precursors and disruption ofthe microenvironment by extensive tumor infiltration or bonemarrow fibrosis (4), and kachexia are most frequent.

Signs and Symptoms of Anemia in Cancer Patients

Anemia of cancer patients often develops rather slowly andcan impact on virtually all organs. The severity of anemia-related symptoms depends on several factors including Hblevel, velocity of onset of anemia, age, comorbidities, extentof the underlying malignancy, intensity of treatment, and bio-logical function of the patients organs (5). In elderly patients,clinical symptoms may occur at higher Hb levels than inyounger patients.

General Symptoms

Symptoms of anemia include a gradual onset in fatigue,lethargy, and inability to concentrate and are identified oftenat hemoglobin levels less than 12g=dL (6). Fatigue is the mostcommon symptom of these patients, with prevalence rangingfrom 60% to more than 90% (7). Severe chronic anemia can leadto cardiorespiratory decompensation and may significantlyimpair the patient’s ability to perform normal daily activities.

Correction of Anemia in Malignant Diseases 491

Cardiovascular Symptoms

In anemia, decreased blood viscosity and decreased bloodoxygen lead to hypoxia of peripheral tissues and increasedproduction of lactate. This results in the accumulation ofvasoactive substances such as adenosine and bradykinincausing vasodilatation and decreased peripheral vessel resis-tance. Finally, these peripheral mechanisms culminate in arise in cardiac workload and subsequent cardiac hyperactiv-ity, which is the most important compensatory mechanismin anemia (8,9). The extent of cardiac compensation dependson the degree of anemia as well as on other factors suchas pre-existing diseases of the heart, lungs, and peripheralvessels. The body’s adaption to anemia is an increase in thecardiac index. This is mostly achieved by acceleration ofthe heart rate rather than by an increased stroke volume(10,11). As the patient experiences this hyperdynamic heartsyndrome, he or she may complain of palpitations, sometimesassociated with pounding pulse, and of significant aggrava-tion of underlying conditions such as angina pectoris, claudi-cation, or cerebral ischemia (12,13).

In mild anemia, the heart rate may only become abnor-mal during episodes of physical exercise, but in more severeforms of anemia, tachycardia associated with a wide pulsepressure occurs even at rest. Electrographic changes can alsooccur in severe anemia and may include depression of the S-Tsegment and inverted T-waves, indicating insufficient oxygendelivery to the heart muscle (14).

The physical examination of the chronic anemic patientusually reveals forceful apical tachycardia and strong periph-eral pulses, and functional systolic murmurs that can be heardabove the heart and the carotid arteries (15). Often, they aremoderate in intensity, and at times may be rough in qualityand raise suspicion of organic valvular heart disease. Withsevere anemia, the murmurs may be a sign of the mitral andtricuspid insufficiency that results from cardiac dilatation.Gallop rhythms of both the S3 and S4 type may occur (15).The patient may complain of fatigue, shortness of breath,and palpitations, particularly during and following exercise.

492 Pohl and Ludwig

Respiratory Symptoms

One of the physiologic compensatory mechanisms in anemia isa rise of the respiratory rate in an attempt to increase bloodoxygenation. If the respiratory rate exceeds a certain level,the patient will experience shortness of breath, a symptommost often developing in a state of exertion when reducedblood oxygen meets increased oxygen requirements of exercis-ing muscles. However, patients suffering from severe anemiamay complain of dyspnea even while resting (16). In a stateof dyspnea, the minute ventilation and the residual air areincreased, but the forced expiratory volume is reduced (15).In addition, the lungs may also be directly affected by anemia.If increased cardiac workload results in congestive heartdisease, pulmonary edema may occur as a consequence ofincreased blood pressure in the pulmonary circulation. Thisleads to severe dyspnea and may even become life threatening.

Skin Changes

The skin of patients suffering frommoderate to severe anemiausually appears pale and cold as a result of blood shifting tovital organs and also because of cutaneous vasoconstriction(15). Pallor, however, is not specific for anemia and may haveseveral other causes such as habitual pallor, myxedema, vaso-vagal dysregulation, and edema. On the other hand, jaundice,cyanosis, racial skin pigmentation, and dilatation of periph-eral vessels may all mask the pallor of anemia (15). The pallorof mucous membranes, particularly of the mouth and phar-ynx, the conjunctivae, the lips, and the nail beds, is morespecific for anemia because these parts of the integumentare less affected by the factors mentioned above. The palmsusually also appear pale, but the creases retain their darkpink color. If the Hb level drops below 7g=dL, however, thecreases take the color of the surrounding skin, which maythus be regarded as a sign of severe anemia (15). Otherchanges of the skin sometimes observed in anemic patientsinclude a loss of normal skin elasticity and brittle and orebroken finger nails, especially in patients with chronic iron-deficiency anemia (17).


Gastrointestinal Symptoms

Disturbances of the gastrointestinal tract are rather commonin anemia. Some of them, such as duodenal ulcer, carcinomaof the gastrointestinal tract or glossitis, and atrophy of thetongue papillae in pernicious anemia, may be manifestationsof the disorder underlying the anemia. A reduction in the per-fusion of the intestinal mucosa may lead to the symptomsmentioned above as well as in malabsorption. In combinationwith the increased metabolic rate in anemia, malabsorptionmay result in a decline in the general state of nutrition.Irregular bowel movements and indigestion have also beenreported in anemic patients.

Renal Symptoms

In mild-to-moderate anemia, renal function will not beaffected, as blood is directed from peripheral tissue to the kid-neys. Slight proteinuria is a common finding in patients withsevere anemia. In severe chronic anemia, renal blood flowmay decrease and result in impaired renal function, includingrenal edema (18).

The kidneys are also among the organs responsiblefor compensating for anemia by increasing erythropoietinsecretion triggered by low Hb levels (19). Conversely, severeimpaired renal function resulting in deficient erythropoietinproduction may itself be a cause of anemia (20). This specificform of anemia can be very successfully treated with recombi-nant human erythropoietin (rHuEPO, epoetin alfa).

Genitourinary Symptoms

Symptoms arising from the genitourinary tract are relativelyfrequent in patients with anemia and may partly result fromimpaired secretion of sexual hormones. Menorrhagia, irregu-lar menstrual cycles, and amenorrhea are among the mostcommon symptoms reported by women. Males may sufferfrom impotence, and the loss or disturbance of libido in bothsexes may contribute significantly to impaired quality of life(QOL) (5,21). Correction of anemia can improve these symp-toms significantly.

494 Pohl and Ludwig

Immune Function

In patients with iron-deficiency anemia, impairment of theimmune system has been observed based on a decrease inB- and T-cell function (22). The immune deficiency commonlyobserved in anemic dialysis patients cannot be completelyreversed by correcting uremia, which implies that anemiamay also play a role in this regard. In vitro studies have shownthat interleukin-2 secretion by peripheral blood mononuclearcells of anemic dialysis patients is significantly reduced com-pared to normal controls and that correction of anemia inthese patients can restore several parameters of immune func-tion. Repeated or chronic infections resulting from thisimmune deficiency may in turn aggravate anemia. Addition-ally, the decreased perfusion of skin and mucous membranesin anemia may contribute to a higher rate of infections, as ithas been shown that anatomic compartments with poor perfu-sion are prone to infection. Repeated or chronic infectionsresulting from this immune deficiency may in turn aggravateanemia. It is important to note, however, that subfebrile tem-peratures in severely anemic patients or even fever of milddegree might be a consequence of the increasedmetabolic rate.

Neurological Symptoms

Cerebral hypoxia in anemia may lead to symptoms such asdepression, sleeplessness, reduced vigilance, headache,vertigo, tinnitus, and dizziness. In anemic patients withchronic renal failure, impaired cognitive function and anincreased frequency of depression have been observed. Impor-tantly, these symptoms were not related to uremia but ratherto low Hb levels. In several studies, it has been shown thatcorrection of anemia by administering blood transfusions orerythropoietin results in significant improvement of cognitivefunction measured by several scales (23,24). Additionally,severe anemia may affect the retina and cause cotton woolspots, exudates, or hemorrhages, which have been describedoccasionally. These retinal changes appear not to be corre-lated to the degree of anemia, and they have been found moreoften in men than in women (25).


Table 3 summarizes the numerous signs and symptomsof anemia in cancer patients.

Incidence and Prevalence of Cancer-RelatedAnemia

The probability of occurrence of anemia in cancer patientsdepends on various factors including age of the patient, thepresence or absence of infections, histology and stage ofthe tumor, and other present comorbidities. There is a high

Table 3 Signs and Symptoms of Anemia in Cancer Patients

Organ Signs and symptoms of anemia

Neuromuscular system FatigueDecreased cognitive functionDizziness, vertigoDepressive moodRetinal damageImpaired cognitive functionMuscular weakness

Cardio-respiratory system Exertional dyspneaIncreased ejection fractionTachycardia, palpitationsCardiac enlargement, hypertrophyIncreased pulse pressure,systolic ejection murmur

Reduced oxygen diffusionCardiac failure

Skin PallorBrittle=broken nailsCold skin

Gastrointestinal system NauseaAnorexiaConstipationMalabsorption, malnutrition

Genitourinary system ProteinuriaWater retentionMenstrual problemsLoss of libido, impotency

Immune system Decreased immune reactivityImpaired B-, T-cell and macrophage function

496 Pohl and Ludwig

incidence of anemia (50–60%) in patients with lymphomas,multiple myeloma, lung tumors, gynecologic and genitourin-ary tumors. The occurrence of anemia in patients with solidtumors is less than that observed for hematological malignan-cies. But incidence of mild-to-moderate anemia can be high inthese diseases (6,7,26).

A recent survey, the European Cancer Anaemia Survey(ECAS), enrolled 15,367 cancer patients representative forthe distribution of most tumor types as seen in the clinical set-ting in Europe (27). They were followed for up to 6 months. Atenrolment, the prevalence of anemia (hemoglobin <12.0 g=dL)was 39% in the entire patient group; anemia rates were35% in newly diagnosed patients, 48% in patients with per-sistent=recurrent disease, and 31% in patients in remission.Patients most frequently anemic had hematological malig-nancies (53.0%), and were receiving chemotherapy (50.0%).Anemia was found to be closely associated with a poorperformance status, but the majority of patients wereleft untreated.

In the subset of 2732 patients who were not anemicat enrollment and were started to be treated with antitumortherapy during the ECAS survey, the incidence of anemia wasstudied. The incidence of anemia was highest in patients withlung cancer (70.9%) and gynecologic malignancies (64.6%). Pati-ents with head and neck, and urogenital cancers had relativelylower incidences of anemia (34.7% and 38.6%, respectively).For patients who received chemotherapy, incidence of anemiawas 67.7% for those with persistent=recurrent disease, 61.3%for those newly diagnosed and not receiving cytotoxic treatmentat enrolment, and 48.3% for those in remission. For patientswho received radiotherapy, incidence of anemia was 23.3% forthose with persistent=recurrent disease, 19.2% for those newlydiagnosed and not receiving cytotoxic treatment at enrolment,and 17.6% for those in remission.

The incidence of anemia in patients with platinum-basedtherapies as commonly used in lung and ovarian cancer wasas high as 75.1%, compared to 54.2% in patients receivingnonplatinum drugs, which accords with previous observations(Table 4).


INDICATIONS FOR ANEMIA TREATMENT

Anemia reduces QOL and may delay or limit effectiveanticancer treatment. It has been shown that anemia is anindependent prognostic factor of disease outcome (28–30).Overall survival and loco-regional recurrences followingradiation therapy may be affected by anemia. Despite thesefacts, anemia remains underestimated and not treated inmany cancer patients.

Initiation of treatment seems indicated if the patientsuffers from anemia-related symptoms and reports a reductionin QOL related to anemia. Young, otherwise healthy indivi-duals may well tolerate even very low hemoglobin levels withonly subclinical impairment of physical function. Patientswith multimorbidity may develop severe symptoms even withminor reductions of hemoglobin levels. This clinical realitymakes individualization of anemia treatment mandatory,but these considerations are often neglected by the medicalcommunity. Most groups base their decision to initiate EPOtherapy on hemoglobin levels only, which may conflict witha patient oriented approach. In addition, the hemoglobin levelat which to initiate therapy, stop treatment, or increase doses

Table 4 Estimated Frequency of Anemia During CancerTreatment

DiseasePercentage of

anemic patients

Percentage ofpatients receiving

transfusions

Multiple myeloma 47 22Cervical cancer 34 16Ovarian cancer 34 15Lung cancer 34 13Non-Hodgkin’s lymphoma 32 16Testicular cancer 32 14Prostate cancer 21 12Colorectal cancer 13 8Breast cancer 12 8


498 Pohl and Ludwig

of EPO are still a matter of debate (31–33). Guidelines issuedby different committees recommend initiation of treatment athemoglobin of �10g=dL. In addition, these guidelines recom-mend consideration of EPO therapy at hemoglobin levels upto 12 g=dL if the patient suffers from symptoms of anemia(34,35). An analysis on the incremental gain in QOL in rela-tion to the increase in hemoglobin achieved during erythro-poietin treatment revealed the biggest benefit of therapywhen hemoglobin is increased from 10 to 12 g=dL. Increasinghemoglobin beyond that level leads to further, albeit propor-tionally much lower gain in QOL (36).

TREATMENT OPTIONS

In cancer patients, anemia usually improves as the diseaseresponds to treatment. In cases where the malignancy seemsrefractory to further therapy or in patients who respond toanticancer treatment but without improvement in theiranemia, or in patients with severe anemia symptoms, itbecomes necessary to treat the anemia. Treatment of anemiaimproves anemia-related symptoms, QOL, and may haveimpact on antitumor therapy.

Therapeutic strategies, based on the clinical situationand on the underlying etiology of anemia, include blood transfu-sions and administration of rHuEPO, and when appropriateiron substitution.

Red Blood Cell Transfusions

Before the introduction of recombinant human erythropoietinin 1985 (rHuEPO, epoetin alfa) (37,38), red blood cell transfu-sions were the only treatment option for improvement of Hblevels to reduce anemia-related symptoms. One package ofred blood cells contains about 1.7�1012 red cells, 0.1�106

leukocytes, 0.2�109 platelets in a total volume of 270mL.The hematocrit is about 70%, hemoglobin value 24 g=dL, andthere is an average of 200–250mg iron per package.


Red blood cell transfusions provide immediate increase ofpatients’ hemoglobin levels and a quick relieve of symptoms,but have usually a short and only transient effect. Thisunavoidable fluctuation is a serious disadvantage.

As red blood cell transfusions were withheld until hemo-globin levels declined to more severe levels (7–8 g=dL) or thepatient experienced signs and symptoms of severe anemia,the perception developed that anemia that did not reach thetransfusion trigger point is clinically unimportant. The meantrigger level in cancer patients in Europe is 8.2 g=dL. Consid-ering new knowledge on the detrimental effect of anemia onQOL, this trigger level appears to be inappropriate in mostcases. Even with a trigger level around 9g=dL, patients ontransfusion therapy spend most of their time with a hemoglo-bin level that is certainly suboptimal.

However, blood transfusions pose several risks to therecipient (Table 5–7). The effectiveness of red blood cell trans-fusions may decrease after frequent applications due toalloimmunization and antibody production. Another disad-vantage, particularly in patients with long standing need oftransfusions, is the iron overload or secondary hemochroma-tosis after repeated transfusions in long-surviving cancerpatients (39). This bears the risk of liver damage, and morerecent findings report the accumulation of iron in the myocar-dium (40). Especially in elderly patients, both the anemia and

Table 5 Adverse Effects of Red Blood Transfusions

Immunologic Nonimmunologic

Alloimmunization InfectionsABO, Rh, HLA Hepatitis, HIV, CMV, EBV

HTLV-I & II, HHV-8Parvovirus, bacteria, plasmodiaParasites, prions, others . . .

Plasma proteins(neutrophil=platelet antigens)

Iron overload

Immunosuppression (# CD4=CD8 ratio,# NK cells)

Volume overload

Graft vs. host disease

500 Pohl and Ludwig

the iron overload reduce the functional capacity of the heart,thereby reducing chances of survival.

Other hazards of red blood cell transfusions include immu-nosuppression, which may enhance tumor growth, adversehemolytic reactions, and infections (1,26). Given the risk ofblood transfusion, extreme caution should be exerted whendealing with patients affected with curable or long-life expe-ctancy cancers such as early breast cancer.

Several patients are skeptic to undergo red blood celltransfusions on personal, religious, or logistical reasons. Bloodtransfusions may be inconvenient for the patient, and usuallyhave to be given in a hospital environment with the associatedresource implication and reduction in patient autonomy.

Among the numerous attempts which have been made toidentify factors that are predictive of a future need for trans-fusion, a low initial hemoglobin level was the only parameterconsistently found been associated with the likelihood offuture transfusion (41–43).

Recombinant human erythropoietin treatment (seebelow) is a physiologic and often a better choice than transfu-sions for the treatment of cancer-related anemia. However,about 30% of cancer patients do not respond to epoietin treat-ment, and for these patients, red cell transfusions may be theonly option. Red blood cell transfusions should be adminis-tered in cancer patients who need fast symptom relieve orwho are unresponsive to rHuEPO.

For the majority of the patients, blood transfusions arean effective way of increasing hemoglobin levels relativelysafely and improving their sense of well being. The beneficialeffects usually last for 1–3 weeks at which point another bloodtransfusion may be needed. Severe anemia with symptoms ofdyspnea or fatigue is still an indication for red blood celltransfusions.

Despite the immediate relief of anemia-related symp-toms by red blood cell transfusions, clinicians often admi-nister transfusions only when the hemoglobin levels dropbelow 10 or even 8 g=dL and=or when patients show signs ofsevere anemia. This common practice results in insufficienttreatment of anemia symptoms in many patients.


Table

6Transfusion

Rea

ctions

Onset

Rea

ction

Tim

eafter

transfusion

Pathom

echamism

Clinical

manifestation

Interm

ediate

Hem

olytic

Acu

teAB0�,Rh�

incompatibility

Fev

er,back

pain,

sensa

tion

ofch

est

compression,

hypoten

sion

,nausea,

vom

iting,mildto

life

threatening

Feb

rile

non

hem

olytic

(3–7%)

Acu

teHLA-,leukocyte-platelet-,

plasm

a-antigen

s,bacteria

orother

pyrogen

s

Chills,fever,hea

dach

e,malaise,

mildto

life-threatening

Non

card

iogen

icpulm

onary

edem

a

Acu

teTransfusion

sof

HLA-

antibod

ies,

agglutination

ofgranulocytesor

leukocyte

Chills,fever,ch

estpain,

cyanosis,hypoten

sion

,pulm

onary

edem

a

Other

allergic

Acu

tePlasm

aantigen

s,IgA

Urticaria,sk

inreaction,

bronch

ospasm

,angioed

ema,

anaphylactic

shock

502 Pohl and Ludwig

Delayed

Delayed

hem

olytic

Anamnestic

alloa

ntibod

yresp

onse,anti-D

,anti-K

,anti-Fy

2–10days

Fev

er,anem

ia,jaundice,

hem

oglobinem

ia,

hem

oglobinuria

Post-transfusion

purp

ura

Dev

elop

men

tof

alloa

ntibod

ies—

platelet

specificantigen

s2–10days

Bleed

ing,hem

orrh

age,

life

threatening

Infections

Hep

atitis,

CMV,EBV,

HIV

,HTLV-I,

bacteria,plasm

odia

Daysto

weeks

Dep

endingon

type

ofinfection

Graft

vs.

host

disea

se

Rea

ctionof

transfused

lymphocytesagainst

hostantigen

sWee

ksto

mon

ths

Skin

rash

,hep

atitis,

enteritis,

bon

emarrow

insu

fficien

cy,mildto

life

threatening

Chronic

immunosuppression

#CD4=CD8ratio,

#NK

cells,

others

Mon

thsto

yea

rsIn

dication

sfordecreased

survival


Recombinant Human Erythropoietin

Erythropoietin is an endogenous hormone, produced primarilyin the kidney, which is the most important factor in regulatingerythropoiesis (44). The glycosylated 30.4kDa glycoproteinhormone exerts antiapoptotic activity on late erythroid pre-cursors and stimulates erythroid cell proliferation and differen-tiation, controlling the dynamic balance between erythropoiesisand erythrocyte loss in order to maintain the red cell volumeand, therefore, ensuring the oxygen supply to the tissues.Erythropoietin binds to the erythropoietin receptor, whichis mainly expressed on erythroid colony-forming units (45).Erythropoietin production is induced by tissue hypoxia, whichmay result from anemia or decreased ambient oxygen tension,increased oxygen affinity for hemoglobin, and any other stimu-lus that decreases delivery of oxygen to the tissues.

In 1985, the cDNA encoding for human erythropoietinwas isolated (37,46). This enabled biotechnical production of

Table 7 Risk Projections of Infective and Immunological BloodTransfusion Complications

Risk factorEstimated risk

per unit

Number ofdeaths=million

units

In fectionViralHepatitis A 1=1,000,000 0Hepatitis B 1=50,000–1=170,000 0–0.14Hepatitis C 1=2,000,000 <0.5HIV 1=2,000,000 <0.5HTLV types I and II 1=19,000 to <1=80,000 0Parvovirus B19 1=10,000 0

Bacterial contamination 1=500,000 0.1–0.25Immunological complicationsAcute hemolytic reactions 1=250,000–1=1,000,000 0.67Delayed hemolytic reactions 1=1000 0.4Transfusion-relatedacute lung injury

—

Transfusion-associated GvHD —

504 Pohl and Ludwig

large amounts of rHuEPO for clinical use. Its efficacy in treat-ing anemia of end-stage renal failure was well demonstrated(47,48). Since then, rHuEPO has been used in the treatmentof anemia associated with acquired immunodeficiency syn-drome, rheumatoid arthritis, neonatal prematurity, hemato-logical and nonhematological malignancies, myelodysplasia,and aplastic anemia. In 1990, it was shown that there is a sig-nificant proportion of cancer patients with a blunted rHuEPOresponse to their anemic condition as compared to patientswith anemia due to iron deficiency (49). Recombinant humanerythropoietin can increase erythrocytes and hemoglobinlevels, hence alleviating the symptoms of anemia while redu-cing patients’ requirement for blood transfusions (50–54).

Another erythropoiesis stimulating protein agent, darbe-poetin alfa has been developed for the treatment of anemia.Studies on human erythropoietin demonstrated that there isa direct relationship between the sialic acid-containing carbo-hydrate content of the molecule and its serum half-life and invivo biological activity, but an inverse relationship with itsreceptor binding affinity. These observations led to thehypothesis that increasing the carbohydrate content wouldlead to a molecule with enhanced biological activity. Hyper-glycosylated rHuEPO analogs were developed to test thishypothesis. Darbepoetin alfa (novel erythropoiesis stimulatingprotein, NESP) contains five N-linked carbohydrate chains(two more than rHuEPO) and has been evaluated in preclinicalanimal studies. Due to its increased sialic acid-containingcarbohydrate content, NESP is biochemically distinct fromrHuEPO, having an increased molecular weight and greaternegative charge. Compared with rHuEPO, it has an appro-ximately threefold longer serum half-life, greater in vivopotency, but has considerable lower receptor affinity. NESPis currently being evaluated in human clinical trials fortreatment of anemia and reduction in its incidence (55).

Continuous erythropoiesis receptor activator (CERA), anovel erythropoietic agent with an extended serum half-life,promoted a greater erythropoietic response than epoetinin animal models (56). CERA contains a single methoxy-polyethylene glycol polymer of approximately 30kDa in size


that appears to extend the half-life of this agent as comparedto epoietin. CERA binds less tightly to the EPO receptor anddissociates more quickly compared with epoetin. A novelmode of action is proposed by which rapid dissociation fromthe erythropoietin receptor together with an extended serumhalf-life result in an enhanced and sustained erythropoieticeffect through continuous modulated stimulation of erythro-poiesis. These pharmacokinetics may lead to enhanced ery-thropoietic activity, less frequent dosing and optimal patientoutcomes (57). In two phase I studies, healthy male volun-teers were randomized to receive CERA or placebo eitherintravenously or subcutaneously. A potent dose-dependenterythropoietic response was observed with both intravenousor subcutaneous administration (56). CERA has a potent,prolonged dose-dependent erythropoietic activity. Phase IIstudies of CERA administered every 3 weeks to patientswith cancer are currently ongoing.

Erythropoietin Treatment—PracticalConsiderations

To address uncertainties regarding indications and efficacy intreating anemia, the American Society of Clinical Oncologyand the American Society of hematology developed an evi-dence-based clinical practice guideline for the use of erythro-poietin in patients with cancer (58). The guideline panelfound good evidence to recommend use of epoietin as atreatment option for patients with chemotherapy-associatedanemia with a hemoglobin level less than 10g=dL. Use ofepoietin for patients with less severe anemia (hemoglobin<12g=dL but never below 10g=dL) should be determined byclinical circumstances. Good evidence from clinical trialssupported the use of subcutaneous epoietin thrice weekly(150U=kg tiw) for a minimum of 4 weeks. Less strong evidencesupported an alternative weekly (40,000 U=week) dosing regi-men, based on common clinical practice. With either adminis-tration schedule, dose escalation should be considered forthose not responding to the initial dose. In the absence ofresponse, continuing epoietin beyond 6–8 weeks does not

506 Pohl and Ludwig

appear to be beneficial. Epoietin should be titrated once thehemoglobin concentration reaches 12 g=dL. For multiple mye-loma and chronic lymphatic leukemia, an expert panel set upthe treatment recommendations for the use of epoetin (35). Itis recommended that treatment is initiated only after otherpossible causes of anemia were eliminated. Epoetin shouldbe administered to any patient with hemoglobin �10g=dL.Patients with hemoglobin 10–12g=dL should receive epoetinif they suffer from significant symptoms of anemia and=orhave progressively decreasing hemoglobin values. Dosageshould be initiated at 10,000 IU three times=week or 40,000IU once=week and be titrated to maintain hemoglobin at 12 g=dL. Nonresponsive patients (<1 g=dL increase over 4 weeks)may have their dose increased to 20,000 IU three times=weekor 60,000 IU once=week. Epoetin treatment should bediscontinued if there is no response to the increased dosage,or hemoglobin >14 g=dL. Treatment should resume forpatients who exceed 14g=dL, at a reduced dosage, if theirhemoglobin level falls below 12g=dL (58).

While rHuEPO outnumbers many of the advantages ofblood transfusions, response rates to this treatment are vari-able and in some trials a large proportion of patients (30–50%)did not response (59). Good risk patients seem to respond bet-ter to rHuEPO treatment. But, it is difficult to predict whichpatients will respond to rHuEPO treatment. It could beshown in one cohort of patients that median time to responsewas approximately 4 weeks, but it can take up to 12 weeks todetermine responsiveness (5). Predictive models for responseto rHuEPO treatment have been established but offer onlylimited clinical value. Studies have indicated that patientswith high levels of rHuEPO before treatment are less likelyto respond to rHuEPO (60,61).

Functional iron deficiency is a major factor limiting theefficacy or rHuEPO therapy and iron supplementation shouldbe considered. A recent study has suggested that intravenousiron replacement is more effective than oral iron but this find-ing requires confirmation (62).

Around 60% of patients will respond to rHuEPO with anHb increase of >2.0 g=dL. The median time to response is


approximately 4 weeks. Ludwig et al. (63) showed that if after2 weeks of treatment with rHuEPO, the serum EPO level was<100 IU=L and the Hb had increased by more than 0.5 g=dL,a positive response to treatment could be predicted with 95%accuracy. Low baseline erythropoietin levels or an increasein the transferrin receptor level after 2 weeks of treatmentpredicted those patients most likely to respond to treatmentwith erythropoietin (59,64).

Tolerance of Erythropoietin

rHuEPO treatment in anemic cancer patients is generallywell tolerated. Pain or erythema at the injection site has beenreported. rHuEPO induced hypertension, which occurs mostfrequently in patients with chronic renal failure, is uncom-mon in anemic cancer patients, but may be seen in those witha history of insufficiently controlled hypertension. rHuEPOmay exacerbate splenomegaly secondary to extramedullaryhematopoiesis through the stimulation of erythropoiesis inmyeloproliferative disorders (65).

In theory, treatment with rHuEPO may stimulate thegrowth of malignant cells if those tumor cells express surfacereceptors for erythropoietin. It was reported that rHuEPO,either alone or in the presence of other cytokines, does notmodulate growth of human cancer cells in vitro (66). Recentfindings, however, showed contradictory results. Tumor speci-mens from patients with breast cancer were shown to expresshigh levels of erythropoietin and of erythropoietin receptorswhile in normal breast tissue no or only low expression wasseen. In addition, incubation of breast cancer cell lines underhypoxic conditions led to upregulation of erythropoietin and oferythropoietin receptors and enhanced DNA synthesis (67).These observations warrant further careful investigationson possible enhancement of tumor growth by erythropoietintreatment in patients with certain cancers.

The occurrence of immune-mediated pure red cell aplasia(PRCA) cases associated with antierythropoietin antibodieshas recently been reported in renal patients treated withrecombinant erythropoietin alfa and=or beta (68). So far, nocases have been reported in cancer patients on chemotherapy.

508 Pohl and Ludwig

Increasing costs for rHuEPO may limit its use especiallyin less developed countries (69).

POSITIVE EFFECTS OF CORRECTION OF ACD

Anemia is a serious complication of cancer and cancer treat-ment leading to various symptoms and dysfunctions of thehuman body. Correction of anemia not only helps to alleviatesymptoms associated with this condition, but may alsoimprove survival and possibly even response to chemotherapyand radiotherapy (Table 8).

Most clinical trials documenting a positive effect ofcorrection of ACD used rHuEPO to improve anemia-relatedsymptoms such as improvement in fatigue and otherparameters of QOL and to decrease in transfusion need(Tables 9 and 10).

Table 9 ECAS Enrollment Data

n Anemic (%)

Evaluable at enrollment 14027 39Solid tumors 11071 36Breast 3216GI-Colorectal 2469Lung 2057Gynecologic 1702Urogenital 917Head and neck 710Hematological malignancies 2956 53Lymphoma=myeloma 2316Leukemia 640

Table 8 Positive Effects of Correction of ACD

Alleviation of anemia-related symptomsImprovement of quality of lifeDecrease of fatiguePreliminary indications for the improvement of response to chemotherapyPreliminary indications for the improvement of response to radiotherapy


Several studies have demonstrated that rHuEPO therapyincreases hemoglobin levels, decreases transfusion require-ments, and improves QOL in anemic cancer patients (70–72).

There is extensive evidence proving the effectiveness ofrHuEPO in the treatment of anemia in cancer (Table 11).

Ongoing trials focus on improvement of results of cancertreatment, in particular on better response to chemotherapy,radiotherapy, or combined radiochemotherapy.

Initial Clinical Trials in Cancer Patients

Initial clinical trials using rHuEPO were performed onpatients with multiple myeloma or non-Hodgkin’s lymphomaand yielded excellent results (73,74). Response rates, definedas an increase of hemoglobin by at least 2 g=dL,were observed in 83–85% of all patients. The first study eval-uating the efficacy of rHuEPO in treating chemotherapy-induced anemia observed a response rate of 85% in patientsreceiving 200 or 300 U=kg=day (75). The high response ratesthat could be seen in these early pilot studies evoked a multi-tude of phase II studies as well as prospective randomized,partly double-blind, placebo-controlled clinical trials onrHuEPO treatment in cancer patients. The largest studyincluded 413 patients, 68% of whom had solid tumors (76).Patients were grouped according to treatment regimens—nochemotherapy, myelosuppressive noncisplatin-containingchemotherapy, and myelosuppressive cisplatin-containing

Table 10 Patients Not Anemic at Enrollment

Not anemicat enrollment

Became anemicduring cancer

treatment

TreatmentChemotherapy 2101 1317 (63%)Radiotherapy 514 100 (20%)Tumor typeSolid tumors 2262 1228 (54%)Hematological malignancies 297 154 (52%)

510 Pohl and Ludwig

chemotherapy—and randomly assigned to receive either pla-cebo or rHuEPO. In all three groups, patients receivingrHuEPO had a statistically significant increase in hematocritcompared with placebo-treated patients (p< 0.004).

The first clinical trials revealed that rHuEPO treat-ment resulted in a significant increase in hemoglobin inapproximately 50–85% of anemic cancer patients. Responserates seemed to be comparable between anemic cancerpatients without chemotherapy and those on either plati-num or nonplatinum-based treatment protocols. Respondersexperienced significant improvements in performance statusand QOL parameters. rHuEPO treatment was well toler-ated by cancer patients without substantial occurrence ofhypertension.

Clinical Trials in Patients with HematologicalMalignancies

Multiple Myeloma and Non-Hodgkin’s Lymphoma

Ludwig et al. (60) performed a phase II study on 20 anemicmyeloma patients, which showed a response rate of 75% torHuEPO treatment. An identical response rate of 75% wasreported by Barlogie and Beck who investigated 28 myelomapatients who had previously been treated with chemotherapyor autologous bone marrow transplantation (77). Responserates of 71% as well as an improvement in performance statuswere observed by Mittelman et al. (78). In a group of patientswith advanced, transfusion-dependent, and chemoresistantmultiple myeloma, the response rate, defined as completeabolition of transfusion dependency, was 35% (79).

The first randomized trial on rHuEPO treatment of mye-loma associated anemia found a response rate of 78% inrHuEPO-treated patients by the eighth week (80). A trial on71 myeloma patients receiving concomitant chemotherapyreported a response rate of 75% in the rHuEPO arm as com-pared to 21% in the control arm, further a trend towardsreduced transfusional needs under rHuEPO treatment (81).These data were confirmed in a placebo-controlled study of145 patients with multiple myeloma and anemia (82).


Table 11 Clinical Trials Documenting Positive Effects ofCorrection of Anemia in Cancer Patients

Ref. Diagnosis Response criteria N

Ludwig

et al. (73)

MM Hb " �2 g=dL 13

Platanias

et al. (75)

Various tumor

types

Hb " >10% 30

Miller

et al. (118)

Various tumor

types

Hb " �1 g=dL 21

Barlogie and

Beck (77)

MM with

renal failure

Hb " �2 g=dL 28

Ludwig

et al. (60)

MM or SCC Hb " �2 g=dL 20 MM, 14 SCC

Cascinu

et al. (119)

Various tumor

types

Hb �10 g=dL 20

Case

et al. (178)

Various tumor

types

Hkt " >6% 153

Tsukuda

et al. (121)

Head and

neck cancer

Hb " �1 g=dL

hemoglobin

18

Markman

et al. (135)

Ovarian

cancer

Development

of anemia

17

Lavey

et al. (149)

Various tumor

types

Hb " >14 g=dL 40

Rose

et al. (89)

CLL Six times " in Hkt 221

Henry

et al. (76)

Various tumor

types

Six times " in Hkt 413

Cascinu

et al. (51)

Various tumor

types

TI 100

Dusenbery

et al. (148)

Cervical

cancer

Hb " 20

Cazzola

et al. (61)

MM=NHL Median average

increase in

Hb=week

84 MM

512 Pohl and Ludwig

Treatment Therapy

Response

rate Study design

Epoietin alfa

150–250 U=kg tiw

Prior CT: all pts.,

9 pts. CT, or RT

during study

85% Phase II

trial

Epoietin 25–300 U=kg

five times=week i.v.

CT 50% Phase II

trial

Epoietin alfa 100 or

200 U=kg five

times=week iv

Cisplatin-based

CT

57% Phase I–II

trial

Epoietin alfa

150 U=g tiw

Pts. stable on

standard CT

78% Phase II

trial

Epoietin alfa

150–300 U=kg tiw

Prior CT: all pts.,

9 pts. CT, or RT

during study

75% Phase II

trial

Epoietin alfa

50–100 U=kg tiw

Cisplatin-based

CT

75% Phase II

trial

Epoietin alfa

150 U=kg tiw

58% vs.

14%

Randomized

trial

Epoietin alfa 3000

or 6000 U i.v. tiw

RT and=or CT 58% vs.

78% vs.

0%

Randomized

trial

Epoietin alfa

100 U=kg daily

Carboplatin-based

CT

6% Phase II,

Prevention

trial

150–300 U=kg tiw RT 80% vs. 5% Phase II

Epoietin alfa,

150 U=kg tiw

Not stated 47% Double-blind,

randomized

placebo-controlled

trial

Epoietin alfa,

100 or 150 U=kg tiw

No CT vs. CT vs.

cisplatin-based CT

32–58% Double-blind,

randomized

placebo-controlled

trial

Epoietin alfa

100 U=kg tiw

Cisplatin-based CT 80% vs. 44% Double-blind,

randomized

placebo-controlled

trial

Epoietin alfa

200 U=kg daily

RT 2.9 g=dL Phase I and

II trial

Epoietin beta

1000–10,000 U daily

Concomitant CT:

79%

–0.04 g=dL

1000 U,

0.22 g=dL

2000 U,

0.43 g=dL

5000 U,

0.58 g=dL

10,000 U

Randomized,

controlled,

multicenter

trial

(Continued)

513

Table 11 Clinical Trials Documenting Positive Effects ofCorrection of Anemia in Cancer Patients (Continued)


Silvestris

et al. (80)

MM Hb " 44

Ludwig

et al. (117)

Various tumor

types

Hb " �2 g=dL 102

DeCampos

et al. (179)

SCLC Total red blood

cell transfusions

36

Tsuji

et al. (153)

Gastric cancer Hb level, day 10 10

Osterborg

et al. (83)

MM=NHL No transfusions

over 8 weeks

or Hb " >2 g=dL

121, 65 MM

Porter

et al. (125)

Pediatric solid

tumors

Transfusion need 24

Wurnig

et al. (126)

Bone tumors Transfusion need 30

Mittelman

et al. (78)

MM Hb", TN 17

DelMastro

et al. (136)

Breast cancer Hb decrease 62

Pawlicki

et al. (123)

Various tumor

types

Hb " >2 g=dL 215

Kasper

et al. (122)

Various tumor

types

Hb " >2 g=dL 48

Kurz

et al. (127)

Gynecologic

tumor

TI 35

Braga

et al. (154)

Gastric or colorectal

cancer

TI 20

Glaspy

et al. (50)

Various tumor

types

Hb " 2342

514 Pohl and Ludwig

Treatment Therapy

Response

rate Study design

Epoietin alfa

150–300 U=kg tiw

CT 78% Randomized

trial

Epoietin alfa

150–300 U=kg tiw

69% CT 62% vs.

52%

CT-pts.

Phase II

trial

Epoietin alfa

150 or 300 U=kg tiw

Carboplatin-based

CT

54 U,

53 U vs.

116 U

Randomized

prevention

trial

200 U=kg daily i.v. Surgery, distal

gastrectomy

14 vs.

10.8 g=dL

Randomized

trial

Epoietin beta

10,000 U daily

Prior CT and

concomitant:

59 pts.,

6 pts. not

60% Randomized

trial

Epoietin alfa

150–300 U=kg tiw

CT 23 vs.

80 mL=kg

Double-blind,

randomized

placebo-controlled

trial

Epoietin alfa

600 U=kg two

times=week i.v.

CT 2.1 vs.

8.4 U RBC

Double-blind,

randomized

prospective

placebo-

controlled

phase III trial

Epoietin alfa

150–300 U=kg tiw

CT 70.6% Phase II trial

Epoietin alfa

150 U=kg tiw

CT 0.8 vs. 3.1 g=dL Randomized

prevention trial

Epoietin alfa

150 U=kg tiw

CT 67% Phase II trial

Epoietin

2000–10,000 U

daily

49% Phase II trial

Epoietin alfa

150 U=kg tiw

CT 78% vs. 33% Double-blind,

randomized

placebo-

controlled,

prospective

multicenter trial

Epoietin alfa

500 U=kg

Surgery 10% vs. 50% Randomized

trial

Epoietin alfa

150–300 U=kg tiw

CT 1.8 g=dL Community-based

trial

(Continued)

515



Tsukuda

et al. (128)

Head and

neck cancer

" �1 g=dL Hb 22

Dammacco

(81)

MM refractory

to CT

Absolute Hb "and=or reduction

in transfusion

71

Oberhoff

et al. (124)

Various tumor

types

TI 227

Sweeney

et al. (150)

Various tumor

types

Hb: 14 or 15 g=dL 48

Kettelhack

et al. (155)

Colorectal

cancer

Preoperative transfusion 102

Rau

et al. (156)

Colorectal

cancer

>3 U autologous blood 54

Frommhold

et al. (147)

Head and

neck cancer

Hb " 50

Demetri

et al. (52)

Various tumor

types

Hb " �1.2 g=dL 2370

Bokkel Huinink

et al. (29)

Ovarian

cancer

>1 RBC transfusion 122

Thatcher

et al. (137)

SCLC Hb >10 g=dL, TI 130

Dunphy

et al. (138)

Head and

neck cancer

Transfusion dependence 30

Siakantaris

et al. (90)

CLL Six times " in Hkt

or Hkt of 0.38

22

Littlewood

et al. (174)

Various tumor

types

Hb ", survival 375

Glaser

et al. (151)

Oral SCC Hb cutoff: 12.5 g=dL,

residual cancer at

surgery, 17-month local

control rate

37

Gabrilove

et al. (131)

Various tumor

types

Hb " 3012

Dammacco MM Hb " �2 g=dL 145

516 Pohl and Ludwig

Treatment Therapy

Response

rate Study design

Epoietin alfa 100,

200, or 400 U=kg

weekly

Platinum-based

CT

54% vs. 0% Randomized

trial

Epoietin alfa

150–300 U=kg tiw

CT 75% Randomized

trial

Epoietin alfa

5000 U daily

CT 72% vs. 58% Randomized

trial

Epoetin alfa

200 U=kg daily

RT 42% vs. 0% Randomized

trial

Epoetin beta

20,000 U daily

Surgery 33% vs. 28% Randomized

multicenter

trial

Epoetin alfa

200 U=kg daily

Surgery 71% vs. 42% Randomized trial

Epoetin alfa

150–300 U=kg tiw

RT 3.2 vs. 0.7 g=dL Phase II trial

10,000–20,000 U tiw CT 53% Randomized

multicenter trial

150–300 U=kg tiw CT 9% vs. 39% Randomized

prevention trial

Epoetin alfa

150 or 300 U=kg tiw

Platinum-based

CT

48%, 20% vs.

66%; 45%,

20% vs. 59%

Randomized

prevention trial

Epoetin alfa

150–450 U=kg tiw

Carboplatin-based

CT

15% vs. 36% Randomized

prevention trial

Epoietin alfa 12 pts. concomitant

CT

77% Phase II trial

Epoietin alfa

150–300 U=kg tiw

CT 2.2 vs.

0.5 g=dL;

17 vs.

11 months

Double-blind,

randomized

placebo-controlled

multicenter

trial

Epoetin alfa

150 U=kg tiw

Neoadjuvant

CT

32% vs. 73%;

89% vs. 70%

Phase II trial

Epoetin alfa

400,000–60,000 U

weekly

CT 2.2 g=dL Multicenter trial

Epoietin alfa Prior CT: 58% Double-blind,

(Continued)

517



(82)

Kunikane

et al. (139)

NSCLC Hb " 72

Quirt

et al. (134)

Cancer related

anemia

Hb " 401

Vansteenkiste

et al. (180)

Lung cancer Hb ", TI, proportionof pts. transfused

297

Scott

et al. (157)

Head and neck

cancer

Transfusion

requirements

58

Olson

et al. (181)

Metastatic breast

cancer

Hb " �20 g=L 180

Glaspy

et al. (182)

Solid tumors Hb ", TI 4298

Osterborg

et al. (84)

Hematologic

malignancies

(NHL, CLL, MM)

Hb " �2g=dL, TI 106 dþ126þ117

Bamias

et al. (140)

Solid tumors Transfusion need,

Hb >10 g=dL

144

Hedenus

et al. (86,88)

Lymphoproliferative

malignancies

Hb " �2 g=dL, TI,

proportion

of pts. Transfused

66

Kosmadakis

et al. (158)

Gastrointestinal

tract cancer

Transfusion need,

Hb, Hkt

31

Abbreviations: Hb: hemoglobin, Hkt: hematocrit, CT: chemotherapy, RT: radiotherapy,Pts: patients, tiw: three times per week, pw: per week, i.v.: intravenously,TI: transfusion independence, MM: multiple myeloma, SCC: squamous cell carcinoma,SCLC: small cell lung cancer, NHL: Non-Hodgkin lymphoma, CLL: chronic lymphatic leukemia.

518 Pohl and Ludwig

Treatment Therapy

Response

rate Study design

150–300 U=kg tiw all pts. placebo-controlled

Epoietin alfa 100 ug

vs. 200 ug=kg tiw

CT Significant Hb " Double-blind,

randomized

placebo-controlled

multicenter trial

Epoietin alfa

150 U=kg tiw

218 pts. CT Significant Hb " Phase II multicenter

trial

Darbepoetin alfa

vs. placebo


placebo-controlled,

randomized multicenter

phase III trial

Epoetin alfa

600 UI=kg

three times

before surgery

Surgery 35% vs. 17% Prospective,

randomized,

placebo-controlled

trial

Epoetin beta

1000 IE or

5000 IE s.c. tiw

CT " Hb: 23 and 17 g Randomized trial

Epoetin alfa

150–300 U=kg

tiw or

10000–20,000 U tiw

CT Decrease in

transfusion need,

Hb " 1.6–2 g=dL

Open-label,

randomized,

community-based trial

Epoietin beta

150 U=kg tiw


placebo-controlled,

randomized

phase III trial

Epoietin alfa

10,000 U tiw

Platinum-based

CT

15.3% vs. 33%;

16.6% vs. 45.8%

Prospective,

randomized,

placebo-controlled

trial

Darbeopoietin

alfa, 1, 2.25, or

4.5 mcg=kg=week

CT 45% vs. 55%

vs. 62% vs. 10%

Double-blind,

randomized

placebo-controlled

trial

Epoetin alfa

300 U=kg tiw

plus iron i.v.

Surgery Prospective,

randomized,

double-blind

trial

519

rHuEPO treatment significantly decreased the incidence oftransfusion compared with placebo and increased mean Hb.

A randomized trial by Cazzola et al. (61) on 146 anemicpatients without transfusion need at the time of enrolmentincluding myeloma and NHL patients found response ratesto rHuEPO treatment of 62%. Another large randomized mul-ticenter trial was performed on transfusion-dependent ane-mic patients with multiple myeloma (83). The same groupinvestigated transfusion-dependent patients with NHL(n¼ 106), CLL (n¼ 126), or MM (n¼ 117) and a low serumerythropoietin in a randomized trial (84). Patients receivedeither epoetin beta 150 IU=kg or placebo subcutaneouslythree times a week for 16 weeks. The response rate was67% and 27% in the epoetin beta vs. the placebo group,respectively (p< 0.0001).

Recently, it was found that several anemic multiplemyeloma patients treated with epoietin alfa achieved anunexpectedly longer survival rate (85).

A randomized, double-blind, placebo-controlled trialshowed that darbepoetin alfa effectively increased hemoglo-bin concentrations in patients with lymphoproliferative disor-ders (86). Another recent trial revealed that epoetin betaadministered once weekly is an effective and convenienttreatment for anemia in patients with either multiplemyeloma, low-grade non-Hodgkin’s lymphoma, or chroniclymphocytic leukemia (87,88).

Chronic Lymphatic Leukemia

In a prospective trial, patients with chronic lymphatic leuke-mia received either rHuEPO treatment or placebo for 12weeks with a response rate of 50% in the rHuEPO treatmentarm. In addition, patients responsive to rHuEPO treatmentshowed significant improvements in energy, self-rated health,and several other QOL parameters (89). A response rate of50% was observed in a randomized trial in anemic CLLpatients (90). Improvement of hemoglobin by erythropoietintreatment may even allow downstaging of CLL. This hasrecently been shown in 25 patients with stage III and IV

520 Pohl and Ludwig

CLL. Twenty patients achieved an increase in hemoglobinand hence a lower stage after erythropoietin treatment. Theauthors also indicated the excellent survival seen in thispatient group (91).

Myelodysplastic Syndrome

Most patients with myelodysplastic syndrome (MDS) aretransfusion-dependent and could also be considered candi-dates for rHuEPO treatment. The use of rHuEPO alone is suc-cessful only in a minority of patients (92). Clinical trialsevaluating the effect of rHuEPO treatment in patients withMDS yielded different results regarding outcome of treatment.These outcomes may be due to different patient selection,various treatment regimens as well as definition of responsecriteria.

Very often, patients with MDS show very high levels ofendogenous erythropoietin. Therefore, it is not surprisingthat rHuEPO treatment of MDS-associated anemia resultedin response rates between 10% and 38% (93–101). A meta-analysis of 17 trials involving a total of 205 MDS patientsfound significant responses to rHuEPO monotherapy in 16%of patients (102). The same authors reported that responserates of patients with MDS can be enhanced by combiningrHuEPO with myeloid growth factors, such as G-CSF indicat-ing a synergistic effect (103). They found an overall responserate of 38% and a response rate of 60% in patients with ringsideroblasts. A series of subsequent studies has confirmedtheir positive results, but their response rates varied tremen-dously (0–80%) (104–107). Other trials combined rHuEPOwith GM-CSF resulting in response rates between 25% and53% (108–110). Combination of rHuEPO with IL-3 (111) or13-cis-retinoic acid and alpha-tocopherol (112) resulted in21% and 35% response rates, respectively. A recent prospec-tive randomized trial in 66 patients compared combinationtherapy with rHuEPO and GM-CSF with GM-CSF plusplacebo. The response rated was only 9% in the combinationtherapy arm as compared to 5% in the GM-CSF=placeboarm (113).


Terpos et al. (114) investigated the potential advantageof a prolonged administration of rHuEPO to achieve highererythroid response rates (RR) in 281 MDS patients. Responseto treatment was evaluated after 12 and 26 weeks of therapy.The overall response rate was 45.1%. These results suggestthat prolonged administration of rHuEPO produces highand long-lasting erythroid response rates in MDS patients.

Recently, Hellstrom-Lindberg et al. (115) developed adecision model for anemic patients with MDS, in which trans-fusion need and serum erythropoietin were used to definethree groups with different probabilities of erythroid responseto treatment with G-CSF and rHuEPO. They included 53patients from a prospective study in an evaluation on treat-ment outcome. Patients with good or intermediate probabilityof response were treated with G-CSF and rHuEPO. Theresponse rates were 61% and 14% in the good and in the inter-mediate predictive groups, respectively. Hence, this validateddecision model for treating the anemia of MDS may be usefulto optimize treatment for patients with this disease.

Currently, darbepoetin alfa therapy has been initiated inpatients with MDS (116). Preliminary results suggested thatdarbepoetin alfa administered twice a week is effective in cor-recting anemia and maintaining Hb in patients with MDS.

Despite wide variations in outcomes of studies on combinedrHuEPO=myeloid growth factor therapy, which mainly seem toresult from differences in patient characteristics, treatmentregimes and response criteria, response rates are generallyhigher than those of rHuEPO monotherapy (Table 12).

Unfortunately, patients dependent to transfusions (�2red blood cell transfusions per month) are less likely toresponse even to combined growth factor therapy (92). Withregard to this, treatment with rHuEPO either with or withoutG-CSF cannot be regarded as standard treatment for MDS atthe moment.

Clinical Trials in Patients with Solid Tumors

Most clinical trials on rHuEPO treatment in cancer patientswith solid tumors cover a wide range of tumor types and often

522 Pohl and Ludwig

state their results globally. Hence, it is difficult to evaluatethe efficacy of rHuEPO treatment according to tumor type.In addition, most trials do not include only newly diagnosedcancer patients. Additional data are required to determinewhether initiating rHuEPO earlier spares more patients fromtransfusions or result in better QOL than waiting until hemo-globin levels concentrations decline to nearly 10 g=dL.

The effect of concomitant chemotherapy was investigatedin 94 anemic cancer patients by comparing rHuEPO respon-siveness in 68 patients on chemotherapy vs. 26 patientswithout chemotherapy. Response was achieved in 52% ofthe patients receiving chemotherapy vs. 62% in the treatedgroup (117).

Response rates between 39% and 79% were described inseveral phase II studies on rHuEPO treatment in anemicpatients with various types of solid tumors (118–123). Severalrandomized, double-blind, placebo-controlled trials followed.Cascinu et al. (51) investigated 100 patients with cisplatin-associated anemia and reported significant increases in hemo-globin levels in their rHuEPO arm from the third treatmentweek onward. Response rates between 32% and 58% wereseen in a double-blind and open-label follow-up study inpatients receiving either no chemotherapy or cisplatin-containing chemotherapy or chemotherapy that not includedcisplatin (76). The daily use of 5000 U rHuEPO subcuta-neously led to a significant reduction in the need for bloodtransfusions in the rHuEPO arm compared to untreated con-trols in a large prospective, randomized trial on 227 patientswith solid tumors and chemotherapy-induced anemia (124).Several other randomized trials showed significant higherresponse rates in the treatment group when compared tothe placebo group (125–128). A meta-analysis of eight rando-mized, placebo-controlled trials demonstrated a clear benefitin terms of reduction of transfusion requirement for the useof rHuEPO in patients undergoing chemotherapy for a solidtumor (129). The relative risk for transfusion among EPOpatients was 0.64, which translated into a 36% relative reduc-tion in the proportion of patients requiring transfusion(p¼ 0.00001). Reduction in transfusion requirements was


Table

12

ClinicalTrials

Documen

tingPositiveEffects

ofCorrectionof

Anem

iain

Patien

tswithMDS

Ref.

Number

rHuEPO

dose

Com

bined

with

Respon

secriterion

Respon

serate

(%)

Phase

IIstudies

Herrm

ann

etal.(1991)

19

10,000U

five

times=wee

kly

—

Hellstrom

etal.(96)

12

600–3000U

wee

kly

i.v.

—þ�1.5

g=dL

Hb

25

Kurzrock

etal.(1991)

16

50U=kg

daily

—TI

13

Ludwig

etal.(98)

10

150–300U=kg

tiw

—þ�2g=dL

hem

oglobin

10

Stoneet

al.(99)

20

100–400U=kg

tiw

—þ�1.2

g=dL

Hb

35

Roseet

al.(100)

100

150–300U=kg

tiw

—>6%

hem

atocrit

28

Bernellet

al.(101)

30

10,000U

three

times=wee

kly

—Hb"

38

Hellstrom

-Lindberg

etal.(103)

21

60–300U=kg

daily

G-C

SF

þ�1.5

g=dL

Hb

38

Imamura

etal.(104)

10

100–400U=kg

daily

G-C

SF

þ�1g=dL

Hb

0

Neg

rin

etal.(105)

44

100U=kg

daily

G-C

SF

þ�2g=dL

hem

oglobin

48

524 Pohl and Ludwig

Rem

ach

aet

al.(106)

32

300U=kg

tiw

G-C

SF

þ�1g=dL

hem

oglobin

50

Mantovani

etal.(107)

33

200–400U=kg

daily

G-C

SF

þ�2g=dL

hem

oglobin

orTI

80

Runde

etal.(108)

10

100–200U=kg

daily

GM-C

SF

TI

25

Stasi

etal.(109)

26

150–300U=kg

everysecond

day

GM-C

SF

þ�1g=dL

hem

oglobin

35

Econom

opou

los

etal.(110)

19

60–120U=kg

tiw

GM-C

SF

þ�2g=dL

hem

oglobin

53

Miller

etal.(111)

22

150–300U=kg

tiw

IL-3

þ�2%

reticu

locytes

21

Besa

etal.(112)

23

150U=kg

tiw

CRA,AT

þ�2g=dL

hem

oglobin

35

Terpos

etal.(114)

281

150U=kgtiw,

�26wee

ks

—Erythroid

resp

onse

rates

45.1

Random

ised

trials

Stein

etal.(1991)

21

1200–1600

U=kg

twotimes=

wee

ki.v.

—þ>4%

Hktor

TI

13vs.

0

Italianstudygroup(1998)

87

150U=kgdaily

—37vs.

11

Thom

psonet

al.(113)

66

150U=kgtiw

GM-C

SF

þ�2g=dL

hem

oglobin

9vs.

5

Abbreviation

s:Hb:hem

oglobin,Hkt:hem

atocrite,

CT:ch

emotherapy,RT:radiotherapy,Pts:patien

ts,tiw:threetimes

per

wee

k,pw:per

wee

k,i.v.:intraven

ously,TI:

transfusion

indep

enden

ce,G-C

SF:granulocyte-stimulatingfactor,GM-C

SF:granulocyte-m

acrop

hage-sti-

mulatingfactor.CRA:13-cis-retinoicacid,andAT:alpha-tocop

herol.


similar across strata defined by methodological quality, EPOdose, hematologic status, tumor type at trial entry, andchemotherapy regimen.

Two major community-based trials suggested the clinicalefficacy or rHuEPO treatment in almost all tumor types in2342 cancer patients receiving chemotherapy (50). Demetriet al. (52) found similar outcomes from 2730 patients withnonmyeloid malignancies treated with rHuEPO.

In addition to significant improvements in hemoglobinlevels and QOL parameters, a possible relationship of hemo-globin levels with survival in anemic cancer patients receiv-ing chemotherapy was seen in 375 patients with solid ornonmyeloid hematological malignancies, suggesting a survi-val benefit in the rHuEPO treatment arm (130).

A large randomized multicenter trial investigated a totalof 3012 patients with nonmyeloid malignancies receivingchemotherapy. Patients received epoetin alfa 40,000U onceweekly, which could be increased to 60,000U once weeklyafter 4 weeks, depending on hemoglobin response (131). Theresults from this large, prospective, community-based trialsuggested that once weekly epoetin alfa therapy increaseshemoglobin levels, decreases transfusion requirements, andimproves QOL in patients with cancer and anemia whoundergo concomitant chemotherapy.

A retrospective subset analysis was performed on 244anemic colorectal cancer patients (132). This analysis demon-strated that epoetin alfa 40,000 U s.c. weekly significantlyincreased Hb and improved QOL in anemic colorectal cancerpatients. Another retrospective subset analysis on this trialevaluated Hb response, transfusion use, and QOL improve-ment in 290 anemic ovarian cancer patients (133). The recentsubset analysis revealed that epoetin alfa 40,000U s.c. weeklysignificantly increased Hb levels, decreased transfusion use,and improved QOL in anemic ovarian cancer patients, withefficacy and tolerability similar to that reported in thecomplete study population.

Quirt et al. (134) performed a prospective open-labelstudy, designed to evaluate the efficacy of rHuEPO in 401patients with cancer-related anemia including 183 patients

526 Pohl and Ludwig

not receiving chemotherapy. Treatment with rHuEPO signif-icantly increased hemoglobin levels and reduced transfusionrequirements.

Even though some authors reported minor differ-ences among response rates in various tumor types, anemicpatients with solid tumors generally benefit from rHuEPOtreatment.

Prevention of Anemia with ErythropoietinTreatment

Encouraging response rates of these former trials led to theinitiation of clinical trials applying rHuEPO prophylactically.A significant difference in hemoglobin nadirs was seen inpatients with ovarian cancer on second-line carboplatin-basedchemotherapy (135). A large randomized multicenter trial in122 patients with ovarian cancer evaluated the effectivenessof rHuEPO treatment and found a lower rate of transfusionneed for patients receiving rHuEPO (29). Similar results wereobtained in a smaller study on 62 patients (136).

In patients with small cell lung cancer, the efficacy ofepoietin alfa in preventing the decline in hemoglobin levelsand in reducing transfusion was demonstrated in SCLCpatients receiving 4–6 cycles of primarily platinum-based che-motherapy (137). A randomized trial with or without erythro-poietin during chemotherapy observed less anemia and fewertransfusions in patients randomized to receive erythropoietinconcurrently with paclitaxel and carboplatin (138). The impactof rHuEPO on the development of anemia during cisplatin-based chemotherapy was also studied in 72 nonsmall-cell lungcancer patients (139). Patients were randomized to receive 100IU=kg rHuEPO, 200 IU=kg rHuEPO, or placebo three times aweek. Patients receiving rHuEPO had significantly elevatedhemoglobin levels after the second chemotherapy cycle as com-pared to the placebo group.

One hundred and forty-four patients with solid tumorsreceiving platinum-based chemotherapy were randomized toreceive either 10,000 U of rHuEPO thrice weekly s.c. or notreatment (140). rHuEPO at a dose of 10,000 U thrice weekly


prevented transfusions and development of significantanemia in those patients.

Sixty-two early-stage breast cancer patients undergoingaccelerated adjuvant chemotherapy were randomized toreceive rHuEPO 150 U=kg three times a week or no treat-ment, and prophylactic rHuEPO treatment proved to be effec-tive (136). Significant improvements in Hb levels were seen inpatients with metastatic breast cancer receiving rHuEPO.Another recent trial evaluating the efficacy of rHuEPO treat-ment demonstrated that anemia treatment improves cogni-tive function in breast cancer patients (141). Recently, itwas shown that rHuEPO treatment maintains hemoglobinand QOL in breast cancer patients receiving conventionaladjuvant chemotherapy (142).

Another trial on patients with advanced nonsmall-cell lungcancer receiving first-line chemotherapy randomized 216patients to get either rHuEPO 40,000 U s.c. once weekly or noEPO treatment at all. Hemoglobin was maintained throughoutthe study in the rHuEPO group, whereas Hb decreased relativeto baseline in the control group (143).

Prevention of chemotherapy-induced anemia facilitatescancer treatment with the full projected dose of chemothera-peutic agents and may intrinsically foster the outcome of localand systemic treatment.

Clinical Trials in Patients UndergoingRadiotherapy

It has been known for a long time that hypoxia which is clo-sely associated with anemia decreases the tumors sensitivityto radiotherapy (144). Several studies produced evidencethat anemia influences the outcome of radiotherapy treat-ment (145,146). A retrospective trial including data from889 patients homogeneously irradiated for head and neckcancer identified hemoglobin to be an independent and power-ful predictor of outcome of radiotherapy including locoregio-nal tumor control and survival (147). Consequently, trialswere initiated to evaluate whether erythropoietin treatmentimproves hemoglobin levels during radiotherapy and to inves-

528 Pohl and Ludwig

tigate whether increased hemoglobin levels improve outcomeof radiotherapy.

In 20 patients with cervical cancer undergoing radiother-apy, rHuEPO treatment led to an increase in hemoglobinlevels (148). In another phase II trial including 40 patientswith tumors above the diaphragm and without evidence ofdistant metastasis, all patients receiving rHuEPO treatmentresponded with increases in hematocrit by 6% during radia-tion therapy. Only 6% of the patients in the control groupresponded with increase in hematocrit (149).

An open-label trial including 48 patients with carcinomaof the lung, uterine cervix, prostate, or breast who receivedradiotherapy randomized patients to rHuEPO treatment vs.no rHuEPO treatment. Forty-two percent of all patientsreceiving rHuEPO treatment showed a steady increase inhemoglobin levels and achieved normal hemoglobin levels ascompared to 0% in the control arm (150).

In a retrospective comparison, Glaser et al. (151) showedsuperior local control rate and overall survival in anemicpatients with squamous cell carcinoma of the oral cavityand oropharynx treated with erythropoietin in addition tocombined radiochemotherapy compared to a historical controlgroup without erythropoietin treatment. Thus, rHuEPOtreatment proved to be effective both in treatment andprevention of anemia in patients undergoing radiotherapy.

Two hundred and fifty-six cervical cancer patients with atleast one high risk characteristic were randomized to receiveeither rHuEPO or transfusions only, if required during adju-vant chemo-radiotherapy. Sequential adjuvant chemo-radio-therapy proved to be an effective and well-tolerated approachin high risk cervical cancer, with activity presumably enhan-ced by rHuEPO support including reduction in transfusionneed, increase of Hb levels and survival (152).

Clinical Trials in Surgical Cancer Patients

In a very small trial on 10 patients with gastric cancer under-going distal gastrectomy, rHuEPO treatment prevented post-


operative anemia as judged by hematocrit, hemoglobin concen-tration, and red blood cell count (153). Braga et al. (154) foundan erythropoietic response induced by rHuEPO in anemic can-cer patients who were submitted to major abdominal surgery.Another randomized, double-blind, placebo-controlled trialinvestigated patients with moderate anemia and right-sidedcolon cancer who were scheduled for hemicolectomy and wererandomized to receive rHuEPO or placebo preoperatively. Thispreoperative treatment protocol resulted in significant increasesin reticulocyte count in the treatment arm prior tosurgery but failed to reduce intraoperative and postoperativetransfusion needs (155).

Rau et al. studied the potential of erythropoietin therapyfor increasing the volume of autologous red blood cell dona-tion in a prospective randomized trial in 54 patients withcolorectal cancer. Seventy-one percent of patients inthe rHuEPO arm were able to donate more than three unitsof blood, as compared to 42% of the patients in the controlarm (156).

Fifty-eight patients undergoing surgical resection ofhead and neck tumors were randomized to receive three dosesof rHuEPO or placebo before undergoing surgery (157). Asignificant improvement in hematopoietic parameters anda trend towards decreased transfusion requirements usingpreoperative rHuEPO could be demonstrated.

The effect of rHuEPO administration on preoperativehemoglobin concentrations and on the number of blood trans-fusions was analyzed in patients undergoing surgery forgastrointestinal tract malignancies; 63 patients were rando-mized to either receive subcutaneous rHuEPO in a dose of300 IU=kg body weight plus 100mg iron intravenously(n¼ 31) or placebo medication and iron (158). Patients whoreceived erythropoietin received significantly fewer transfu-sions intraoperatively and postoperatively. Postoperatively,the study group had significantly higher hematocrit, hemoglo-bin, and reticulocyte count values compared to the controlgroup. The use of erythropoietin was also associated with areduced number of postoperative complications and improved1-year survival.

530 Pohl and Ludwig

Pre- and postoperative treatment with rHuEPO appearsto be a promising way of improving anemia and reducingtransfusion need in surgical cancer patients.

Clinical Trials on Patients Receiving Bone MarrowTransplantation

Laboratory studies have shown that serum erythropoietinlevels after allogeneic but not autologous bone marrow trans-plantation are suboptimal for the degree of anemia (159).Endogenous erythropoietin may be suppressed by the under-lying malignancy and chemotherapy, by cyclosporine A,amphotericin B, and by the release of cytokines such as tumornecrosis factor. Randomized studies of rHuEPO after allo-geneic and autologous BMT have been conducted. Most ofthe trials showed that treatment with rHuEPO speeds thetime to red cell transfusion independence and reduces thenumber of red cell transfusions needed after allogeneic trans-plantation (160,161), other studies remained controversial(162,163).

In autologous bone marrow transplantation, the resultsof erythropoietin substitution were disappointing. A phaseII trial in 18 patients who were treated with a combinationof rHuEPO and rHuGM-CSF showed a tendency towardsmore rapid neutrophil recovery but no apparent impact onred cell transfusion requirements (162).

No reduction of transfusions was seen in 35 lymphomapatients treated with rHuEPO (163).

These data underline the activity of rHuEPO treatmentin patients undergoing allogeneic bone marrow transplanta-tion but fails to demonstrate a substantial effect in patientsafter autologous stem cell transplantation.

Clinical Trials and the Impact of AnemiaTreatment on QOL

Several clinical trials describe an improvement of QOL due tocorrection of anemia. Frequently used instruments for eva-


luation of QOL are multidimensional instruments such asEORTCAn, LASA, and the Functional Assessment of CancerTherapy-Anemia (FACT-Anemia) subscales. A study by Roseet al. (89) demonstrated an improvement of QOL in patientstreated with rHuEPO who obtained an increase in hemato-crit. Two other large studies addressed QOL showing that itimproves in association with a rise in Hb (50,52). Demetriet al. further showed that the improvement in QOL, whichoccurred as a function of a rise in Hb, seemed to occurindependently of tumor response.

A large, randomized, placebo-controlled, double-blind trialof rHuEPO in patients with either solid tumors or hematologi-cal malignancies confirmed that treatment with rHuEPO notonly increased the Hb and reduced transfusion need, but alsoled to a significant improvement in QOL in the rHuEPO-treated group as compared to the placebo-treated group (71).

A randomized prospective trial was performed to evalu-ate the use of rHuEPO in cachectic patients with solid tumorsnot receiving chemotherapy to determine if increasing hemo-globin (Hb) resulted in increased exercise capacity, metabo-lism, and energy efficiency during a maximum work load.The randomized, prospective study included 108 patientswho received oral indomethacin 50mg twice daily (n¼ 58;control group), or oral indomethacin 50mg twice daily withepoetin alfa 4000–10,000 IU by subcutaneous injection threetimes weekly (n¼ 50; study group). The study group showedsignificantly greater mean exercise capacity, mean oxygenuptake, mean CO2 production, and respiration. These resultsdemonstrated that early use of epoetin alfa prevents anemiain patients with progressive cancer who are not receiving che-motherapy. Normalization of Hb levels resulted in improvedwhole-body metabolism and energy efficiency, which wasassociated with greater exercise capacity and better dailyQOL (164).

However, more recently, a further critical review of 13rHuEPO studies was conducted by Bottomley et al. with aparticular emphasis on the quality of live issues (165). Thisreview showed that although some studies suggested animprovement in the patients’ QOL after treatment with

532 Pohl and Ludwig

rHuEPO, there were many methodological limitations in sev-eral of the studies, which made complete interpretation of thedata difficult. Improvements in QOL can only be expected inpatients with a substantial increase in Hb and not in thosewithout an erythropoietic response. As this is usually the casein 30–40% of patients in any study, any trial aiming to assesschanges in QOLmust be powered accordingly to lead to mean-ingful results. Alternatively, one could only compare changesin QOL in patients responding to erythropoietin with QOL innonresponders in untreated controls.

The incremental gain in QOL has been shown to be high-est in patients whose hemoglobin increased from 10 to 12g=dL.With further increases in Hb, much less pronounced increasesin QOL are seen. These considerations may suggest that largerand more robustly designed and adequately evaluatedstudies are needed for definite evaluation of the magnitudeQOL improvement in cancer patients by erythropoietintreatment.

Clinical Trials and the Impact of AnemiaTreatment and Outcome on Cancer Treatment

Anemia correlates with several impediments leading to worseoutcomes in response to anticancer treatment. Hemoglobinlevel is an important prognostic factor for treatment outcomein chemotherapy (166).

Anemia correlates with tumor hypoxia which decreasesradiosensitivity (167), stimulates angiogenesis (168), inducesclonal selection of more aggressive tumor phenotypes (169),and enhances multidrug resistance (170). Consistent with theprognostic significance of anemia, survival of patients whohad required blood transfusion after chemotherapy was signifi-cantly shorter than that of patients who had not (171).

However, it remains to be proven whether anemia is asurrogate variable or a causal factor for disease outcome.

A comprehensive literature review on the association ofanemia and survival revealed that anemia is a strong predictorof poorer survival in cancer patients (172). As discussed above,anemia may have a negative impact on the outcome of radio-


therapy and chemotherapy in various solid tumors. In carc-inomas of both the uterine cervix and the head and neck, tumorhypoxia and anemia were associated with poor local control andoverall survival for patients undergoing radiotherapy (173).

The previously already mentioned randomized, placebo-controlled trial in 375 patients with solid tumors ofhematological malignancies on concurrent nonplatinum-based chemotherapy found a tendency towards increased sur-vival in the rHuEPO arm as compared to the placebo arm(174). As mentioned before, a retrospective comparison of ane-mic patients with squamous cell cancer of head and neck hada significantly higher response rate and overall survival whentreated with erythropoietin during combined chemo-radio-therapy compared to a historical control group, which didnot receive erythropoietin for anemia therapy (151).

These results indicate that anemia is not only a surro-gate marker for a more advanced and aggressive malignancyand poor prognosis, but that the negative effects of anemia onoutcome of cancer treatment possibly can be overcome bysuccessful treatment with erythropoietin. Several studiesdesigned to test this hypothesis in detail are presentlyongoing in various cancers and in several treatmentsituations such as neoadjuvant and adjuvant chemo-therapy, palliative chemotherapy, chemo-radiotherapy, andradiotherapy. Results are eagerly awaited in order to expandour knowledge about the true impact of erythropoietintherapy.

Controversial results were found by a recent studyinvestigating the use of erythropoietin for 12 months as anadjunct to chemotherapy to prevent anemia in patients withmetastatic breast cancer receiving first-line chemotherapy(175). Patients with hemoglobin levels �13g=dL wereincluded in this trial; the goal of treatment was to keephemoglobin concentration within the normal range. Aimsof the study were to assess the effect of rHuEPO treatmenton survival. The study had to be terminated early because ofan observed higher mortality in the group treated witherythropoietin, although subsequent follow-up showed aconvergence of the survival curves at 19 months. The

534 Pohl and Ludwig

observed difference in the number of early deaths wasmainly due to an increase in incidence of disease progressionas well as an increase of thrombotic and vascular events inthe rHuEPO group. Although the benefits of rHuEPO arewell established in its approved applications, the use of itshould be considered only in the context of well-designedclinical trials.

CONCLUSION

Anemia is a common complication in cancer patients, particu-larly in those on cancer therapy and uncontrolled disease andcauses a broad spectrum of symptoms, which may vary fromnegligible to life threatening. Anemia is a negative prognosticfactor associated with shortened survival. Low hemoglobinlevels correlate with tumor hypoxia, and hypoxia induces amore malignant tumor phenotype, radio- and chemoresis-tance and enhances angiogenesis. These sequels probablycontribute to the reduced efficacy of cancer therapy seen incancer patients.

Correction of anemia in cancer patients with recom-binant erythropoietin leads to increased hemoglobin levels,reduction of transfusion needs, improved QOL, and exercisecapacity and possibly also to enhanced responsiveness oftumors to antineoplastic therapy, higher response rates, andeventually increased survival. Data to substantiate the latterconsiderations are yet scarce and carefully designedand conducted studies are needed to either prove or discardthese assumptions. Independently, on the outcome of thesestudies, it seems already fair to state that at the time beingno other drug in recent medical history has contributed somuch to the well being of anemic patients with renal diseases,solid tumors, and hematologic malignancies than erythro-poietin.

Continued research in the field of ACD and discussionand understanding of the impact that even mild-to-moderateanemia can have on patient outcomes will contribute to theoptimization of management of ACD.


ACKNOWLEDGMENT

Supported by the Austrian Forum Against Cancer.

REFERENCES

1. Koeller JM. Clinical guidelines for the treatment of cancer-related anemia. Pharmacotherapy 1998; 18:156–169.

2. Means RT Jr. Advances in the anemia of chronic disease. IntJ Hematol 1999; 70:7–12.

3. Ludwig H. Epoetin in cancer-related anaemia. Nephrol DialTransplant 1999; 14(suppl 2):85–92.

4. Erslev AJ. Erythropoietin and anemia of cancer. Eur JHaematol 2000; 64:353–358.

5. Ludwig H, Fritz E. Anemia of cancer patients: patient selec-tion and patient stratification for epoetin treatment. SeminOncol 1998; 25:35–38.

6. Smith RE Jr, Tchekmedyian S. Practitioners’ practical modelfor managing cancer-related anemia. Oncology (Huntington)1992.

7. Reed WR, Hussey DH, DeGowin RL. Implications of theanemia of chronic disorders in patients anticipating radio-therapy. Am J Med Sci 1994; 308:9–15.

8. Sproule BJ, Mitchell HJ, Miller WF. Cardiopulmonaryphysiological responses to heavy exercise in patientswith anemia. J Clin Invest 1960; 33:378–388.

9. Duke M, Abelmann WH. The hemodynamic response tochronic anemia. Circulation 1969; 39:503–515.

10. Brannon ES. The cardiac output in patients with chronicanemia as measured by the technique of right atrial catheter-ization. J Clin Invest 1945; 332.

11. Sharpey-Schafer EP. Cardiac output in severe anemia. ClinSci 1944; 11:400.

12. Bartels EC. Anemia as the cause of severe congestive heartfailure. Ann Intern Med 1937; 11:400.

536 Pohl and Ludwig

13. Pickering GW, Wayne EJ. Observations on angina pectorisand intermittent claudication in anemia. Clin Sci 1933; 1:305.

14. Sanghvi LM. Electrocardiogram in chronic severe anemia.Am Heart J 1958; 56:79.

15. Wintrobe MM, Lukens JN, Lee RG. The approach to thepatient with anemia. In: Lee GR, Bithell TC, Forester J,et al., eds. Wintrobe’s Clinical Hematology. 9th ed. Malvern,PA: Lea and Felbiger, 1993:715–744.

16. Varat MA, Adolph RJ, Fowler NO. Cardiovascular effects ofanemia. Am Heart J 1972; 83:415–426.

17. Anderson NP. Syndrome of spoon nails, anemia, cheilitis anddysphagia. Arch Dermatol Syph 1938; 37:816.

18. Bradley SE BG. Renal function during chronic anemia inman. Blood 1947; 2:192–202.


20. Spivak JL. The mechanisms of action of erythropoietin. Int JCell Cloning 1986; 4:139–166.


22. Gafter U, Kalechman Y, Orlin JB, Levi J, Sredni B. Anemiaof uremia is associated with reduced in vitro cytokine secre-tion: immunopotentiating activity of red blood cells. KidneyInt 1994; 45:224–231.

23. Pickett JL, Theberge DC, Brown WS, Schweitzer SU,Nissenson AR. Normalizing hematocrit in dialysis patientsimproves brain function. Am J Kidney Dis 1999; 33:1122–1130.

24. Nissenson AR, Goodnough LT, Dubois RW. Anemia: not justan innocent bystander? Arch Intern Med 2003; 163:1400–1404.


25. Aisen ML, Bacon BR, Goodman AM, Chester EM. Retinalabnormalities associated with anemia. Arch Ophthalmol1983; 101:1049–1052.

26. Ludwig H, Fritz E. Anemia in cancer patients. Semin Oncol1998; 25:2–6.

27. Ludwig H, Birgegard G, Olmi P, et al. European Cancer Ane-mia Survey (ECAS): Prospective evaluation of anemia in over15,000 cancer patients (abstr). Ann Oncol 2003; 13(suppl):623.

28. Manegold C. The causes and prognostic significance of lowhemoglobin levels in tumor patients. Strahlenther Onkol1998; 174(suppl 4):17–19.

29. Bokkel Huinink WW, de Swart CA, van Toorn DW, Morack G,Breed WP, Hillen HF, van der Hoeven JJ, Reed NS,Fairlamb DJ, Chan SY, Godfrey KA, Kristensen GB, van Tin-teren H, Ehmer B. Controlled multicentre study of the influ-ence of subcutaneous recombinant human erythropoietin onanaemia and transfusion dependency in patients with ovariancarcinoma treated with platinum-based chemotherapy. MedOncol 1998; 15:174–182.

30. Fein DA, Lee WR, Hanlon AL, Ridge JA, Langer CJ, CurranWJ Jr, Coia LR. Pretreatment hemoglobin level influenceslocal control and survival of T1–T2 squamous cell carcinomasof the glottic larynx. J Clin Oncol 1995; 13:2077–2083.

31. Mercadante S, Gebbia V, Marrazzo A, Filosto S. Anaemia incancer: pathophysiology and treatment. Cancer Treat Rev2000; 26:303–311.

32. Glaspy J, Cavill I. Role of iron in optimizing responses ofanemic cancer patients to erythropoietin. Oncology(Huntingt) 1999; 13:461–473.

33. Gordon MS. Managing anemia in the cancer patient: oldproblems, future solutions. Oncologist 2002; 7:331–341.

34. Rizzo JD, Lichtin AE, Woolf SH, Seidenfeld J, Bennett CL,Cella D, Djulbegovic B, Goode MJ, Jakubowski AA, Lee SJ,Miller CB, Rarick MU, Regan DH, Browman GP, GordonMS. Use of epoetin in patients with cancer: evidence-basedclinical practice guidelines of the American Society of Clinical

538 Pohl and Ludwig

Oncology and the American Society of Hematology. J ClinOncol 2002; 20:4083–4107.

35. Ludwig H, Rai K, Blade J, Dammacco F, Degos L, Itri L,Kyle R, Liso V, Littlewood TJ, Mandelli F, Meloni G,Molica S, Osterborg A, Pangalis GA, San Miguel J, SchmittB, Voliotis D. Management of disease-related anemia inpatients with multiple myeloma or chronic lymphocytic leu-kemia: epoetin treatment recommendations. Hematol J2002; 3:121–130.

36. Sabbatini P. The relationship between anemia and quality oflife in cancer patients. Oncologist 2000; 5(suppl 2):19–23.

37. Jacobs K, Shoemaker C, Rudersdorf R, Neill SD, Kaufman RJ,Mufson A, Seehra J, Jones SS, Hewick R, Fritsch EF. Isolationand characterization of genomic and cDNA clones of humanerythropoietin. Nature 1985; 313:806–810.

38. Lin FK, Suggs S, Lin CH, Browne JK, Smalling R, Egrie JC,Chen KK, Fox GM, Martin F, Stabinsky Z. Cloning andexpression of the human erythropoietin gene. Proc Natl AcadSci USA 1985; 82:7580–7584.

39. Kushner JP, Porter JP, Olivieri NF. Secondary iron overload.Hematology (Am Soc Hematol Educ Program) 2001; 47–61.

40. Anderson LJ, Wonke B, Prescott E, Holden S, Walker JM,Pennell DJ. Comparison of effects of oral deferiprone and sub-cutaneous desferrioxamine on myocardial iron concentrationsand ventricular function in beta-thalassemia. Lancet 2002;360:516–520.

41. Barrett-Lee PJ, Bailey NP, O’Brien ME, Wager E. Large-scaleUK audit of blood transfusion requirements and anaemia inpatients receiving cytotoxic chemotherapy. Br J Cancer 2000;82:93–97.

42. Coiffier B, Guastalla JP, Pujade-Lauraine E, Bastit P.Predicting cancer-associated anaemia in patients receivingnon-platinum chemotherapy: results of a retrospectivesurvey. Eur J Cancer 2001; 37:1617–1623.

43. Ortega A, Dranitsaris G, Puodziunas A. A clinical andeconomic evaluation of red blood cell transfusions in


patients receiving cancer chemotherapy. Int J Technol AssessHealth Care 1998; 14:788–798.

44. Faulds D, Sorkin EM. Epoetin (recombinant human erythropoietin). A review of its pharmacodynamic and pharma-cokinetic properties and therapeutic potential in anaemiaand the stimulation of erythropoiesis. Drugs 1989; 38:863–899.

45. Lacombe C, Da Silva JL, Bruneval P, Fournier JG, WendlingF, Casadevall N, Camilleri JP, Bariety J, Varet B, TambourinP. Peritubular cells are the site of erythropoietin synthesisin the murine hypoxic kidney. J Clin Invest 1988; 81:620–623.

46. Lin FK, Suggs S, Lin CH, Browne JK, Smalling R, Egrie JC,Chen KK, Fox GM, Martin F, Stabinsky Z. Cloning andexpression of the human erythropoietin gene. Proc Natl AcadSci USA 1985; 82:7580–7584.

47. Eschbach JW, Egrie JC, Downing MR, Browne JK, AdamsonJW. Correction of the anemia of end-stage renal diseasewith recombinant human erythropoietin. Results of a com-bined phase I and II clinical trial. N Engl J Med 1987;316:73–78.

48. Winearls CG, Oliver DO, Pippard MJ, Reid C, Downing MR,Cotes PM. Effect of human erythropoietin derived fromrecombinant DNA on the anaemia of patients maintainedby chronic haemodialysis. Lancet 1986; 2:1175–1178.


50. Glaspy J, Bukowski R, Steinberg D, Taylor C, TchekmedyianS, Vadhan-Raj S, Procrit Study Group. Impact of therapywith epoetin alfa on clinical outcomes in patients with non-myeloid malignancies during cancer chemotherapy incommunity oncology practice. J Clin Oncol 1997; 15:1218–1234.

51. Cascinu S, Fedeli A, Del Ferro E, Luzi FS, Catalano G.Recombinant human erythropoietin treatment in cisplatin-associated anemia: a randomized, double-blind trial withplacebo. J Clin Oncol 1994; 12:1058–1062.

540 Pohl and Ludwig

52. Demetri GD, Kris M, Wade J, Degos L, Cella D, Procrit StudyGroup. Quality-of-life benefit in chemotherapy patients trea-ted with epoetin alfa is independent of disease response ortumor type: results from a prospective community oncologystudy. J Clin Oncol 1998; 16:3412–3425.

53. Abels R. Erythropoietin for anaemia in cancer patients. Eur JCancer 1993; 29A(suppl 2):2–8.

54. Henry DH. Recombinant human erythropoietin treatment ofanemic cancer patients. Cancer Pract 1996; 4:180–184.

55. Egrie JC, Browne JK. Development and characterization ofnovel erythropoiesis stimulating protein (NESP). Br J Cancer2001; 84(suppl 1):3–10.

56. Reigner B, Jordan A, Panier J, Glaspy F. CERA (continuouserythropoiesis receptor activator), a novel erythropoieticagent: Dose-dependent response in phase I studies [abstr].Proc Am Soc Clin Oncol 2003; 22:732.

57. Haselbeck A, Reigner B, Jordan P, Pannier A, Glaspy J. Pre-clinical and phase I pharmacokinetic and mode-of-action stu-dies of CERA (continuous erythropoiesis receptor activator), anovel erythropoietic agent with an extended serum half-life[abstr]. Proc Am Soc Clin Oncol 2003; 22:732.

58. Rizzo JD, Lichtin AE, Woolf SH, Seidenfeld J, Bennett CL,Cella D, Djulbegovic B, Goode MJ, Jakubowski AA, Lee SJ,Miller CB, Rarick MU, Regan DH, Browman GP, GordonMS. Use of epoetin in patients with cancer: evidence-basedclinical practice guidelines of the American Society of ClinicalOncology and the American Society of Hematology. Blood2002; 100:2303–2320.

59. Beguin Y. Prediction of response and other improvements onthe limitations of recombinant human erythropoietintherapy in anemic cancer patients. Haematologica 2002; 87:1209–1221.

60. Ludwig H, Pecherstorfer M, Leitgeb C, Fritz E. Recombinanthuman erythropoietin for the treatment of chronic anemia inmultiple myeloma and squamous cell carcinoma. Stem Cells1993; 11:348–355.


61. Cazzola M, Messinger D, Battistel V, Bron D, Cimino R,Enller-Ziegler L, Essers U, Greil R, Grossi A, Jager G. Recom-binant human erythropoietin in the anemia associated withmultiple myeloma or non-Hodgkin’s lymphoma: dose findingand identification of predictors of response. Blood 1995;86:4446–4453.

62. Auerbach M, Barker L, Bahrain H, Trout R, Mcllwain M,Ballard H. Intravenous iron (IV Fe) optimises the responseto erythropoietin (Epo) in patients with anaemia of cancerand cancer chemotherapy: results of a multicenter, open-label, randomized trial [abstr]. Blood 2001; 98(suppl).


64. Adamson JW, Ludwig H. Predicting the hematopoieticresponse to recombinant human erythropoietin (epoetin alfa)in the treatment of the anemia of cancer. Oncology 1999;56:46–53.

65. Iki S, Yagisawa M, Ohbayashi Y, Sato H, Urabe A. Adverseeffect of erythropoietin in myeloproliferative disorders.Lancet 1991; 337:187–188.

66. Ehmer B, Roshan Y, Mocks J, Franke W. A meta-analysis onthe safety of epoitin-B therapy in cancer patients. In: SmythJF, Boogaerts MA, Ehmer BR, eds. rhuErythropoietin inCancer Supportive Treatment. New York: Marcel Dekker,1996:13–34.

67. Acs G, Acs P, Beckwith SM, Pitts RL, Clements E, Wong K,Verma A. Erythropoietin and erythropoietin receptor expres-sion in human cancer. Cancer Res 2001; 61:3561–3565.

68. Casadevall N, Nataf J, Viron B, Kolta A, Kiladjian JJ,Martin-Dupont P, Michaud P, Papo T, Ugo V, TeyssandierI, Varet B, Mayeux P. Pure red-cell aplasia and antierythro-poietin antibodies in patients treated with recombinanterythropoietin. N Engl J Med 2002; 346:469–475.

69. Barosi G, Marchetti M, Liberato NL. Cost-effectiveness ofrecombinant human erythropoietin in the prevention ofchemotherapy-induced anaemia. Br J Cancer 1998; 78:781–787.

542 Pohl and Ludwig

70. Demetri GD, Gabrilove JL, Blasi MV, Hill RJ Jr, Glaspy J.Benefits of epoetin alfa in anemic breast cancer patientsreceiving chemotherapy. Clin Breast Cancer 2002; 3:45–51.

71. Littlewood TJ, Bajetta E, Nortier JW, Vercammen E,Rapoport B. Effects of epoetin alfa on hematologic parametersand quality of life in cancer patients receiving nonplat-inum chemotherapy: results of a randomized, double-blind, placebo-controlled trial. J Clin Oncol 2001; 19:2865–2874.

72. Crawford J, Cella D, Cleeland CS, Cremieux PY, DemetriGD, Sarokhan BJ, Slavin MB, Glaspy JA. Relationshipbetween changes in hemoglobin level and quality of lifeduring chemotherapy in anemic cancer patients receivingepoetin alfa therapy. Cancer 2002; 95:888–895.

73. Ludwig H, Fritz E, Kotzmann H, Hocker P, Gisslinger H,Barnas U. Erythropoietin treatment of anemia associatedwith multiple myeloma. N Engl J Med 1990; 322:1693–1699.

74. Oster W, Herrmann F, Gamm H, Zeile G, Lindemann A,Muller G, Brune T, Kraemer HP, Mertelsmann R. Erythro-poietin for the treatment of anemia of malignancy associatedwith neoplastic bone marrow infiltration. J Clin Oncol 1990;8:956–962.

75. Platanias LC, Miller CB, Mick R, Hart RD, Ozer H, McEvillyJM, Jones RJ, Ratain MJ. Treatment of chemotherapy-induced anemia with recombinant human erythropoietin incancer patients. J Clin Oncol 1991; 9:2021–2026.

76. Henry DH. Clinical application of recombinant erythropoietinin anemic cancer patients. Hematol Oncol Clin North Am1994; 8:961–973.

77. Barlogie B. Treatment of the anemia of multiple myeloma:the role of recombinant human erythropoietin. Semin Hema-tol 1993; 30:25–27.

78. Mittelman M, Zeidman A, Fradin Z, Magazanik A, LewinskiUH, Cohen A. Recombinant human erythropoietin in thetreatment of multiple myeloma-associated anemia. ActaHaematol 1997; 98:204–210.


79. Musto P, Falcone A, D’Arena G, Scalzulli PR, Matera R,Minervini MM, Lombardi GF, Modoni S, Longo A, Carote-nuto M. Clinical results of recombinant erythropoietin intransfusion-dependent patients with refractory multiplemyeloma: role of cytokines and monitoring of erythropoiesis.Eur J Haematol 1997; 58:314–319.

80. Silvestris F, Romito A, Fanelli P, Vacca A, Dammacco F.Long-term therapy with recombinant human erythropoietin(rHu-EPO) in progressing multiple myeloma. Ann Hematol1995; 70:313–318.

81. Dammacco F, Silvestris F, Castoldi GL, Grassi B, Bernasconi C,Nadali G, Perona G, De Laurenzi A, Torelli U, Ascari E, RossiFerrini PL, Caligaris-Cappio F, Pileri A, Resegotti L. The effec-tiveness and tolerability of epoetin alfa in patients withmultiplemyeloma refractory to chemotherapy. Int J Clin Lab Res 1998;28:127–134.

82. Dammacco F, Castoldi G, Rodjer S. Efficacy of epoetin alfa inthe treatment of anaemia of multiple myeloma. Br J Haema-tol 2001; 113:172–179.

83. Osterborg A, Boogaerts MA, Cimino R, Essers U, HolowieckiJ, Juliusson G, Jager G, Najman A, Peest D. The EuropeanStudy Group of Erythropoietin (Epoetin Beta) Treatment inMultiple Myeloma and Non-Hodgkin’s Lymphoma. Recombi-nant human erythropoietin in transfusion-dependent anemicpatients with multiple myeloma and non-Hodgkin’s lym-phoma—a randomized multicenter study. Blood 1996;87:2675–2682.

84. Osterborg A, Brandberg Y, Molostova V, Iosava G,Abdulkadyrov K, Hedenus M, Messinger D. Randomized,double-blind, placebo-controlled trial of recombinant humanerythropoietin, epoetin beta, in hematologic malignancies. JClin Oncol 2002; 20:2486–2494.

85. Mittelman M, Neumann D, Peled A, Kanter P, Haran-GheraN. Erythropoietin induces tumor regression and antitumorimmune responses in murine myeloma models. Proc NatlAcad Sci USA 2001; 98:5181–5186.

86. Hedenus M, Hansen S, Taylor K, Arthur C, Emmerich B,Dewey C, Watson D, Rossi G, Osterborg A. Randomized,

544 Pohl and Ludwig

dose-finding study of darbepoetin alfa in anaemic patientswith lymphoproliferative malignancies. Br J Haematol2002; 119:79–86.

87. Cazzola M, Beguin Y, Kloczko J, Spicka I, Coiffier B. Once-weekly epoetin beta is highly effective in treating anaemicpatients with lymphoproliferative malignancy and defectiveendogenous erythropoietin production. Br J Haematol 2003;122:386–393.

88. Hedenus M, Adriansson M, Miguel JS, Kramer MH, Schip-perus MR, Juvonen E, Taylor K, Belch A, Altes A, MartinelliG, Watson D, Matcham J, Rossi G, Littlewood TJ. Efficacyand safety of darbepoetin alfa in anaemic patients withlymphoproliferative malignancies: a randomized, double-blind, placebo-controlled study. Br J Haematol 2003; 122:394–403.

89. Rose E, Rai K, Revicki D, Brown R, Reblando J. Clinical andhealth status assessments in anemic chronic lymphocyticleukemia (CLL) patients treated with epoetin alfa (EPO)[abstr]. Blood 1994; 84(suppl 1):526.

90. Siakantaris MP, Angelopoulou MK, Vassilakopoulos TP,Dimopoulou MN, Kontopidou FN, Pangalis GA. Correctionof disease related anaemia of B-chronic lymphoproliferativedisorders by recombinant human erythropoietin: mainte-nance is necessary to sustain response. Leuk Lymphoma2000; 40:141–147.

91. Pangalis GA, Siakantaris MP, Angelopoulou MK, Vassilako-poulos TP, Dimopoulou MN, Kyrtsonis MC, KonstantopoulosK, Tsaftaridis P, Vaiopoulos GA, Kontopidou FN. Downsta-ging Rai stage III B-chronic lymphocytic leukemia patientswith the administration of recombinant human erythropoie-tin. Haematologica 2002; 87:500–506.

92. Stein RS. The role of erythropoietin in the anemia of myelo-dysplastic syndrome. Clin Lymphoma. In press.

93. Cazzola M, Ponchio L, Pedrotti C, Farina G, Cerani P, LucottiC, Novella A, Rovati A, Bergamaschi G, Beguin Y. Predictionof response to recombinant human erythropoietin(rHuEpo) in anemia of malignancy. Haematologica 1996; 81:434–441.


94. Matsuda A, Kishimoto K, Yoshida K, Yagasaki F, Ito Y,Sakata T, Kawai N, Ino H, Hirashima K, Bessho M. Long-term follow-up of patients with aplastic anemia and refrac-tory anemia responding to combination therapy with recom-binant human granulocyte colony-stimulating factor anderythropoietin. Int J Hematol 2002; 76:244–250.

95. Stebler C, Tichelli A, Dazzi H, Gratwohl A, Nissen C, SpeckB. High-dose recombinant human erythropoietin for treat-ment of anemia in myelodysplastic syndromes and paroxys-mal nocturnal hemoglobinuria: a pilot study. Exp Hematol1990; 18:1204–1208.

96. Hellstrom E, Birgegard G, Lockner D, Helmers C, Ost A,Wide L. Treatment of myelodysplastic syndromes with recom-binant human erythropoietin. Eur J Haematol 1991; 47:355–360.

97. Casadevall N, Belanger C, Goy A, Varet B, Lang J, Poisson D.High-dose recombinant human erythropoietin administeredintravenously for the treatment of anaemia in myelodysplas-tic syndromes. Acta Haematol 1992; 87(suppl 1):25–27.

98. Ludwig H, Fritz E, Leitgeb C, Krainer M, Kuhrer I, SagasterP, Umek H. Erythropoietin treatment for chronic anemia ofselected hematological malignancies and solid tumors. AnnOncol 1993; 4:161–167.

99. Stone RM, Bernstein SH, Demetri G, Facklam DP, Arthur K,Andersen J, Aster JC, Kufe D. Therapy with recombinanthuman erythropoietin in patients with myelodysplasticsyndromes. Leuk Res 1994; 18:769–776.

100. Rose EH, Abels RI, Nelson RA, McCullough DM, Lessin L.The use of r-HuEpo in the treatment of anaemia related tomyelodysplasia (MDS). Br J Haematol 1995; 89:831–837.

101. Bernell P, Stenke L, Wallvik J, Hippe E, Hast R. A sequentialerythropoietin and GM-CSF schedule offers clinical benefitsin the treatment of anaemia in myelodysplastic syndromes.Leuk Res 1996; 20:693–699.

102. Hellstrom-Lindberg E. Efficacy of erythropoietin in themyelodysplastic syndromes: a meta-analysis of 205 patientsfrom 17 studies. Br J Haematol 1995; 89:67–71.

546 Pohl and Ludwig

103. Hellstrom-Lindberg E, Birgegard G, Carlsson M, Carneskog J,Dahl IM, Dybedal I, Grimfors G, Merk K, Tangen JM,WinqvistI. A combination of granulocyte colony-stimulating factor anderythropoietin may synergistically improve the anaemia inpatients with myelodysplastic syndromes. Leuk Lymphoma1993; 11:221–228.

104. Imamura M, Kobayashi M, Kobayashi S, Yoshida K, MikuniC, Ishikawa Y, Matsumoto S, Sakamaki S, Niitsu Y, HinodaY. Failure of combination therapy with recombinant granulo-cyte colony-stimulating factor and erythropoietin in myelo-dysplastic syndromes. Ann Hematol 1994; 68:163–166.

105. Negrin RS, Stein R, Doherty K, Cornwell J, Vardiman J,Krantz S, Greenberg PL. Maintenance treatment of theanemia of myelodysplastic syndromes with recombinanthuman granulocyte colony-stimulating factor and erythro-poietin: evidence for in vivo synergy. Blood 1996; 87:4076–4081.

106. Remacha AF, Arrizabalaga B, Villegas A, Manteiga R, Calvo T,Julia A, Fernandez F, I, Gonzalez FA, Font L, Junca J, del ArcoA, Malcorra JJ, Equiza EP, de Mendiguren BP, Romero M.The Spanish Erythropathology Group. Erythropoietin plusgranulocyte colony-stimulating factor in the treatment ofmyelodysplastic syndromes. Identification of a subgroup ofresponders. Haematologica 1999; 84:1058–1064.

107. Mantovani L, Lentini G, Hentschel B, Wickramanayake PD,Loeffler M, Diehl V, Tesch H. Treatment of anaemia in mye-lodysplastic syndromes with prolonged administration ofrecombinant human granulocyte colony-stimulating factorand erythropoietin. Br J Haematol 2000; 109:367–375.

108. RundeV, Aul C, Ebert A, Grabenhorst U, SchneiderW. Sequen-tial administration of recombinant human granulocyte-macro-phage colony-stimulating factor and human erythropoietin fortreatment of myelodysplastic syndromes. Eur J Haematol1995; 54:39–45.

109. Stasi R, Pagano A, Terzoli E, Amadori S. Recombinant humangranulocyte-macrophage colony-stimulating factor pluserythropoietin for the treatment of cytopenias in patientswith myelodysplastic syndromes. Br J Haematol 1999; 105:141–148.


110. Economopoulos T, Mellou S, Papageorgiou E, Pappa V,Kokkinou V, Stathopoulou E, Pappa M, Raptis S. Treatmentof anemia in low risk myelodysplastic syndromes withgranulocyte-macrophage colony-stimulating factor plusrecombinant human erythropoietin. Leukemia 1999; 13:1009–1012.

111. Miller AM, Noyes WE, Taetle R, List AF. Limited erythro-poietic response to combined treatment with recombinanthuman interleukin 3 and erythropoietin in myelodysplasticsyndrome. Leuk Res 1999; 23:77–83.

112. Besa EC, Kunselman S, Nowell PC, The Central Pennsylva-nia Oncology Group. A pilot trial of 13-cis-retinoic acid andalpha-tocopherol with recombinant human erythropoietin inmyelodysplastic syndrome patients with progressive or trans-fusion-dependent anemias. Leuk Res 1998; 22:741–749.

113. Thompson JA, Gilliland DG, Prchal JT, Bennett JM, LarholtK, Nelson RA, Rose EH, Dugan MH, GM=EPO MDS StudyGroup. Effect of recombinant human erythropoietincombined with granulocyte=macrophage colony-stimulatingfactor in the treatment of patients with myelodysplasticsyndrome. Blood 2000; 95:1175–1179.

114. Terpos E, Mougiou A, Kouraklis A, Chatzivassili A, Michalis E,Giannakoulas N, Manioudaki E, Lazaridou A, Bakaloudi V,Protopappa M, Liapi D, Grouzi E, Parharidou A, SymeonidisA, Kokkini G, Laoutaris NP, Vaipoulos G, AnagnostopoulosNI, Christakis JI, Meletis J, Bourantas KL, Zoumbos NC,Yataganas X, Viniou NA. Prolonged administration of erythro-poietin increases erythroid response rate in myelodysplasticsyndromes: a phase II trial in 281 patients. Br J Haematol2002; 118:174–180.

115. Hellstrom-Lindberg E, Gulbrandsen N, Lindberg G, Ahlgren T,Dahl IM, Dybedal I, Grimfors G, Hesse-Sundin E, Hjorth M,Kanter-Lewensohn L, Linder O, Luthman M, Lofvenberg E,Oberg G, Porwit-MacDonald A, Radlund A, Samuelsson J, Tan-gen JM, Winquist I, Wisloff F. A validated decision model fortreating the anaemia of myelodysplastic syndromes witherythropoietin þ granulocyte colony-stimulating factor: signifi-cant effects on quality of life. Br J Haematol 2003; 120:1037–1046.

548 Pohl and Ludwig

116. Vinh TT, Lewis BH, Yao B, Yim J. Initial experience withdarbepoetin alfa in patients with myelodysplastic syndrome(MDS) [abstr]. Proc Am Soc Clin Oncol 2003; 22:613.

117. LudwigH,SundalE,PecherstorferM,LeitgebC,BauernhoferT,Beinhauer A, Samonigg H, Kappeler AW, Fritz E. Recombinanthuman erythropoietin for the correction of cancer associatedanemia with and without concomitant cytotoxic chemotherapy.Cancer 1995; 76:2319–2329.

118. Miller CB, Platanias LC, Mills SR, Zahurak ML, Ratain MJ,Ettinger DS, Jones RJ. Phase I–II trial of erythropoietin inthe treatment of cisplatin-associated anemia. J Natl CancerInst 1992; 84:98–103.

119. Cascinu S, Fedeli A, Fedeli SL, Catalano G. Cisplatin-associated anaemia treated with subcutaneous erythropoie-tin. A pilot study. Br J Cancer 1993; 67:156–158.

120. Case DC Jr, Bukowski RM, Carey RW, Fishkin EH, HenryDH, Jacobson RJ, Jones SE, Keller AM, Kugler JW, NicholsCR. Recombinant human erythropoietin therapy for anemiccancer patients on combination chemotherapy. J Natl CancerInst 1993; 85:801–806.

121. Tsukuda M, Mochimatsu I, Nagahara T, Kokatsu T, Sawaki S,Kubota A, Furkawa M, Arai Y. Clinical application of recombi-nant human erythropoietin for treatments in patients withhead and neck cancer. Cancer Immunol Immunother 1993;36:52–56.

122. Kasper C, Terhaar A, Fossa A, Welt A, Seeber S, NowrousianMR. Recombinant human erythropoietin in the treatment ofcancer-related anaemia. Eur J Haematol 1997; 58:251–256.

123. Pawlicki M, Jassem J, Bosze P, Lotan C, Kurteva GP, SiddiquiM, Kosmidis PA, Rigatos GA, Kansu E, Durkovic P, Aziz Z,Al Idrissi H, Roth A, Cozma G. A multicenter study of recombi-nant human erythropoietin (epoetin alpha) in the managementof anemia in cancer patients receiving chemotherapy. Antican-cer Drugs 1997; 8:949–957.

124. Oberhoff C, Neri B, Amadori D, Petry KU, Gamucci T,Rebmann U, Nowrousian MR, Voigtmann R, Monfardini S,Armand JP, Herrmann R, Netter-Pinon J, Tubiana-MathieuN, Zwierzina H. Recombinant human erythropoietin in


the treatment of chemotherapy-induced anemia and pre-vention of transfusion requirement associated with solidtumors: a randomized, controlled study. Ann Oncol 1998;9:255–260.

125. Porter JC, Leahey A, Polise K, Bunin G, Manno CS. Recom-binant human erythropoietin reduces the need for erythro-cyte and platelet transfusions in pediatric patients withsarcoma: a randomized, double-blind, placebo-controlledtrial. J Pediatr 1996; 129:656–660.

126. Wurnig C, Windhager R, Schwameis E, Kotz R, Zoubek A,Stockenhuber F, Kurz RW. Prevention of chemotherapy-induced anemia by the use of erythropoietin in patients withprimary malignant bone tumors (a double-blind, randomized,phase III study). Transfusion 1996; 36:155–159.

127. Kurz C, Marth C, Windbichler G, Lahousen M, Medl M,Vavra N, Sevelda P. Erythropoietin treatment under polyche-motherapy in patients with gynecologic malignancies: aprospective, randomized, double-blind placebo-controlledmulticenter study. Gynecol Oncol 1997; 65:461–466.

128. Tsukuda M, Yuyama S, Kohno H, Itoh K, Kokatsu T, Kawai S.Effectiveness of weekly subcutaneous recombinant humanerythropoietin administration for chemotherapy-inducedanemia. Biotherapy 1998; 11:21–25.

129. Quirt I, Micucci S, Moran LA, Pater J, Browman G, SystemicTreatment Program Committee. Erythropoietin in the man-agement of patients with nonhematologic cancer receivingchemotherapy. Cancer Prev Control 1997; 1:241–248.

130. Littlewood TJ, Bajetta E, Nortier JW, Vercammen E,Rapoport B. Effects of epoetin alfa on hematologic para-meters and quality of life in cancer patients receiving non-platinum chemotherapy: results of a randomized, double-blind, placebo-controlled trial. J Clin Oncol 2001; 19:2865–2874.

131. Gabrilove JL, Cleeland CS, Livingston RB, Sarokhan B,Winer E, Einhorn LH. Clinical evaluation of once-weekly dos-ing of epoetin alfa in chemotherapy patients: improvementsin hemoglobin and quality of life are similar to three-times-weekly dosing. J Clin Oncol 2001; 19:2875–2882.

550 Pohl and Ludwig

132. Rushing DA, Einhorn LH, Wildgust M. Epoetin alfa 40,000 Uonce weekly significantly improves hemoglobin and quality oflife in anemic colorectal cancer patients [abstr]. Proc Am SocClin Oncol 2003; 22:771.

133. Campos SM. Epoetin alfa 40,000 U once weekly significantlyimproves hemoglobin and quality of life in anemic patientswith ovarian cancer [abstr]. Proc Am Soc Clin Oncol 2003;22:774.

134. Quirt I, Robeson C, Lau CY, Kovacs M, Burdette-Radoux S,Dolan S, Tang SC, McKenzie M, Couture F. Epoetin alfa ther-apy increases hemoglobin levels and improves quality of lifein patients with cancer-related anemia who are not receivingchemotherapy and patients with anemia who are receivingchemotherapy. J Clin Oncol 2001; 19:4126–4134.

135. Markman M, Reichman B, Hakes T, Rubin S, Jones W,Lewis JL Jr, Barakat R, Curtin J, Almadrones L, HoskinsW. The use of recombinant human erythropoietin to preventcarboplatin-induced anemia. Gynecol Oncol 1993; 49:172–176.

136. Del Mastro L, Venturini M, Lionetto R, Garrone O, Melioli G,Pasquetti W, Sertoli MR, Bertelli G, Canavese G, Costantini M,Rosso R. Randomized phase III trial evaluating the role oferythropoietin in the prevention of chemotherapy-inducedanemia. J Clin Oncol 1997; 15:2715–2721.

137. ThatcherN, De Campos ES, Bell DR, StewardWP, Varghese G,Morant R, Vansteenkiste JF, Rosso R, Ewers SB, Sundal E,Schatzmann E, Stocker H. Epoetin alpha prevents anaemiaand reduces transfusion requirements in patients undergoingprimarily platinum-based chemotherapy for small cell lungcancer. Br J Cancer 1999; 80:396–402.

138. Dunphy FR, Harrison BR, Dunleavy TL, Rodriguez JJ, Hil-ton JG, Boyd JH. Erythropoietin reduces anemia and trans-fusions: a randomized trial with or without erythropoietinduring chemotherapy. Cancer 1999; 86:1362–1367.

139. Kunikane H, Watanabe K, Fukuoka M, Saijo N, Furuse K,Ikegami H, Ariyoshi Y, Kishimoto S. Double-blind rando-mized control trial of the effect of recombinant human ery-thropoietin on chemotherapy-induced anemia in patients


with non-small cell lung cancer. Int J Clin Oncol 2001; 6:296–301.

140. Bamias A, Aravantinos G, Kalofonos C, Timotheadou N,Siafaka V, Vlahou I, Janinis D, Pectasides D, Pavlidis N,Fountzilas G. Prevention of anemia in patients with solidtumors receiving platinum-based chemotherapy by recombi-nant human erythropoietin (rHuEpo): a prospective, openlabel, randomized trial by the Hellenic Cooperative OncologyGroup. Oncology 2003; 64:102–110.

141. O’Shaughnessy J. Effects of epoetin alfa on cognitive func-tion, mood, asthenia, and quality of life in women with breastcancer undergoing adjuvant chemotherapy. Clin BreastCancer 2002; (suppl 3):116–120.

142. Hudis C, Williams D, Gralow JR, PROCRIT Study Group.Epoetin alfa maintains hemoglobin and quality of life inbreast cancer patients receiving conventional adjuvant che-motherapy: final report [abstr]. Proc Am Soc Clin Oncol2003; 22:767.

143. Crawford J, Robert F, Perry M, Belani P, Sarokhan B. Epoetinalfa 40,000 U once weekly maintains hemoglobin in advancednon-small-cell lung cancer patients receiving first-linechemotherapy [abstr]. Proc Am Soc Clin Oncol 2003; 22:13.

144. McCormack M, Nias AH, Smith E. Chronic anaemia,hyperbaric oxygen and tumour radiosensitivity. Br J Radiol1990; 63:752–759.

145. Grant DG, Hussain A, Hurman D. Pre-treatment anaemiaalters outcome in early squamous cell carcinoma of the larynxtreated by radical radiotherapy. J Laryngol Otol 1999;113:829–833.

146. Grogan M, Thomas GM, Melamed I, Wong FL, Pearcey RG,Joseph PK, Portelance L, Crook J, Jones KD. The importanceof hemoglobin levels during radiotherapy for carcinoma of thecervix. Cancer 1999; 86:1528–1536.

147. Frommhold H, Guttenberger R, HenkeM. The impact of bloodhemoglobin content on the outcome of radiotherapy. The Frei-burg experience. Strahlenther Onkol 1998; 174(suppl 4):31–34.

552 Pohl and Ludwig

148. Dusenbery KE, McGuire WA, Holt PJ, Carson LF, Fowler JM,Twiggs LB, Potish RA. Erythropoietin increases hemoglobinduring radiation therapy for cervical cancer. Int J RadiatOncol Biol Phys 1994; 29:1079–1084.

149. Lavey RS, Dempsey WH. Erythropoietin increases hemoglo-bin in cancer patients during radiation therapy. Int J RadiatOncol Biol Phys 1993; 27:1147–1152.

150. Sweeney PJ, Nicolae D, Ignacio L, Chen L, Roach M, III,Wara W, Marcus KC, Vijayakumar S. Effect of subcutaneousrecombinant human erythropoietin in cancer patients receiv-ing radiotherapy: final report of a randomized, open-labelled,phase II trial. Br J Cancer 1998; 77:1996–2002.

151. Glaser CM, Millesi W, Kornek GV, Lang S, Schull B,Watzinger F, Selzer E, Lavey RS. Impact of hemoglobin levelanduseofrecombinanterythropoietinonefficacyofpreoperativechemoradiation therapy for squamous cell carcinoma of the oralcavity and oropharynx. Int J Radiat Oncol Biol Phys 2001;50:705–715.

152. Blohmer JU, Wurschmidt F, Petry U, Weise G, Sehouli J,Kimmig R. Sixth interim analysis of a prospective,randomized, open and controlled AGO- and NOGGO-intergroup study: sequential adjuvant chemo-radiotherapywith vs without epoetin alfa for pts with high-risk cervicalcancer [abstr]. Proc Am Soc Clin Oncol 2003; 22.

153. Tsuji Y, Kambayashi J, Shiba E, Sakon M, Kawasaki T, MoriT. Effect of recombinant human erythropoietin on anaemiaafter gastrectomy: a pilot study. Eur J Surg 1995; 161:29–33.

154. Braga M, Gianotti L, Gentilini O, Vignali A, Di C. Erythro-poietic response induced by recombinant human erythropoie-tin in anemic cancer patients candidate to major abdominalsurgery. Hepatogastroenterology 1997; 44:685–690.

155. Kettelhack C, Hones C, Messinger D, Schlag PM. Rando-mized multicentre trial of the influence of recombinanthuman erythropoietin on intraoperative and postoperativetransfusion need in anaemic patients undergoing right hemi-colectomy for carcinoma. Br J Surg 1998; 85:63–67.


156. Rau B, Schlag PM, Willeke F, Herfarth C, Stephan P, FrankeW. Increased autologous blood donation in rectal cancer byrecombinant human erythropoietin (rhEPO). Eur J Cancer1998; 34:992–998.

157. Scott SN, Boeve TJ, McCulloch TM, Fitzpatrick KA, KarnellLH. The effects of epoetin alfa on transfusion requirementsin head and neck cancer patients: a prospective, rando-mized, placebo-controlled study. Laryngoscope 2002; 112:1221–1229.

158. Kosmadakis N, Messaris E, Maris A, Katsaragakis S,Leandros E, KonstadoulakisMM, Androulakis G. Perioperativeerythropoietin administration in patients with gastrointestinaltract cancer: prospective randomized double-blind study. AnnSurg 2003; 237:417–421.


160. Klaesson S, Ringden O, Ljungman P, Lonnqvist B,Wennberg L. Reduced blood transfusions requirements afterallogeneic bone marrow transplantation: results of a rando-mised, double-blind study with high-dose erythropoietin.Bone Marrow Transplant 1994; 13:397–402.

161. Link H, Boogaerts MA, Fauser AA, Slavin S, Reiffers J,Gorin NC, Carella AM, Mandelli F, Burdach S, Ferrant A. Acontrolled trial of recombinant human erythropoietin afterbone marrow transplantation. Blood 1994; 84:3327–3335.

162. Pene R, Appelbaum FR, Fisher L, Lilleby K, Nemunaitis J,Storb R, Buckner CD. Use of granulocyte-macrophagecolony-stimulating factor and erythropoietin in combinationafter autologous marrow transplantation. Bone MarrowTransplant 1993; 11:219–222.

163. Chao NJ, Schriber JR, Long GD, Negrin RS, Catolico M,Brown BW, Miller LL, Blume KG. A randomized study of ery-thropoietin and granulocyte colony-stimulating factor (G-CSF) versus placebo and G-CSF for patients withHodgkin’s and non-Hodgkin’s lymphoma undergoing autolo-gous bone marrow transplantation. Blood 1994; 83:2823–2828.

554 Pohl and Ludwig

164. Daneryd P. Epoetin alfa for protection of metabolic andexercise capacity in cancer patients. Semin Oncol 2002; 29:69–74.

165. Bottomley A, Thomas R, Van Steen K, Flechtner H, Djulbego-vic B. Human recombinant erythropoietin and quality of life:a wonder drug or something to wonder about? Lancet Oncol2002; 3:145–153.

166. Van Belle SJ, Cocquyt V. Impact of haemoglobin levels on theoutcome of cancers treated with chemotherapy. Crit RevOncol Hematol 2003; 47:1–11.

167. Becker A, Stadler P, Lavey RS, Hansgen G, Kuhnt T,Lautenschlager C, Feldmann HJ, Molls M, Dunst J. Severeanemia is associated with poor tumor oxygenation in headand neck squamous cell carcinomas. Int J Radiat Oncol BiolPhys 2000; 46:459–466.

168. Dunst J, Pigorsch S, Hansgen G, Hintner I, Lautenschlager C,Becker A. Low hemoglobin is associated with increased serumlevels of vascular endothelial growth factor (VEGF) in cancerpatients. Does anemia stimulate angiogenesis? StrahlentherOnkol 1999; 175:93–96.

169. Semenza GL, Agani F, Feldser D, Iyer N, Kotch L, Laughner E,Yu A. Hypoxia, HIF-1, and the pathophysiology of commonhuman diseases. Adv Exp Med Biol 2000; 475:123–130.

170. Liang BC. Effects of hypoxia on drug resistance phenotypeand genotype in human glioma cell lines. J Neurooncol1996; 29:149–155.

171. Takigawa N, Segawa Y, Okahara M, Maeda Y, Takata I,Kataoka M, Fujii M. Prognostic factors for patients withadvanced non-small cell lung cancer: univariate and multi-variate analyses including recursive partitioning and amal-gamation. Lung Cancer 1996; 15:67–77.

172. Caro JJ, Salas M, Ward A, Goss G. Anemia as an indepen-dent prognostic factor for survival in patients with cancer:a systemic, quantitative review. Cancer 2001; 91:2214–2221.

173. Girinski T, Pejovic-Lenfant MH, Bourhis J, Campana F,Cosset JM, Petit C, Malaise EP, Haie C, Gerbaulet A,Chassagne D. Prognostic value of hemoglobin concentrations


and blood transfusions in advanced carcinoma of the cervixtreated by radiation therapy: results of a retrospective studyof 386 patients. Int J Radiat Oncol Biol Phys 1989; 16:37–42.

174. Littlewood TJ, Bajetta E, Nortier JW, Vercammen E,Rapoport B. Effects of epoetin alfa on hematologic para-meters and quality of life in cancer patients receiving non-platinum chemotherapy: results of a randomized, double-blind, placebo-controlled trial. J Clin Oncol 2001; 19:2865–2874.

175. Leyland-Jones B. Breast cancer trial with erythropoietinterminated unexpectedly. Lancet Oncol 2003; 4:459–460.

176. Mughal T. Anemia in patients with cancer: an overview. In:Bokemeyer C, Ludwig H, eds. Anemia in Cancer. Elsevier,2001:15–23.

177. Bokemeyer C. Causes of anaemia in cancer patients withemphasis on treatment factors. In: Bokemeyer C, Ludwig H,eds. Anemia in Cancer. : Elsevier, 2001:43–54.

178. Case DC Jr, Bukowski RM, Carey RW, Fishkin EH, Henry DH,Jacobson RJ, Jones SE, Keller AM, Kugler JW, Nichols CR.Recombinant human erythropoietin therapy for anemic cancerpatients on combination chemotherapy. J Natl Cancer Inst1993; 85:801–806.

179. de Campos E, Radford J, Steward W, Milroy R, Dougal M,Swindell R, Testa N, Thatcher N. Clinical and in vitro effectsof recombinant human erythropoietin in patients receivingintensive chemotherapy for small-cell lung cancer. J ClinOncol 1995; 13:1623–1631.

180. Vansteenkiste J, Pirker R, Massuti B, Barata F, Font A, FieglM, Siena S, Gateley J, Tomita D, Colowick AB, Musil J.Double-blind, placebo-controlled, randomized phase III trialof darbepoetin alfa in lung cancer patients receivingchemotherapy. J Natl Cancer Inst 2002; 94:1211–1220.

181. Olsson AM, Svensson JH, Sundstrom J, Bergstrom S, EdeklingT, Carlsson G, Hansen J, Svensson B, Albertsson M. Erythro-poietin treatment in metastatic breast cancer—effects on Hb,quality of life and need for transfusion. Acta Oncol 2002;41:517–524.

556 Pohl and Ludwig

182. Glaspy J, Degos L, Dicato M, Demetri GD. Comparableefficacy of epoetin alfa for anemic cancer patients receiving pla-tinum- and nonplatinum-based chemotherapy: a retrospectivesubanalysis of two large, community-based trials. Oncologist2002; 7:126–135.


19

Putative Negative Effects of theCorrection of Anemia in ACD

GUNTER WEISS

Department of General Internal Medicine,Clinical Immunology and Infectious Diseases,

Medical University, Innsbruck, Austria

INTRODUCTION: ACD AS A DEFENSESTRATEGY OF THE BODY!

Being aware of the fact that ACD is the most frequent anemiain hospitalized patients, the question arises if this condition isjust a side effect of an ongoing and very well controlledimmune response or whether or not nature had an ideabehind the development of this condition. As outlined in pre-vious chapters of this book, lowering the hemoglobin concen-tration has several pathophysiological advantages for fightingthe pathologic process underlying ACD.

559

First, the withdrawal of iron from the circulation and itsstorage within the reticuloendolial system, which is a basicpathophysiological mechanism underlying ACD, limits theavailability of this essential metal for the growth of micro-organisms and tumor cells. Iron is a central compound ofenzymes involved in the citric acid cycle such as mitochon-drial aconitase, in oxidative phosphorylation such as NADHubiquinone oxidoreductase or NADH succinate oxidoreduc-tase, and in DNA synthesis such as ribonucelotide reductase(1). Thus, limitation of iron availability is a very effectivedefense strategy of the body to control the growth of patho-gens as outlined in the chapter by Dr. Weinberg.

Second, the development of anemia limits oxygen trans-port capacity in general, and rapid proliferating tissues aremost affected since oxygen is an essential compound forenergy metabolism and thus for the proliferation and differ-entiation of cells (2). Hypoxia may also induce a counterbalan-cing effect in inducing the release of hormones such asvascular endothelial growth factor (VEGF), which inducesangiogenesis and neovascularization to provide a sufficientsupply of nutrients to tumor cells (for review see Ref. 3).

Third, limitation of iron may strengthen the immuneresponse directed against invading pathogens and tumorcells. Apart from all the immunomodulatory effects, whichare detailed in the chapter by Cardoso and colleagues, ironnegatively affects the activity of cell mediated immune effec-tor function. This notion arises from in vitro and in vivo obser-vations that iron loading of macrophages reduces theirresponsiveness to IFN-g, a central T-helper cell type 1 derivedcytokine, which is of pivotal importance to control an acuteinfection or the emerge of malignant cells (4). Following treat-ment with iron salts or transferrin bound iron, the inductionof IFN-g mediated immune effector pathways is impaired (5–7).These pathways include IFN-g induced expression of MCHclass II antigens and ICAM-1, formation and release of tumornecrosis factor alpha (TNF-a), degradation of the essentialamino acid tryptophan to from kynurenine via induction ofindoleamine-2,3-dioxygenase (8) or formation of neopterin, apyrazino-pyrimidino-derivative produced upon interaction of

560 Weiss

GTP and the IFN-g inducible enzyme GTP-cyclohydrolase I(9). In contrast, the induction of an iron deficient state, eitherby dietary iron restriction or by treatment with an iron chela-tor such as desferroxamine, increases the activity of theseIFN-g induced effector branches as compared to normal ironstatus. In a line with this, macrophages loaded with iron loosetheir ability to kill intracellular pathogens such as Listeria,Legionella, or Ehrlichia spp. by IFN-g mediated pathways(10). In contrast, iron chelation restores the immune effectorpotential of these cells, which are then able to clear theseinfections upon stimulation with IFN-g (11–13). Part of thisphenomenon may relate to a direct inhibitory effect of irontowards the expression of the critical immune effectorenzyme inducible nitric oxide synthase (iNOS, 14). This isof importance since high output formation of NO by macro-phages is one of the central antimicrobial and antitumorweapons of innate immunity (15,16). Iron inhibits the tran-scription of iNOS by blocking the binding affinity of the criti-cal transcription factors NF-IL6 and hypoxia inducible factor1 (HIF-1) to the iNOS promoter (17,18). Interestingly,another enzyme of innate immunity, which confers resistancetoward infection with intracellular pathogens such as Leish-mania, Salmonella, or Mycobacteria spp., is also linked toiron homeostasis. This protein, named natural resistanceassociated macrophage protein 1 (NRAMP-1), is expressedin the late phagosome of monocytes and neutrophils andconsists of 12 transmembrane domains with a divalent iontransport motif (19,20). Evidence has been provided thatNRAMP-1 is a transporter of iron across the phagolysosomalmembrane, thus modulating the metal’s concentrations inthe cytoplasm and phagolysosome (21–23), which is of impor-tance since (i) iron catalyzes the formation of toxic radicalsthat are of importance for host defenses, (ii) limitation of ironavailability negatively affects the proliferation of microbes,and (iii) the reduction of metabolically active iron enhancesT-helper cell mediated immune effector pathways includingthe formation of NO directed against these micro-organisms.

Part of the immunomodulatory effects of iron may notonly be related to a direct interaction of iron with these

Putative Negative Effects of the Correction of Anemia in ACD 561

immune pathways but also to the fact that by inhibitingIFN-g activity iron modulates the Th-1=Th-2 balance. Thisnotion arises from the Th-1=Th-2 paradigm (24) in whichTh-1 and Th-2 derived cytokines negatively affect the expres-sion and activity of each other. In particular, the Th-1 cyto-kine IFN-g blocks the production of IL-4 or IL-13 by Th-2cells and vice versa. In the presence of iron, IFN-g activityis impaired and the negative inhibitory effect of this cytokineon Th-2 cytokine formation is weakened; thus, the Th-2 effec-tor branch with its anti-inflammatory and deactivatingactions is strengthened (25–27)—an unfavorable conditionin case of a tumor or infectious disease.

The practical relevance of these immunomodulatoryeffects of iron on innate and cellular immune effector functioninvolving the Th-1=Th-2 pathways has been supported byvarious clinical data and in vitro observations. For example,in children with cerebral malaria, a severe complication ofP. falciparum malaria with high lethality, the addition ofthe iron chelator desferrioxamine to a standard antimalarialtreatment resulted in an improved clinical course as reflectedby a shorted duration of coma and fever and an increasedclearance of plasmodia from the circulation (28). Interest-ingly, the children receiving desferrioxamine had higherlevels of Th-1 cytokines and NO, while serum concentrationof Th-2 cytokines (IL-4) was lower (26). This suggests thatwithdrawal of iron may increase Th-1 mediated immune func-tion also in vivo. However, the additional treatment withdesferrioxamine did not result in an improved survival (29),which may also be referred to the fact that desferrioxamineis a hydrophilic compound with a weak capacity to cross mem-branes and affect intracellular targets (30,31).

Other examples of infections, where iron substitution oroverload has a well-documented aggravating potential,include HIV, hepatitis C, tuberculosis, parasitic diseases(e.g., Leishmania, Pneumocystis carinii), bacterial infectionswith gram-positive and gram-negative rods, and fungal infec-tions (e.g., Candida) (reviewed in Ref. 2 and in the chaptersby Drs. Weinberg and Boelaert). For some of these mainlyintracellular infections, the progression of the disease could

562 Weiss

be related to a loss of the macrophages’ ability to kill thepathogens by cytokine-dependent (mostly IFN-g) effectorpathways. Moreover, since immune cells are of different sus-ceptibility to iron restriction or iron overload, changes in ironavailability affect B-, T- or NK-cell expression and differentia-tion especially in the relative expansion between Th-1 vs.Th-2 and CD4þ vs. CD8þ cells in vitro and in vivo (32–34;please refer also to the chapter on ‘‘Iron and immunity’’ in thisbook). In addition, the increased availability of iron feeds thepathogens, which have developed sophisticated strategies toacquire iron for their growth (35,36).

Moreover, iron also has negative effects towards the‘‘first line of cellular host defense,’’ the neutrophil granulo-cyte. This assertion is supported by the finding that iron ther-apy of chronic hemodialysis patients impairs the potential ofneutrophils to kill bacteria and reduces the capacity to phago-cytose foreign particles and by the observations of others thatiron overload in vitro and in vivo results in neutrophil dys-function (37,38). Thus, excess iron may aggravate the clinicalcourse of infectious and malignant disease as outlined below.

SEVERE ACD IS ASSOCIATED WITH A POORPROGNOSIS: DOES THIS IMPLY THAT ANEMIACORRECTION MAY BE BENEFICIAL?

The treatment and correction of ACD has various beneficialeffects as outlined in the previous chapter by Dr. Ludwigand colleagues. However, being aware of the fact that thedevelopment of ACD may partly result from a defense strat-egy of the body, one has to ask whether the correction ofACD by various treatment procedures may also exert nega-tive side effects towards the clinical course of the underlyingdisease or via the induction of complications.

The rationale for correction of anemia is based on at leasttwo pillars. First, the development of anemia has many nega-tive effects for the body such as reduced cardiac performance,reduced physical activity, and fatigue, which would warrantcorrection especially when the patients are older and have


secondary diseases such as coronary heart disease, chronicrenal failure, and diabetes mellitus (39,40). Second, it is wellestablished that a more advanced anemia is associated with apoorer prognosis at least in malignant diseases (41–43), and itwas therefore hypothesized that correction of anemia mayimprove the prognosis of the underlying disease. This hypoth-esis is not proven so far by prospective, randomized controlledtrials in patients with infectious or malignant diseases. Theteaching that correction of anemia may ameliorate the under-lying disease is misleading (43). A more severely anemic ACDpatient represents a more advanced stage of the tumor orinfectious disease (in case of HIV), and the low hemoglobinlevels are probably not the reason for a poor prognosis butrather a reflection of an uncontrolled malignant or infectiousdisease and an exhausted immunity. By escaping from thecontrol of the immune system, the tumor further proliferateswhile the immune system tries to regain control by furtherstimulation of T-cells, NK-cells, or macrophages, which canbe monitored by determination of circulating cytokine andimmune activation markers. Increased amounts and activ-ities of circulating cytokines such as TNF-a, IL-1, IL-6, orIFN-g then exert their effects on erythroid cell proliferation,erythrophagocytosis, iron transport, and the acute phaseresponse in the liver including hepcidin production and ery-thropoietin formation, thus further reducing hemoglobinlevels and worsening the anemia. These interactions havebeen documented in the literature by the observation of anegative association between hemoglobin levels and theamounts of circulating cytokines and immune effector mole-cules such as sTNF-rec I and II, TNF-a, IFN-g, or neop-terin (44–47). Neopterin is a soluble pyrazine-pyramidinoderivative, which is produced and released by activatedmonocytes=macrophages upon IFN-g mediated induction ofGTP-cyclohydrolase I. Neopterin has turned and the mole-cule. It has thus turned out be an excellent and easy-to-mea-sure parameter to monitor cellular immune activity in vitroand in vivo (9). Secondly, the amounts of circulating neopterinor TNF-a have turned out to be positively associated with apoor clinical course in several malignant diseases such as

564 Weiss

colon carcinoma, lung cancer, and ovarian carcinoma, and thepredictive value of these parameters was superior to TNM stageor the presence=absence of distant metastasis (44,47–50). Thesedata support the hypothesis that by overcoming the control bythe immune system a tumor is enabled to proliferate while theimmune system tries to counteract these activities by furtherstimulating immune effector pathways as reflected by increasedTNF-a and neopterin levels. Immune activation then alsocauses a more severe ACD in order to limit malignant cell pro-liferation and oxygen tension also being aware of the fact thata more progressed anemia has many detrimental effects as out-lined in the previous chapter. Thus, the sole correction of ane-mia is unlikely to improve the prognosis of the underlyingdisease at this stage since it is a reflection of an insufficient orexhausted immune response against the underlying malignantor chronic infectious diseases. On the other hand, increasing thehemoglobin concentration in cancer patients may affect theresponse to therapy as it can positively influence the numberof chemotherapy cycles a patient will receive or affect radio-sen-sitivity as outlined later in this chapter.

Nonetheless, treatment of anemia increases oxygentension, which may primarily be of benefit for rapidly prolifer-ating tissues and pathogens. Moreover, an increasedavailability of iron will feed pathogens and malignant cellsand finally, a negative effect of ACD correction on immuneeffector function weakens the already impaired immunityand counteracts the immune control of pathogen prolifera-tion. It is important to note that the effect of ACD therapydepends on the underlying disease. While the points stressedabove may hold true for ACD in malignant and some infec-tious diseases, the situation with ACD in autoimmune dis-eases such as rheumatoid arthritis or Crohn’s disease is justthe opposite (51,52). In the later situation, a negative effectof ACD therapy on immune function is desirable since itmay improve the pathology of the underlying disease byweakening the misguided activation primarily of the Th-1mediated immune effector branch. Finally, the situation inHIV may be different from most other infections as outlinedin the chapter of Dr. Boelart. While iron and transfusion


appear to be detrimental, the correction of anemia by erythro-poietin may be of clinical benefit since slowing down T-cellsactivity has previously been shown to hamper the progressionof disease (53–55).

POTENTIAL HAZARDS OF SPECIFICTHERAPEUTIC REGIMEN

Iron

As stressed in the previous chapters, iron has multiple effectsthroughout the body and both iron deficiency and iron over-load are associated with adverse effects. ACD patients sufferfrom a relative or functional iron deficiency related to thediversion of iron from the circulation to storage sites of thereticuloendothelial system (RES), where iron is incorporatedinto ferritin and hemosiderin and thus is neither availablefor erythropoiesis nor as for the growth of pathogens (seethe chapter by Drs. Mulero and Brock in this book). Thisdiversion of iron is primarily caused by pro- and anti-inflam-matory cytokines that upregulate the expression of proteinsfor iron uptake, such as divalent metal transporter-1 (DMT-1)or transferrin receptor, and downregulate the iron exportprotein, ferroportin-1 on macrophages (56–58). Thus, aftererythrophagocytosis, iron is effectively stored by macrophagesand not recycled (59). In addition, under chronic inflamma-tory conditions, the uptake of ferrous iron by the duodenumis reduced due to impaired expression of DMT-1 in the duode-nal enterocyte. In this context, the recent identification of thecytokine and iron inducible acute phase protein, hepcidin, haswidened our knowledge since this peptide is able to modulateduodenal iron uptake and presumably also macrophage ironrelease (for review see Ref. 60).

Thus, administration of iron to subjects with ACD willnot make sense if the idea behind this therapeutic regimenis to correct the functional iron deficiency. First, iron will bepoorly absorbed due to downregulation of the iron absorptionmachinery in the duodenum, a process that may be controlledby the acute phase protein hepcidin (60). Only a minimum

566 Weiss

of the small amounts of iron taken up after oral administra-tion or iron given intravenously will reach the sites of erythro-poiesis. This is due to the fact that, based on the effective irondiversion strategy, the metal is directed into the storage sitesof the RES—a process with is supported by a-1-antitrypsin,another acute phase protein. a-1-Antitrypsin binds to trans-ferrin receptors of erythroid progenitor cells and thus blocksiron uptake into this compartment leading to arrest of growthand differentiation of BFU-e and CFU-e (61). At the sametime, a-1-antitrypsin has no effect on iron accumulation bycells of the RES (for reviews, see Refs. 62, 63). Moreover,tumor cells and micro-organism have developed sophisticatedstrategies to acquire iron (35,36) that are superior to those oferythroid progenitor cells. Thus, iron not immediately clearedfrom the circulation by RES will be captured by thesepathogens and used for their growth and proliferation.

Although not being effective in improving anemia inACD patients suffering from infectious or malignant disease,iron has multiple side effects. These include the formation ofhighly toxic hydroxyl radicals via the catalytic action of themetal by the Haber–Weiss reaction (64), which cause tissuedamage, endothelial dysfunction, an increase in the risk foracute cardiovascular events (65,66), and the promotion carci-nogenesis via malignant cell transformation (67–69). Studieshave shown that increased iron availability in serum or tis-sues is associated with an increased risk for several tumors(70,71). Moreover, in African iron overload, a strong ass-ociation between iron accumulation in the RES and theincidence of hepatocellular carcinoma has been found (72).Although in hereditary hemochromatosis the RES is relativelyiron depleted, functional defects of monocytes=macrophageshave been detected (59) and hemochromatotic patients appearto be at a higher risk to develop cancer (73,74).

The carcinogenic potential of iron is further promoted byits deactivating effects on cellular immune effector functionand in being a growth factor for rapidly proliferatingtissues (69). Hyperoxia—e.g., as a result of excessive anemiacorrection—decreases iron uptake by alveolar macrophages,reduces the incorporation of iron into ferritin, and increases


iron-mediated hyperoxic injury towards other cells (75). Byaffecting immunity, iron therapy also increases the risk ofinfectious complications or septicemia in ACD patients (76).Iron therapy of chronic hemodialysis patients for the correc-tion of renal anemia impairs the potential of neutrophils tokill bacteria and reduces their capacity to phagocytose foreignparticles, and iron overload in vitro and in vivo results inneutrophil dysfunction (77–79).

Thus, limitation of iron availability as it occurs in ACDappears to be beneficial to better fight the underlying disease.This assertion is supported by the finding that in certaininfectious diseases as well as in cancer, the pathophysiologi-cal or pharmacological withdrawal of iron may harbor clinicalpromise (80–82). In conclusion, as long as prospective studieswill provide clear information on iron therapy, iron should benever given alone to ACD patients suffering from a malignantor infectious disease. This therapeutic intervention may beassociated with an unfavorable clinical course and presum-ably with an increased risk of fatal disease due to the reasonsdescribed above (Table 1).

However, due to its immune deactivation effects, irontherapy may be of benefit in ACD in connection with autoim-mune disorders, especially when used in combination withrecombinant erythropoietin (51,52). ACD patients sufferingfrom true iron deficiency may also benefit from iron therapy.Examples are patients with ACD who have additionalbleeding, e.g., due to gastro-intestinal tumors. Since the total

Table 1 Putative Negative Effects of Iron Therapy in ACD

Negative effect on cell mediated immune effector functiona

Tissue damage via hydroxyl radical formationIn being an essential growth factor stimulation of tumor cell and micro-organism growth by counteracting the iron withholding strategya

May not reach the sites of erythropoiesis due to effective iron diversionstrategy of the immune system

Endothelial dysfunction and increased risk for cardiovascular eventsPoorly absorbed from the gut during ACD

aThese effects relate to ACD on the basis of an infectious or malignant disease.

568 Weiss

lack of available iron also has detrimental effects on immu-nity, especially in connection with immune cell differentiationand proliferation, the relevant question is of how to correctabsolute iron deficiency. Since iron is poorly absorbed underchronic inflammatory conditions, iron given intravenouslymight be the appropriate route of administration. As statedabove, there are some reservations against such a procedure,however, the advantages of iron supplementation in patientswith absolute iron deficiency may be superior to the poten-tial negative effects detailed above (82a), a notion whichhas to be clarified in prospective clincial trials.

Transfusion

The application of red blood cell transfusions is an efficienttherapeutic strategy to rapidly correct low hemoglobin levelsin severely anemic ACD patients. However, different studiesraised the concern that blood transfusions especially in can-cer patients may be associated with an unfavorable clinicalcourse. In surgical intervention studies where the effect ofperioperative transfusion on long-term survival was investi-gated in patients who underwent surgery for gastric cancer(83) or esophageal cancer (84), or who underwent coronarybypass surgery (85), blood transfusions during surgerywere associated with an unfavorable clinical course. In mostof these studies, the negative effect on survival of bloodtransfusion was not due to recurrence of cancer but ratherrelated to other factors such as organ dysfunction (86). Atleast in one study, the depletion of leukocytes from theblood transfused perioperatively had no effect on long-termsurvival (87).

Several factors may contribute to a negative associationbetween transfusion and the clinical outcome.

i. Perioperative blood transfusions increase the inci-dence of infectious complications even after elec-tive operations (88). For each pack of blood cellstransfused, the odds ratio to develop a nosocomialinfection was increased by 1.5-fold—a risk withwas aggravated by patients’ ages (89). Although,


in the Western world transfusions are extensivelychecked for the presence of infectious agents, thereis still a minimal risk for transfusion of new or notyet diagnosed micro-organisms.

ii. In a cohort of almost 38,000 women, those subjectswho ever received a blood transfusion had an up to2.7-fold higher risk to develop low-grade non-Hodgkin lymphoma as compared to women whohave never been transfused. This association wasindependent from social, dietary, or behavioralstatus (90).

iii. Increased amounts of transfused blood increase theiron burden of the body, resulting in the detrimen-tal effects of iron therapy listed above (Table 1).

iv. Blood transfusion affects the patient’s immunestatus other than by modulating iron homeostasis(91). Red blood cell transfusions containing lym-phocytes may in rare cases lead to a low-grade‘‘graft vs. host disease’’ that negatively influencesimmune surveillance (92). Administration of bloodtransfusions to gastric cancer patients resultedin a significant change in the CD4=CD8 ratio ascompared to nontransfused cancer patients (93).Administration of blood transfusions to HIV-positive subjects may induce immune disturbancesthat have devastating effects (53).

v. The storage time of the transfused blood may be ofimportance. In a Danish study of 740 patientsundergoing surgery for colorectal carcinoma, themean survival was 4.6 years in nontransfused sub-jects and 3.0 years in perioperatively transfusedsubjects. Interestingly, the mean survival oftransfused patients receiving blood, which hadbeen stored for less than 21days, was 2.5 yearswhile it was 3.7 years when the blood was storedlonger prior to transfusion. The hazard ratio of dis-ease recurrence after curative resection was 1.5 insubjects transfused with blood stored for a short

570 Weiss

period of time as compared to nontransfused sub-jects or the other transfusion group (94).

vi. The timing of the blood transfusion may affect theclinical outcome. The administration of bloodtransfusions in the first 8 days after mastectomyshortened the metastasis free survival of breastcancer patients (95).

vii. In rare cases, blood transfusions may induce severeallergic reaction or lead to cardiac decompensation,especially when applied too rapidly to patientswith pre-existing cardiac insufficiency.

A caveat is in order. Some of the studies referred to abovemay be biased by the fact that a more sustained anemia maybe a reflection of a more advanced disease as outlined in para-graph 2 of this chapter. Thus, such patients would have aworse prognosis per se and the transfusion of blood for thecorrection of anemia may then not account for the negativeclinical outcome.

Blood transfusions are widely used as a rapid andeffective therapeutic intervention. They may be beneficialfor saving the lives of patients, especially in the context ofsevere anemia due to the complication of ACD by another con-dition (as outlined in the chapter by Dr. Vamvakas in thisbook). Future studies should address the impact of red bloodcell transfusion on the clinical outcome of ACD patients andhow leukocyte depletion of transfusions may affect the courseof the underlying disease.

Recombinant Human Erythropoietin

Therapy with recombinant human erythropoietin (Epo) is awidely used approach to treat ACD, and several studies haveshown that Epo therapy results in an increase of hemoglobinlevels in ACD patients with autoimmune, infectious, orneoplastic diseases (96,97). Accordingly, the increase in hemo-globin levels was associated with a decreased need for bloodtransfusions and an increased quality of life for patients, anissue with increasing importance in clinical case management(summarized in this book in Chapters 14, 17, and 18


Lichtin, Goodnough and Ludwig). The rational for Epo ther-apy is based on the fact that Epo levels appear to be toolow for the degree of anemia in ACD when compared to irondeficiency anemia in some (98) but not all studies (99,100).Moreover, cytokines such as TNF-a or IFN-g have negativeeffects on endogenous Epo formation and activity in ACD(101,102).

Although Epo is widely used as a therapeutic agent, itsmechanism of action in ACD has not been clearly elucidated(103). First of all, Epo is a cytokine that binds to specificreceptors and induces a signal transduction network in thecell leading to transcriptional and post-translation regulationof target genes (as outlined in the chapter by Dr. Jelkmannand colleagues in this book). After binding to its surface recep-tor, Epo stimulates signal transduction molecules and tran-scription factors, which are also induced or deactivated byother specific cytokines. Another possibility is that Epo maycounteract antiproliferative effects of cytokines on erythroidprogenitor cells (101,104) with stimulation of BFU-e andCFU-e differentiation and proliferation. A third possibility isthat Epo may induce hemoglobin synthesis in erythroid pro-genitors via stimulation of TfR expression with subsequentuptake of iron and by induction of porphyrin biosynthesisleading to the formation of heme (105–107). This is supportedby the observation that a poor response to Epo treatment isassociated with increased levels of proinflammatory cytokineson the one hand and a poor iron availability, as determinedby increased Zn-protoporphyrin levels on the other hand(108–111).

Thus, Epomay have specific effects (immune-modulation,effects on proliferation, and differentiation of cells) apartfrom stimulation of erythropoiesis. Such effects may be ofimportance in connection with the therapeutic applicationof Epo since this cytokine=hormone may harbor both putativepositive as well as detrimental effects towards the diseaseunderlying ACD. A consideration of these effects will nowfollow.

First, Epo may exert immunomodulatory effects by inter-fering with the signal transduction pathways of the cytokine-

572 Weiss

cascade. Erythropoietin activates three members of the signaltransducer and activator of transcription (STAT) family.Especially, the activation of STAT1a, which is also inducedby Type I and II interferons, may influence the effectivity ofEpo in inducing erythroid progenitor cell differentiation(112). In patients with end stage renal disease, the adminis-tration of Epo had significant effects on cytokine levels.Long-term administration of Epo decreased TNF-a levels insuch patients, and good responders to Epo therapy (113)had significantly higher CD28 expression on T-cells andreduced IL-10, IL-12, IFN-g, and TNF-a levels compared topoor responders (114). Such anti-inflammatory effects of thehormone may be of benefit as has been shown in rheumatoidarthritis, where combined treatment with Epo and iron notonly increases hemoglobin levels but also results in a reduc-tion of disease activity (51). Also, Epo protects themyocardiumfrom ischemia=reperfusion injury by its anti-inflammatoryeffects (115) and is beneficial in autoimmune encephalitis(116) possibly by its antiapoptotic effects (117). Moreover,Epo therapy increases the expression of the complement regu-latory proteins, decay accelerating factor and CD59, whichrestrict complement activation and inhibit hemolysis (118).

Erythropoietin therapy exerts additional beneficialeffects not linked to hemoglobin correction in diseases withan autoimmune background or where deactivation of theimmune system (e.g., after stroke or myocardial infarction)is desirable. On the other hand, it is unclear whether theseanti-inflammatory effects, which especially modify Th-1mediated immune pathways, are detrimental in infectiousor malignant disease by counter-acting the immunologicalcontrol and defense strategy against the pathogen and=or byinhibiting apoptosis of malignant cells.

Several tumor cells and cell lines have been shown toeither produce Epo and=or to express EpoR on their surface.Interestingly, the effect of Epo on such cells is contradictory.Erythropoietin therapy may be beneficial in multiple mye-loma since Epo administration led to tumor regression in amurine myeloma model (119), but the opposite effect wasdescribed in a case report (120). EpoRs have been detected


on many malignant cell lines, including mammary, ovarian,uterus, prostate cancer, hepatocellular, and renal carcinomaas well as on myeloid cell lines (121–126). In a study investi-gating EpoR expression on leukemia cells from 150 patientswith acute myeloid or lymphoblastic leukemia, the authorsfound that 60% of AML cells expressed EpoR and in 16% ofthese, the proliferation of cells could be stimulated by Epotreatment in vitro. Interestingly, all patients with FAB-M6expressed EpoR. Patients with both, EpoR expression and invitro response to Epo, had a shorter duration of completeremission than those without EpoR (P¼ 0.005, 127).

Administration of Epo to EpoR-expressing human renalcarcinoma cells in vitro stimulates their proliferation; thisobservation might have implications on the clinical course ofpatients with hypernephroma (121). Even more importantly,the functional significance of EpoR expression for the courseof the malignant diseases has been shown for tumors of thefemale reproductive tract, such as breast, ovarian, and uter-ine cancer (123–126). High amounts of EpoR expression werefound in 90% of biopsies from human breast carcinomapatients and in 60% of these, concomitant production of Epoby the tumor could be detected (124). In a rat mammary ade-nocarcinoma model, the inhibition of EpoR mediated signaltransduction resulted in a delay of tumor growth and a 45%reduction in maximal tumor depth (125). The production ofEpoR and Epo by breast cancer cells appears to be regulatedby hypoxia, and in clinical specimens of breast carcinoma, thehighest levels of EpoR were associated with neoangiogenesis,tumor hypoxia, and infiltrating tumors (128). Thus, Epo sig-naling may contribute to the promotion of human cancer byhypoxia (123,124,128,129). In a line with these observations,a recent study investigating the effect of Epo administrationin metastatic mammalian cancer was terminated by theindependent reviewing committee because of higher mortalityamong patients receiving Epo (PRT=EPO-INT76 study). Thepurpose of this double blind, randomized, placebo controlledstudy was to evaluate the impact of maintaining hemoglobinusing Epo-alpha in metastatic breast cancer subjects receiv-ing chemotherapy. A total of 938 patients were investigated;

574 Weiss

at the time of study termination, 101 patients had died in theEpo group vs. 78 in the placebo branch (130).

In a study with 24 different tumor cell lines, the authorsfound that malignant cells secret small amounts of Epo andthat production is increased by hypoxic stimuli (126). More-over, implantation of EpoR expressing cell lines into nude micewith subsequent inhibition of EpoR signaling by using an EpoRantagonist resulted in inhibition of angiogenesis and destruc-tion of tumor masses. In contrast, Epo treatment promotedangiogenesis and tumor survival. Thus, in Epo-sensitive celllines, the inhibition of the Epo signaling pathway could be apromising therapeutic target to control malignancy (126).

Part of the effect of Epo in promoting malignant cellgrowth may be related to the fact that Epo can directly stimu-late neoangiogenesis. A recent investigation showing that Epoincreases inflammation and ischemia induced neovasculariza-tion by enhancing endothelial progenitor cell mobilization(131). In addition, Epo stimulated the proliferation of capil-laries in myocardial tissue in a comparable magnitude toVEGF (132). Whether or not Epo also effects VEGF produc-tion remains to be shown. However, the angiogenic effect ofEpo may be of benefit in cardiology to promote neovasculari-zation following myocardial infarction or bypass surgery. Incontrast, Epo induced neoangiogenesis is an unfavorable con-dition in malignancy since this may further promote tumorgrowth and the development of metastases.

Erythropoietin formation by tumor cells may be of impor-tance for tumor vascularization in the following way. Hypoxiastimulates Epo production by tumor cells, which then pro-motes neoangiogenesis. Thus, the survival and growth ofEpoR-positive tumor cells could be supported via the hor-mone’s antiapotoic and immunomodulatory effects, and byits stimulatory potential towards tumor cell growth andneovascularization (126,128,130,131).

Therefore, in the near future, it may be essential to assessthe EpoR status of tumor cells to avoid potential devastatingand life shortening effects of ACD therapy with Epo in suchpatients. It may turn out that Epo is not an appropriate ther-apy for ACD in all malignant diseases or that the hormone


may even be contra-indicated in connectionwith some tumors.Moreover, the clinical impact of the neoangiogenic effects ofEpo and its immunomodulatory potential should be assessedas soon as possible in large, double blinded, and multicenterplacebo controlled studies. Such studies are of pivotal impor-tance for the well being of our patients in order to gather moreinformation on the net results of Epo therapy on the clinicalcourse of the diseases underlying ACD (Table 2).

Other possible side effects of Epo, such as pure red cellaplasia, hypertension, or increased risk for thrombosis, arediscussed elsewhere in this book (please refer to the chaptersof Dr. Goodnough and Drs. Kaur and Lichtin).

GENERAL CONSIDERATION CONCERNINGANEMIA CORRECTION IN ACD

As pointed out in this chapter, the therapeutic armoryused for ACD correction may exert divergent effects on the

Table 2 Effects of Epo Not Linked to Erythropoiesis and itsPossible Clinical Impact

� Anti-inflammatory action by down-regulating Th-1 mediated immuneeffector function, increasing complement deactivation and inhibitingapoptosis

Positive: in autoimmune disorders, immune disturbances such as MI,stroke

Putative negative in infectious and malignant diseases by counteractingimmune control

� Many tumor cells express EpoR on their surfacePositive: Epo treatment may restrict tumor growth in multiple myelomaNegative: Epo functionally induces proliferation and=or malignanttransformation of, e.g., mammary carcinoma, ovarian carcinoma,myeloid leukemic cells and may thus favor carcinogenesis and lead toa detrimental clinical course

� Epo stimulates angiogenesis by promoting endothelial progenitor cellmobilization and capillary growth

Positive: neovascularization after myocardial infarction or bypasssurgery

Negative: Epo mediated stimulation of angiogenesis may favor tumorgrowth

576 Weiss

underlying disease. What is beneficial in autoimmune andrheumatic diseases may turn out to be detrimental for thetreatment of ACD caused by malignant or chronic infectiousdiseases.

The association of anemia with a poor prognosis in malig-nant disease appears to reflect an insufficient=exhaustedimmune response indicating that the tumor has overcomethe control by the immune system. In turn, the immune sys-tem tries to keep the balance by further inducing the stimula-tion of T-, NK-cells, and macrophages with one of theconsequences being a worsening of ACD. In this case, correc-tion of ACD per se may not positively influence the underlyingbattle between malignant cells, microbes, and immunity.

On the other hand, the effectiveness of radiation therapyfor malignancy may be improved by an increased oxygen ten-sion in tissues since more oxygen increases the antitumor effectof radiation therapy via radical formation within neoplastic tis-sues (133–135). Accordingly, a beneficial effect of higher hemo-globin levels has been found for the use of radiation therapy inhead and neck tumors (39). In contrast, a recent double blindrandomized study provided evidence that the rise of hemoglo-bin as a consequence of therapy with rhEpo was associated witha poorer survival in patients with head and neck cancer as com-pared to cancer patients who did not receive rhEpo (135a.)

Anemia induces expression of VEGF that in turn promotesneoangiogenesis and may thus favor vascularization of tumortissues (42,136). Correction of anemia by Epo may also promoteneoangiogenesis and tumor vascularization (131,132).

In conclusion, it will be essential to determine in futurewell-designed, prospective studies and the desired therapeuticendpoint for different diseases, i.e., what is the optimal hemoglo-bin level in ACD patients, especially those suffering from amalignant disease. These trials will need to consider the clinicalcourse of theunderlyingdisease, the balance and efficiency of theimmune response, the quality of life, and cardiac performanceand renal function. A recent study in subjects with chronic renalfailure found that hematocrit levels between 33% and 36% wereassociated with the best outcomes in terms of mortality andmorbidity (41), while both higher and lower hematocrits are


associated with an unfavorable clinical course at least in endstage renal disease (40). We have to be aware that not all thera-peutic procedures may be warranted under all conditions withACD. Inmalignant diseases, the evaluation ofEpoR in tumor tis-sue may help to identify those patients for whom Epo therapy inconjunction with a cautious use of iron may be beneficial.

Only prospective, randomized, double blinded, multicen-ter studies can help us to decide, which therapeutic regimenis the best in ACD depending on the underlying disease, towhich extent ACD should be corrected, what is the net effectof ACD correction on the clinical course of the underlying dis-ease, and if such an effect can be referred to improvement ofhemoglobin levels or to another action of the therapeutic sub-stance (e.g., iron, erythropoietin, transfusion). New insightsinto the pathophysiology of ACD may also open the door tonew therapeutic principles, such as antagonizing hepcidinactivity in the circulation, modulating macrophage ironrelease or duodenal iron absorption via ferroportin-1, or inhi-biting signal transduction for Epo or VEGF in responsivemalignant cells.

The future of ACD therapy appears to be one of the mostexciting fields in modern medicine when one considers thehigh incidence of ACD and how little we know about the effectof anemia correction on the course of the underlying disease.

ACKNOWLEDGMENT

The continuous support by grants from the Austrian Res-earch Funds and the Austrian National Bank is gratefullyacknowledged.

REFERENCES

1. Weiss G. Iron acquisition by the reticuloendothelial system.In: Templeton D, ed. Molecular and Cellular Iron Transport.New York: Marcel Dekker, 2002:467–488.

2. Weinberg ED. Iron loading and disease surveillance. EmergInfect Dis 1999; 5:346–350.

578 Weiss

3. Choi KS, Bae MK, Jeong JW, Moon HE, Kim KW. Hypoxia-induced angiogenesis during carcinogenesis. J Biochem MolBio 2003; 36:120–127.

4. Gordon S. Alternative activation of macrophages. Nat Immu-nol Rev 2003; 3:133–146.

5. Weiss G, Fuchs D, Hausen A, Reibnegger G, Werner ER,Werner-Felmayer G, Wachter H. Iron modulates interferon-g effects in the human myelomonocytic cell line THP-1. ExpHematol 1992; 20:605–610.

6. Recalcati S, Pometta R, Levi S, Conte D, Cairo G. Response ofmonocyte iron regulatory protein activity to inflammation:abnormal behavior in genetic hemochromatosis. Blood 1998;91:2565–2572.

7. Oexle H, Kaser A, Most J, Bellmann-Weiler R, Werner ER,Werner-Felmayer G, Weiss G. Pathways for the regulationof interferon-gamma-inducible genes by iron in human mono-cytic cells. J Leukoc Biol 2003; 74:287–294.

8. Taylor MW, Feng GS. Relationship between interferon-gamma, indoleamine 2,3-dioxygenase, and tryptophancatabolism. FASEB J 1991; 5:2516–2522.

9. Murr C, Wiedner B, Wirleitner B, Fuchs D. Neopterin as amarker for immune system activation. Curr Drug Metab2002; 3:175–187.


11. Byrd TF, Horwitz MA. Lactoferrin inhibits or promotesLegionella pneumophila intracellular multiplication innonactivated and interferon-g activated human monocytesdepending upon its degree of iron saturation: iron-lactoferrinand nonphysiologic iron chelates reverse monocyte actionagainst Legionella pneumophila. J Clin Invest 1991; 88:1103–1112.

12. Barnewell RE, Rikihisa Y. Abrogation of gamma interferon-induced inhibition of Ehrlichia chaffeensis infection inhuman monocytes with transferrin iron. Infect Immun1994; 62:4804–4810.


13. Alford CE, King TE, Campell PA. Role of transferrin, trans-ferrin receptors, and iron in listericidal activity. J Exp Med1991; 174:459–466.


15. MacMicking J, Xie QW, Nathan C. Nitric oxide and macro-phage function. Annu Rev Immunol 1997; 15:323–350.

16. Bogdan C. Nitric oxide and the immune response. Nat Immu-nol 2001; 2:907–916.

17. Mellilo G, Taylor LS, Brooks A, Musso T, Cox GW, Varesio L.Functional requirement of the hypoxia responsive element inthe activation of the inducible nitric oxide synthase promoterby the iron chelator desferrioxamine. J Biol Chem 1997;272:12236–12242.

18. Dlaska M, Weiss G. Central role of transcription factor NF-IL6 for cytokine and iron-mediated regulation of murineinducible nitric oxide synthase expression. J Immunol 1999;162:6171–6177.

19. Forbes JR, Gros P. Divalent metal transport by NRAMP pro-teins at the interface to host–pathogen interaction. TrendsMicrobiol 2001; 9:397–403.

20. Blackwell JM, Searle S, Goswami T, Miller EN. Understand-ing the multiple functions of NRAMP1. Microbes Infect 2000;2:317–321.

21. Baker ST, Barton CH, Biggs TE. A negative autoregulatorylink between Nramp1 function and expression. J Leukoc Biol2000; 67:501–507.

22. Kuhn DE, Baker BD, Lafuse WP, Zwilling BS. Diff-erential iron transport into phagosomes isolated from theRAW264.7 macrophage cell lines transfected with NRAMP-1Gly169 or NRAMP1Asp169. J Leukoc Biol 1999; 66:113–119.

23. Biggs TE, Baker ST, Bothman MS, Dhital A, Barton HC,Perry HV. Nramp1 modulates iron homeostasis in vivo andin vitro: evidence for a role in cellular iron release involving

580 Weiss

de-acidification of intracellular vesicles. Eur J Immunol 2001;31:2060–2070.

24. Romagnani S. The Th1=Th2 paradigm. Immunol Today 1997;18:263–267.

25. Thuma PE, Weiss G, Herold M, Gordeuk VR. Serumneopterin, interleukin-4, and interleukin-6 concentrationsin cerebral malaria patients and the effect of iron chelationtherapy. Am J Trop Med Hyg 1996; 154:164–168.

26. Weiss G, Thuma PE, Mabeza F, Werner ER, Herold M,Gordeuk VR. Modulatory potential of iron chelation therapyon nitric oxide formation in cerebral malaria. J Infect Dis1997; 175:226–230.

27. Mencacci A, Cenci F, Boelaert JR, Bucci P, Mosci P, Fed’Ostiani C, Bistoni F, Romani L. Iron overload alters innateand T helper cell responses to Candida albicans in mice.J Infect Dis 1997; 175:1467–1476.

28. Gordeuk V, Thuma P, Brittenham G, McLaren G, Parry D,Backenstose A. Effect of iron chelation therapy on recoveryfrom deep coma in children with cerebral malaria. N EnglJ Med 1992; 327:1473–1477.

29. Mabeza GF, Loyevsky M, Gordeuk VR, Weiss G G. Iron chela-tion therapy for malaria: a review. Pharmacol Ther 1998;81:53–75.

30. Loyevsky M, Lytton SD, Mester B, Libman J, Shanzer A,Cabantchik ZI. The antimalarial action of desferal involvesa direct access route to erythrocytic (Plasmodium falciparum)parasites. J Clin Invest 1993; 91:218–224.

31. Hershko C, Gordeuk VR, Thuma PE, Theanacho EN, SpiraDT, Hider RC. The antimalarial effect of iron chelators:studies in animal models and in humans with mild falci-parum malaria. J Inorg Biochem 1992; 47:267–277.

32. De Sousa M, Reimao R, Porto G, Grady RW, Hilgartner MW,Giardina P. Iron and lymphocytes: reciprocal regulatory inter-actions. Curr Stud Hematol Blood Transfus 1992; 58:171–177.

33. Omara FO, Blakley BR. The effects of iron deficiency and ironoverload on cell mediated immunity in the mouse. Br J Nutr1994; 72:899–909.


34. Oppenheimer SJ. Iron and its relation to immunity and infec-tious disease. J Nutr 2001; 131:S616–S633.

35. Eide DJ. Metal ion transport in eukaryotic microorganisms:insights from Saccharomyces cerevisiae. Adv Microb Physiol2000; 43:1–38.

36. Genco CA, White-Dixon D. Emerging strategies in microbialhaem capture. Mol Microbiol 2001; 39:1–11.

37. Patruta SI, Edlinger R, Sunder-Plassmann G, Horl WH. Neu-trophil impairment associated with iron therapy in hemodia-lysis patients with functional iron deficiency. J Am SocNephrol 1998; 9:655–663.

38. VanAsbeck BS, Marx JJ, Struyvenberg A, vanKats JH,Verhoef J. Deferoxamine enhances phagocytic function ofhumanpolymorphonuclearphagocytes.Blood1984;64:714–720.

39. Henke M. Correction of cancer anemia—impact on diseasecourse, prognosis and treatment efficacy, particularly forpatients undergoing radiotherapy. Onkologie 2001; 24:450–454.

40. Murphy ST, Parfrey PS. The impact of anemia correction oncardiovascular disease in end stage renal disease. SeminNephrol 2000; 20:350–355.

41. Collins AJ, Ma JZ, Ebben J. Impact of hematocrit on morbid-ity and mortality. Semin Nephrol 2000; 20:345–349.

42. Thomas GM. Raising hemoglobin. An opportunity to increasesurvival? Oncology 2002; 63(suppl 2):19–28.

43. Munstedt K, Kovacic M, Zygmunt M, Von Georgi R. Impact ofhemoglobin levels before and during chemotherapy on survi-val of patients with ovarian cancer. Int J Oncol 2003; 23:837–843.

44. Fuchs D, Shearer GM, Boswell RN, Lucey DR, Clerici M,Reibnegger G, Werner ER, Zajac RA, Wachter H. Negativecorrelation between blood cell counts and serum neopterinconcentration in patients with HIV-1 infection. AIDS 1991;5:209–212.

45. Feelders RA, Vreugdenhil G, Eggermont AM, Kuiper-KramerPA, van Eijk HG, Swaak AJ. Regulation of iron metabolism in

582 Weiss

the acute-phase response interferon gamma and tumournecrosis factor alpha induce hypoferraemia, ferritin produc-tion and a decrease in circulating transferrin receptors incancer patients. Eur J Clin Invest 1998; 28:520–527.

46. Nakashima J, Tachibana M, Ueno M, Miyajima A, Baba S,Murai M. Association between tumor necrosis factor in serumand cachexia in patients with prostate cancer. Clin CancerRes 1998; 4:1743–1748.

47. Denz H, Huber P, Landmann R, Orth B, Wachter H, Fuchs D.Association between the activation of macrophages, changesof iron metabolism and the degree of anaemia in patientswith malignant disorders. Eur J Haematol 1992; 48:244–248.

48. Reibnegger GJ, Bichler AH, Dapunt O, Fuchs DN, Fuith LC,Hausen A, Hetzel HM, Lutz H H, Werner ER, Wachter H.Neopterin as a prognostic indicator in patients with carci-noma of the uterine cervix. Cancer Res 1986; 46:950–955.

49. Weiss G, Kronberger P, Conrad F, Bodner E, Wachter H,Reibnegger G. Neopterin and prognosis in patients withadenocarcinoma of the colon. Cancer Res 1993; 53:260–265.

50. Murr C, Bergant A, Widschwendter M, Heim K, Schrocks-nadel H, Fuchs D. Neopterin is an independent prognosticvariable in females with breast cancer. Clin Chem 1999; 45:1998–2004.

51. Kaltwasser JP, Kessler U, Gottschalk R, Stucki G. Effect ofrecombinant human erythropoietin and intravenous iron onanemia and disease activity in rheumatoid arthritis. J Rheu-matol 2001; 28:2430–2436.

52. Gasche C, Dejaco C, Waldhoer T, Tillinger W, Reinisch W,Fueger GF, Gangl A, Lochs H. Intravenous iron and erythro-poietin for anemia associated with Crohn disease. A rando-mized, controlled trial. Ann Intern Med 1997; 26:782–787.

53. Sullivan P. Associations of anemia, treatment for anemia,and survival in patients with human immunodeficiency virusinfection. J Infect Dis 2002; 185(suppl 2):S138–S142.

54. Kramer A, Biggar RJ, Hampl H, Friedman RM, Fuchs D,Wachter H, Goedert JJ. Immunologic markers of progression


to acquired immunodeficiency syndrome are time-dependentand illness-specific. Am J Epidemiol 1992; 136:71–80.

55. Gordeuk VR, Delanghe JR, Langlois MR, Boelaert JR JR.Iron status and the outcome of HIV infection: an overview.J Clin Virol 2001; 20:111–115.

56. Ludwiczek S, Aigner E, Theurl I, Weiss G. Cytokine-mediated regulation of iron transport in human monocyticcells. Blood 2003; 101:4148–4154.


58. Tilg H, Ulmer H, Kaser A, Weiss G. Role of IL-10 for induc-tion of anemia during inflammation. J Immunol 2002;169:2204–2209.

59. Moura E, Verheul AF, Marx JJ. A functional defect in heredi-tary haemochromatosis monocytes monocyte-derived macro-phages. Eur J Clin Invest 1998; 28:164–173.

60. Ganz T. Hepcidin, a key regulator of iron metabolism and med-iator of anemia of inflammation. Blood 2003; 102:783–788.

61. Graziadei I, Gaggl S, Kaserbacher R, Braunsteiner H,Vogel W. The acute-phase protein alpha 1-antitrypsin inhi-bits growth and proliferation of human early erythroid pro-genitor cells (burst-forming units-erythroid) and of humanerythroleukemic cells (K562) in vitro by interfering withtransferrin iron uptake. Blood 1994; 83:260–268.

62. Means RT. Advances in the anemia of chronic disease. IntJ Hematol 1999; 70:7–12.


64. Rosen GM, Pou S, Ramos CL, Cohen MS, Britigan BE. Freeradicals and phagocytic cells. FASEB J 1995; 9:200–209.

65. Sullivan JL. Stored iron and myocardial perfusion deficits.Am Heart J 2002; 143:193–195.

66. Gaenzer H, Marschang P, Sturm W, Neumayr G, Vogel W,Patsch J, Weiss G. Association between increased iron stores

584 Weiss

and impaired endothelial function in patients with hereditaryhemochromatosis. J Am Coll Cardiol 2002; 40:2189–2194.

67. Okada S. Prevention of free-radical mediated tissue damageand carcinogenesis induced by low-molecular-weight iron.Biometals 2003; 16:99–101.

68. Lee NT, Richardson DR. The role of iron in cell cycle progres-sion and the proliferation of neoplastic cells. Biochim BiophysActa 2002; 1603:31–46.

69. Weinberg ED. Iron therapy and cancer. Kidney Int 1999;69(suppl):S131–S134.

70. Stevens RG, Jones DY, Micozzi MS, Taylor PR. Body ironstores and the risk of cancer. N Engl J Med 1988;319:1047–1052.

71. Weinberg ED. The role of iron in cancer. Eur J Cancer Prev1996; 5:19–36.

72. Gangaidzo IT, Gordeuk VR. Hepatocellular carcinoma andAfrican iron overload. Gut 1995; 37:727–730.

73. Niederau C, Fischer R, Sonnenberg A, Stremmel W,Trampisch HJ, Strohmeyer G. Survival and causes of deathin cirrhotic and in noncirrhotic patients with primary hemo-chromatosis. N Engl J Med 1985; 313:1256–1262.

74. Fracanzani AL, Conte D, Fraquelli M, Taioli E, Mattioli M,Losco A, Fargion S. Increased cancer risk in a cohort of 230patients with hereditary hemochromatosis in comparison tomatched control patients with non-iron-related chronic liverdisease. Hepatology 2001; 33:647–651.

75. Wesselius LJ LJ, Williams WL, Bailey K, Vamos S, O’Brien-Ladner AR, Wiegmann T. Iron uptake promotes hyperoxicinjury to alveolar macrophages. Am J Respir Crit Care Med1999; 159:100–106.

76. Bullen J, Griffiths E, Rogers H, Ward G. Sepsis: the criticalrole of iron. Microbes Infect 2000; 2:409–415.

77. Patruta SI SI, Edlinger R, Sunder-Plassmann G, Horl WH.Neutrophil impairment associated with iron therapy in hemo-dialysis patients with functional iron deficiency. J Am SocNephrol 1998; 9:655–663.


78. Hoepelman IM, Jaarsma EY, Verhoef J, Marx JJ. Effect ofiron on polymorphonuclear granulocyte phagocytic capacity.Role of oxidation state and effect of ascorbic acid. Br J Hae-matol 1988; 70:495–500.

79. VanAsbeck BS, Marx JJ, Struyvenberg A, vanKats JH, Ver-hoef J. Deferoxamine enhances phagocytic function of humanpolymorphonuclear phagocytes. Blood 1984; 62:714–720.

80. Buss JL, Torti FM, Torti SV. The role of iron chelation in can-cer therapy. Curr Med Chem 2003; 10:1021–1034.

81. Hoen B. Iron and infection: clinical experience. Am J KidneyDis 1999; 34:S30–S34.

82. Weinberg ED. Roles of iron in neoplasia. Biological TraceElement Res 1992; 34:123–140.

82a. Auerbach M, Ballard H, Trout JR, et al. Intravenous ironoptimizes the response to recombinant human erythropoietinin cancer patients with chemotherapy-related anemia: a mul-ticenter, open-label, randomized trial. J Clin Oncol 2004;22:1301-1307

83. Hyung WJ, Noh SH, Shin DW, Huh J, Huh BJ, Choi SH,Min JS. Adverse effects of perioperative transfusion onpatients with stage III and IV gastric cancer. Ann Surg Oncol2002; 9:5–12.

84. Langley SM, Alexiou C, Bailey DH, Weeden DF. The influ-ence of perioperative transfusin on survival after esophagealresection for carcinoma. Ann Thorac Surg 2002; 73:1704–1709.

85. Engoren MC, Habib RH, Zacharias A, Schwann TA, RiordanCJ, Durham SJ. Effect of blood transfusion on long-term sur-vivalafter cardiac operation. Ann Thorac Surg 2002; 74:1180–1186.

86. Vincent JL, Baron JF, Reinhart K, Gattimoni L, Thijs L,Webb A, Meier-Hellmann A, Nollet G, Peres-Bota D. Anemiaand blood transfusion in critically ill paitents. JAMA 2002;288:1499–1507.

87. van de Watering LM, Brand A, Houbiers JG, Klein WM,Hermans J, vande Velde C. Perioperative blood transfusions,

586 Weiss

with or without allogenic leukocytes, relate to survival, not tocancer recurrance. Br J Surg 2001; 88:267–272.

88. Braga M, Vignali A, Radaelli G, Gianotti L, DiCarlo V. Asso-ciation between perioperative blood transfusion and psoto-perative infection in patietns having elective operations forgastrointestinal cancer. Eur J Surg 1992; 158:531–536.

89. Taylor RW, Manganaro L, O’ Brien J, Trottier SJ, Parkar N,Veremakis C. Impact of allogenic packed red blood cell trans-fusion on nosocomial infection rates in critically ill patients.Crit Care Med 2002; 30:2249–2254.

90. Cerhan JR, Wallace RB, Dick F, kemp J, Parker AS, ZhengW,Sellers TA, Folson AR. Blood transfusions and the riskof non-Hodgkin’s lymphoma subtypes and chronic lymphocy-tic leukemia. Cancer Epidemiol Biomarkers Prev 2001;10:361–368.

91. Cunningham-Rundles S, Giardina PJ, Grady RW, Califano C,Mc Kenzie P, DeSousa M. Effect of transfusional iron over-load on immune response. J Infect Dis 2000; 182:S115–S121.

92. Vamvakas EC, Blajchman MA. Deleterious clinical effects oftransfusion associated immunomodulation: fact or fiction?Blood 2001; 97:1180–1192.

93. Murata N, Idezuki Y, Konishi T, Watanabe H, Ushirokoji Y,Shinohara K, Shibusawa M, Haga S, Hiraishi M, Bandai Y,Yamamura T, Yumoto S, Gunji A, Nishigaki K. Influence ofperioperative blood trnasfusion on the prognosis of patietnswith gastric cancer receiving anticancer chemotherapy.Gastric Cancer 2000; 3:24–27.

94. Mynster T, Nielsen HJ. Storage time of transfused blood anddisease recurrence after colorectal cancer surgery. Dis ColonRectum 2001; 44:955–964.

95. Pysz M. Blood transfusions in breast cancer patietns under-going mastectomy: possible importance of timing. J SurgOncol 2000; 75:258–263.

96. Marsh WA, Rascati KL. Meta-analyses of the effectiveness oferythropoietin for end stage renal disease and cancer. ClinTher 1999; 21:1443–1455.


97. Rizzo JD, Lichtin AE, Woolf SH, Seidenfeld J, Bennett CL,Cella D, Djulbegovic B, Goode MJ, Jakubowski AA, Lee SJ,Miller CB, Rarick MU, Regan DH, Browman GP, GordonMS. Use of epoetin in patients with cancer: evidence-basedclinical practice guidelines of the American Society of ClinicalOncology and the American Society of Hematology. J ClinOncol 2002; 20:4083–4107.

98. Miller CB, Jones RJ, Piantadosi S, Abeloff MD, Spivak JL.Decreased erythropoietin response in patients with anemiaof cancer. New Engl J Med 1990; 332:1688–1692.

99. Cazzola M, Ponchio L, deBenedetti F. Defective iron supplyfor erythropoiesis and adequate endogenous erythropoietinproduction in the anemia associated with systemic onset juve-nile chronic arthritis. Blood 1996; 87:4824–4830.

100. Corazza F, Beguin Y, Bergmann P, Andre M, Ferster A,Devalck C, Fondu P, Buyse M, Sariban E. Anemia in childrenwith cancer is associated with decreased erythropoietin activ-ity and not with inadequate erythropoietin production. Blood1998; 92:17993–1798.

101. Means RT, Krantz SB. Inhibition of human erythroid colonyformation by IFN-gamma can be corrected by human recom-binant erythropoietin. Blood 1991; 78:2564–2570.

102. Jelkmann W, Hellwig-Buergel T. Tumor necrosis factor p55receptor (TNF-RI) mediates the in vitro inhibition ofhepatic erythropoietin production. Exp Hematol 1999; 27:224–228.

103. Goodnough LT, Skikne B, Brugnara C. Erythropoietin, ironand erythropoiesis. Blood 2000; 96:823–833.

104. Means RT. Pathogenesis of the anemia of chronic disease:a cytokine mediated anemia. Stem Cells Dayt 1995; 13:32–37.

105. Weiss G, Houston T, Kastner S, Grunewald K, Brock J. Reg-ulation of cellular iron metabolism by erythropoietin: activa-tion of iron regulatory protein and up-regulation of transferrin-receptor in erythroid cells. Blood 1997; 89:680–687.

106. Busfield SJ, Tilbrook PA, Callus BA, Spandaccini A, KuhnLC, Klinken PS. Complex regulation of transferrin receptors

588 Weiss

during erythropoietin-induced differentiation of J2E erythroidcells. Eur J Biochem 1997; 249:77–85.

107. Zoller H, Decristoforo C, Weiss G. Erythroid 5-aminolevuli-nate synthase, ferrochelatase and DMT1 expression in ery-throid progenitors: differential pathways for erythropoietinand iron-dependent regulation. Br J Haematol 2002; 118:619–626.


109. Cooper AC, Breen CP, Vyas B, Ochola J, Kemeny DM, Mac-dougall IC. Poor response to recombinant erythropoietin isassociated with loss of T-lymphocyte CD28 expression andaltered interleukin-10 production. Nephrol Dial Transplant2003; 18:133–140.

110. Goicoechea M, Martin J, de Sequera P, Quiroga Ja, Ortiz A,Carreno V, Caramelo C. Role of cytokines in the response toerythropoietin in hemodialysis patients. Kidney Int 1998;54:1337–1343.

111. Cooper AC, Mikhail A, Lethbridge MW, Kemeny DM, Mac-dougall IC. Increased expression of erythropoiesis inhibitingcytokines by T cells in patients exhibiting a poor responseto erythropoietin therapy. J Am Soc Nephrol 2003; 14:1776–1784.

112. Kirito K, Uchida M, Takatoku M, Nakajima K, Hirano T,Mirura Y, Komatsu N. A novel function of Stat 1 and Stat3proteins in erythropoietin induced erythroid differentiationof a human leukemia cell line. Blood 1998; 92:462–471.

113. Aquilera A, Bajo MA, Diez JJ, Ruiz-Caravaca Ml, Alvarez V,Codoceo R, Selgas R. Effects of human recombinant erythro-poietin on inflammatory status in peritoneal dialysispatients. Adv Perit Dial 2002; 18:200–205.

114. MacDougall IC, Cooper AC. Erythropoietin resistance: therole of inflammation and pro-inflammatory cytokines.Nephrol Dial Transplant 2002; 17(suppl 11):39–43.

115. Calvillo L, Latini R, Kajstura J, Leri A, Anversa P, Ghezzi P,Salio M, Cerami A, Brines M. Recombinant human erythro-


poietin protects the myocardium from ischemia-reperfusioninjury and promotes beneficial remodeling. Proc Natl AcadSci USA 2003; 100:4802–4806.

116. Agnello D, Bigini P, Villa P, Mennini T, Cerami A, Brines M,Ghezzi P. Erythropoietin exerts an anti-inflammatory effecton the CNS in a model of experimental auto-immune ence-phalitis. Brain Res 2002; 952:128–134.

117. Variano M, Dello Russo C, Pozzoli G, Battaglia A, Tringali G,Aloe-Spiriti MA, Preziosi P, Navarra P. . Eur J Neurosci2002; 16:684–692.

118. Ohi H, Tamano M, Sudo S, Okada N. Recombinant EPO ther-apy increases erythrocyte expression of complement regula-tory proteins. Am J Kidney Dis 2003; 41:179–185.

119. MittelmanM, NeumannD, Peled A, Kanter P, Haran-Ghera N.Erythropoietin induces tumor regression and antitumorimmune response in murine myeloma models. Proc NatlAcad Sci USA 2001; 98:5181–5186.

120. Olujohungbe A, Handa S, Holmes J. Does erythropoietinaccelerate malignant transformation in multiple myeloma?Postgrad Med 1997; 73:163–164.

121. Westenfelder C, Baranowski RL. Erythropoietin stimulatesproliferation of human renal carcinoma cells. Kidney Int2000; 58:647–657.

122. Brower V. Normal and neoplastic prostate cells have EPOreceptors. Lancet Oncol 2003; 4:69.

123. Acs G, Acs P, Beckwith SM, Pitts RL, Clements E, Wong K,Verma A. Erythropoietin and erythropoietin receptor expres-sion in human cancer. Cancer Res 2001; 61:3561–3565.

124. Acs G, Zhang PJ, Rebbeck TR, Acs P, Verma A. Immunohis-tochemical expression of erythropoietin and erythropoietinreceptor in breast carcinoma. Cancer 2002; 95:969–981.

125. Arcasoy MO, Amin K, Karayal AF, Chou SC, Raleigh JA,Varia MA, Haroon ZA. Functional significance of erythropoie-tin receptor expression in breast cancer. Lab Invest 2002;82:911–918.

590 Weiss

126. Yasuda Y, Fujita Y, Matsuo T, Koinuma S, Hara S,Onozaki M, Hashimoto M, Musha T, Ogawa K, NakamuraY, Shiozaki H, Utsumi H. Erythropoietin regulates tumorgrowth of human malignancies. Carcinogenesis 2003; 24:1021–1029.

127. Takeshita A, Shinjo K, Higuchi M, Miyawaki S, Takemoto Y,Kishimoto Y, Saito K, Takuchi H, Kuriyama K, Kimura Y,Asou N, Takahashi M, Hotta T, Kanamaru A, Ueda R, OhnoR. Quantitative expression of erythropoietin receptor (EPO-R) on acute leukaemia cells: relationships between theamount of EPO-R and CD phenotypes, in vitro proliferativeresponse, the amount of other cytokine receptors and clinicalprognosis. Br J Haematol 2000; 108:55–63.

128. Acs G, Zhang PJ, McGrath CM, Acs P, McBroom J, Mohyel-din A, Liu S, Lu H, Verma A. Hypoxia-inducible erythropoie-tin signaling in squamous dysplasia and squamous cellcarcinoma of the uterine cervix and its potential role in cervi-cal carcinogenesis and tumor progression. Am J Pathol 2003;162:1789–1806.

129. Harris AL. Hypoxia—a key regulatory factor in tumorgrowth. Nat Cancer Rev 2002; 2:38–47.

130. Leyland-Jones B. Breast cancer trial terminated unexpect-edly. Lancet Oncol 2003; 4:459–460.

131. Heeschen A Aicher C, Lehmann R, Fichtlscherer S, Ulbrich C,Mildner-Rihm C, Martin H, Zeiher AM, Dimmeler S. Erythro-poietin is a potent physiological stimulus for endothelial pro-genitor cell proliferation. Blood 2003 Apr 17 [epub ahead].

132. Jaquet K, Krause K, Tawakol-Khodai M, Geidel S. Erythro-poietin and VEGF exhibit equal angiogenetic potential.Microvasc Res 2002; 64:326–333.

133. Reed WR, Hussey DH, DeGowin RL. Implications of theanemia of chronic disorders in patients anticipating radio-therapy. Am J Med Sci 1994; 308:9–15.

134. Manegold C. The causes and prognostic significance of lowhemoglobin levels in tumor patients. Strahlenther Onkol1998; 174(suppl 4):17–19.


135. Shasha D. The negative impact of anemia on radiotherapyand chemoradiation outcomes. Semin Hematol 2001;38(suppl 7):8–15.

135a. M Henke, R Laszig, C Rube, U Schafer, KD Haase, B Schil-cher, S Mose, KT Beer, U Burger, C Dougherty, H Fromm-hold. Erythropoietin to treat head and neck cancer patientswith anaemia undergoing radiotherapy: randomised, double-blind, placebo-controlled trial. Lancet 2003; 362:1255–1260.

136. Dunst J, Becker A, Lautenschlager C, Markau S, Becker H,Fischer K, Haensgen G. Anemia and elevated systemic levelsof vascular endothelial growth factor (VEGF). StrahlentherOnkol 2002; 178:436–441.

592 Weiss

20

Anemia of Chronic Disease inHematologic Disorders and

Oncology

ROBERT T. MEANS

Hematology Oncology Division,Department of Medicine, Ralph H.

Johnson VA Medical Center and theMedical University of South Carolina,

Charleston, South Carolina, U.S.A.

GORDON D. McLAREN

Hematology Oncology Division,Department of Medicine, University

of California, Irvine and VA LongBeach Healthcare System,

Long Beach, South Carolina, U.S.A.

FREQUENCY OF ANEMIA OF CHRONICDISEASE IN HEMATOLOGICAND ONCOLOGIC DISORDERS

With infectious and inflammatory syndromes, malignant dis-eases comprise one of the major categories of clinical disorderstraditionally associated with the anemia of chronic disease. Inthe series of anemic patients reported by Cash and Sears from

PART VII: SPECIFIC CONDITIONS OF ACD

593

an urban general hospital, 19% of anemia of chronic disease(ACD) patients had a nonhematologic malignancy (1). (Patientswith hematologic diagnoses were specifically excluded from thatseries.)

It is difficult to define the frequency of ACD in specificmalignant syndromes. As will be discussed in more detailbelow, the frequency of ACD in various solid tumor syn-dromes reflects underlying mechanisms contributing toanemia specific to those diseases; in addition, ACD is typicallymore common in advanced disease than it is in localized can-cer. For example, the vast majority of anemias in colon cancer(particularly at early stages) will reflect blood loss with orwithout consequent iron deficiency. In solid tumors withoutassociated blood loss, ACD is likely to be the dominant anemiasyndrome observed early in the course of disease. In the seriesby Cash and Sears referred to above, 23% of anemic patientshad a solid tumor. Of these patients, 44% met the study’scriteria for ACD (hypoferremia with a midrange or higherferritin) (1). In a series of solid tumor patients referred forradiation therapy (2), approximately half were anemic, andapproximately three-quarters of the patients whose anemiawas investigated further were felt to have ACD.

In the hematologic malignancies, the situation is compli-cated by the potential effect of marrow infiltration and replace-ment to a greater degree than is typically observed in solidtumors. Anemia in the myeloproliferative disorders, myelo-dysplastic syndromes, and acute leukemias is a consequenceof clonal abnormalities in the hematopoetic stem cell, and can-not reasonably be attributed to ACD. The same is true for par-oxysmal nocturnal hemoglobinuria, and, to a great degree, foraplastic anemia, particularly later in its clinical course.Approximately 40% of patients with Hodgkin’s disease areanemic at presentation, and the majority of this is attributedto ACD (3). For the non-Hodgkin’s lymphomas, anemia is initi-ally present in 32% of patients (4), and is similarly attributedto ACD in most cases. In multiple myeloma, anemia is presentat diagnosis in 73% of cases (5). It is difficult to distinguishwhat proportion of these cases will have anemia due to mar-row replacement, and what proportion will have anemia due

594 Means and McLAREN

to suppressive effects of cytokines (6). The same difficultyexists for hairy cell leukemia (7,8). In contrast, anemia inchronic lymphocytic leukemia usually results either fromautoimmune hemolysis, pure red cell aplasia, or marrowreplacement.

PATHOGENESIS OF ACD IN HEMATOLOGICAND ONCOLOGIC DISORDERS

The pathogenetic mechanisms involved in ACD have beendiscussed elsewhere in this book. Rather than repeating pre-viously presented material, this section will address the spe-cific role of various pathogenetic processes in hematologicand oncologic disorders.

Inhibitory Cytokines and OncologicDisorders

As has been described earlier, the cytokines whichmediate theimmune and inflammatory response can be implicated in allthe pathogenetic processes contributing to the developmentof ACD (9–11). A number of cytokines, particularly tumornecrosis factor (TNF), have been shown to exhibit increasedproduction in various malignancies (12,13).

In most cases, the cytokines are produced as part of thehost response to the neoplastic cells. In some cases, however,the cytokines are actually produced by the neoplastic cellsthemselves. This is particularly true for TNF and lympho-toxin, which are produced by neoplastic B-lymphocytes in avariety of diseases including multiple myeloma, lymphoma,hairy cell leukemia, and (to a lesser degree) chronic lympho-cytic leukemia (6–8,14,15). Production of these cytokinesmay contribute to the anemia and hematopoetic suppressionassociated with these diseases (7,8).

Blunted Erythropoietin Production

A blunting of the expected increment in erythropoietin pro-duction in response to anemia is a characteristic component

ACD in Hematology and Oncology 595

of ACD. A blunted erythropoietin response to anemia has beenreported in cancer patients generally (16), and in those withspecific hematologic and nonhematologic malignancies (17–20). However, this observation and its significance are notagreed upon universally. In a study of 111 children withmalig-nancies (evenly divided between solid tumors and leukemia),the erythropoietin concentration was felt to be appropriatefor the degrees of anemia observed (21). The major contributorto the anemia in these patients was felt to be an impaired ery-thropoietic response to erythropoietin, as manifested by adecreased concentration of the serum soluble transferrinreceptor (sTfR) (21). It is not clear if the discrepancy betweenthis study and other studies of erythropoietin production inmalignancy can be explained by the high proportion of leuke-mic patients, or by some difference in the biology of childrencompared to adults. An intermediate position is supported bya similar study carried out in adult lung cancer patients thatsimilarly found that the primary factor in their anemia wasimpaired erythropoiesis, but, in addition, observed a relativelyblunted erythropoietin response (20).

Impairment of the erythropoietin response to anemia canalso be caused by chemotherapeutic agents, particularly(although not exclusively) those which are platinum-based,with or without radiation therapy (22–27), and by the use ofnephrotoxic antibiotics, such as amphotericin B (28).

Non-ACD Factors Contributing to Anemia inCancer Patients

The natural history of malignancies, and the physiologic con-sequences of the diagnostic and therapeutic procedures usedin the treatment of these diseases, means that other processesin addition to those involved in ACD may be contributing tothe anemia observed in a particular patient or group ofpatients. Several of these factors are outlined in Table 1.For this reason, a number of authors prefer to use the term‘‘anemia of cancer’’ to refer to the combination of ACDmechanisms and some or all of the other processes outlinedin Table 1 (29). Most of these authors use ‘‘anemia of cancer’’


to describe the anemia found at diagnosis, or which persistsseveral months after treatment has stopped, and refer toanemia observed during active therapy as ‘‘chemotherapy-induced anemia’’ (30).

DIAGNOSIS OF ACD IN CANCER PATIENTS

In cancer patients as in other individuals, the goal of anemiadiagnosis is to identify an appropriate treatment strategy.Diagnostic criteria for ACD in cancer patients are the sameas for ACD in other syndromes. The majority of anemic cancerpatients will have at least some laboratory findings sugges-tive of ACD. Most commonly, this will include a low serumiron and typically a low serum transferrin or total iron-bind-ing capacity (TIBC) concentration. The objective in cancerpatients should be to rule out possibilities other than ACDand treatment-induced anemia, since the treatment of thesetwo modalities is essentially the same. This means eliminat-ing the possibilities of blood loss, iron deficiency, hemolysis,and malnutrition.

Blood loss without iron deficiency can generally beeliminated as a possibility by a careful history and physical

Table 1 Factors Other than ACD Contributing to Anemia inCancer Patients

Blood lossIatrogenicTumor-related

Marrow replacementMalnutritionHemolysis

AutoimmuneMechanicalDrug-induced

TreatmentChemotherapyRadiation therapyAntibiotics


examination, including stool examination for occult blood.Methods for distinguishing ACD from iron deficiency havebeen discussed earlier in the book. A serum ferritin concen-tration below the lower limit of normal for men (usually25–35mg=L) is diagnostic of iron deficiency. In general, aserum ferritin concentration >200mg=L rules out iron defi-ciency (31). In patients with significant underlying illnessessuch as cancer, which are liable to falsely elevate the serumferritin concentration out of proportion to iron stores, thesTfR concentration may be used to identify iron deficiency.In complicated anemic patients undergoing bone marrowexamination, the combination of a normal serum ferritin con-centration and an elevated sTfR concentration detected irondeficiency with a high degree of sensitivity and specificity(32). Measurement of the iron regulatory protein hepcidin,which some investigators have reported is able to distinguishACD from iron deficiency (33,34), is not yet sufficientlystudied or available to be of clinical utility.

Hemolysis may occur by autoimmune mechanisms inpatients with lymphoproliferative malignancies, or (muchmore rarely) by mechanical methods in some individuals withsolid tumors. Suspicion for hemolysis can be provided byotherwise unexplained elevations in the serum bilirubin.The diagnostic evaluation involves the usual direct antiglobu-lin test, and the examination of a peripheral smear.

Malnutrition is suspected on the basis of history andphysical examination, and by valuation of the albumin andprealbumin concentrations. Serum transferrin concentrationmay be decreased by malnutrition, leading to potentialconfusion with ACD.

TREATMENT

General Principles

As with all etiologies of ACD, the first line of therapy shouldbe to treat the underlying disease. Remission of the malig-nancy is typically associated with resolution of the anemia.The situation in cancer treatment is somewhat more


complicated by the fact that treatment typically inducesanemia in and of itself.

Erythropoietin Therapy

As with other etiologies of ACD, erythropoietin is an effectivetherapy. It is also efficacious for ‘‘chemotherapy-inducedanemia,’’ as defined above.

Epoetin

Recombinant human erythropoietin (epoetin alpha) has beenshown to be an effective therapy for anemia in cancerpatients. This efficacy has been demonstrated in both solidand nonmyeloid hematologic malignancies, whether receivingchemotherapy or not (35–39). Epoetin therapy has also beenreported to improve the quality of life of cancer patients sepa-rately from the changes in hemoglobin concentration (39).

Epoetin can be administered on a weight basis (100–300U=kg subcutaneously three times a week), but is more typi-cally given as a fixed dose of 40,000 U subcutaneously oncea week. Failure to show a hemoglobin increment by 4–6 weeksshould prompt a dose increase to 60,000 U per week.

Darbepoetin

Darbepoetin alpha (also called novel erythropoiesis stimulat-ing protein or NESP) is a recombinant erythropoietin analogwith modified glycosylation. The glycosylation pattern of dar-bepoetin results in a longer half-life, and permits longer dos-ing intervals. Both darbepoetin and epoetin exert theirerythropoietic effects through the same receptor (40). Darbe-poetin is effective in the same diseases and clinical settingsas epoetin (41,42).

Darbepoetin has been studied at both weekly andbiweekly dosing intervals (41–43). Like epoetin, it can bedosed on a weight basis, but is more typically given at a fixeddose of 200mg subcutaneously every other week. Failure torespond should result in a dose increment to 300mg. A head-to-head comparison of darbepoetin 200 mg every other weekto epoetin 40,000 U subcutaneously each week showed


comparable efficacy in preliminary analysis. (44). A retrospec-tive cohort study comparing patients treated with epoetin anddarbepoetin at the same doses showed comparable hemoglobinresponse, transfusion frequency (9.4% epoetin, 8.0%darbepoetin), and need for dose increment (14.0% epoetin,11.4% darbepoetin) (45).

Iron Supplementation with Erythropoietic Therapy

Many investigators recommend routine oral iron supplemen-tation during treatment of ACD with erythropoietin products(46). This is based on studies in ACD patients with rheumatoidarthritis in whom iron supplementation was shown to be amajor predictor of response to erythropoietin (47). Unfortu-nately, iron supplementation has generally not been employedin a systematic manner in erythropoietin trials involving can-cer patients, making an evidence-based recommendation diffi-cult. One small study reported that intravenous ironsupplementation in chemotherapy patients treated with ery-thropoietin led to a greater and more consistent hemoglobinresponse than did oral supplementation (48).

Failure to respond to either epoetin or darbepoetin ther-apy after the dose increases described above should prompt asearch for iron deficiency or blood loss, an untreated inflamma-tory=infectious process, or another contributing cause ofanemia.

ACKNOWLEDGMENTS

The work was supported in part by the US Department ofVeterans Affairs Veterans Health Administration Researchfunds (RTM and GDM) and grant HL69418 from the U.S.National Heart, Lung, and Blood Institute (RTM).

REFERENCES

1. Cash JM, Sears DA. The spectrum of diseases associated withthe anemia of chronic disease: a study of 90 cases. Am J Med1990; 87:638–644.


2. Reed WR, Hussey DH, DeGowin RL. Implications of the ane-mia of chronic disorders in patients anticipating radiotherapy.Am J Med Sci 1994; 308:9–15.

3. Kaplan HS. Hodgkin’s Disease. 2nd ed. Boston: HarvardUniversity Press, 1998:689.

4. Moullet I, Salles G, Ketterer N, Dumontet C, Bouafia F,Neidhart-Berard EM, Thieblemont C, Felman P, Coiffier B.Frequency and significance of anemia in non-Hodgkin’slymphoma patients. Ann Oncol 1998; 9:1109–1115.

5. Kyle RA, Gertz MA, Witzig TE, Lust JA, Lacy MQ, DispenzieriA, Fonseca R, Rajkumar AV, Offord JR, Larson DR, PlevakME, Therneau TM, Griepp PR. Review of 1027 patients withnewly diagnosed multiple myeloma. Mayo Clin Proc 2003;78:21–33.

6. Garrett IR, Durie BGM, Nedwin GE, Gillespie A, Bringman T,Sabatini M, Bertolini DR, Mundy GR. Production of lympho-toxin, a bone-resorbing cytokine, by human myeloma cells. NEngl J Med 1987; 317:526–532.

7. Taniguchi T, Kuratsune H, Kanamaru A, Tokumine Y,Tagawa S, Machii T, Kitani T. Inhibition against CFU-C andCFU-E colony formation by soluble factors derived from hairycells. Blood 1989; 73:907–913.

8. Lindemann A, Ludwig W-D, Oster W, Mertlesmann R,Herrmann F. High level secretion of tumor necrosis factor-alpha contributes to hematopoietic failure in hairy cell leuke-mia. Blood 1989; 73:880–884.


10. Means RT. Pathogenesis of the anemia of chronic disease: acytokine-mediated anemia. Stem Cells 1995; 13:32–37.

11. Means RT. Advances in the anemia of chronic disease. Int JHematol 1999; 70:7–12.

12. Balkwill F, Burke F, Talbot D, Tavernier J, Osborne R, NaylorS, Durbin H, Fiers W. Evidence for tumor necrosisfactor=cachectin production in cancer. Lancet 1987; 2:1229–1232.


13. Teppo A-M, Maury CPJ. Radioimmunoassay of tumor necrosisfactor in serum. Clin Chem 1987; 33:2024–2027.

14. Hadjur S, Jirik FR. Increased sensitivity of Fancc-deficienthematopoietic cells to nitric oxide and evidence that thisspecies mediates growth inhibition by cytokines. Blood 2003;101:3877–3884.

15. Sappino A-P, Seelentag W, Pelte M-F, Alberto P, Vassalli P.Tumor necrosis factor=cachectin and lymphotoxin geneexpression in lymph nodes from lymphoma patients. Blood1990; 75:958–962.


17. Kettelhack C, Schoeter D, Matthias D, Schlag PM. Serumerythropoietin levels in patients with solid tumours. Eur JCancer [A] 1994; 30A:1289–1291.

18. Beguin Y, Lampertz S, Bron D, Fillet G. Serum erythropoietin inchronic lymphocytic leukaemia.Br JHaematol 1996; 93:154–156.

19. Leng HMJ, Albrecht CF, Kidson SH, Folb PI. Erythropoietinproduction in anemia associated with experimental cancer.Exp Hematol 1999; 27:806–810.

20. Dowlati A, R’zik S, Fillet G, Beguin Y. Anemia of lung canceris due to impaired erythroid marrow response to erythropoie-tin stimulation as well as relative inadequacy of erythropoietinproduction. Br J Haematol 1997; 97:297–299.

21. Carazza F, Beguin Y, Bergmann P, Andre M, Ferstner A,Devalck C, Fonu P, Buyse M, Sariban E. Anemia in childrenis associated with decreased erythropoietic activity and notwith inadequate erythropoietin production. Blood 1998;92:1793–1798.

22. Smith DH, Goldwasser E, Vokes EE. Serum immunoerythro-poietin levels in patients with cancer receiving cisplatin-basedchemotherapy. Cancer 1991; 68:1101–1105.

23. Schapira L, Antin JH, Ransil BJ, Antman KH, Eder JP,McGarigle CJ, Goldberg MA. Serum erythropoietin levels inpatients receiving intensive chemotherapy and radiotherapy.Blood 1990; 76:2354–2359.


24. Saijo Y, Nakai Y, Saito J, Sugawara S, Suzuki S, Numata Y,Motomiya M. Changes in serum erythropoietin levels duringchemotherapy for lung cancer. Chemotherapy 1992; 38:281–285.

25. Cerruti A, Castello G, Balleari E, Bogliolo G, Lerza R, Pannac-ciulli I. Serum erythropoietin increase in patients receivingadjuvant therapy with 5-fluorouracil and leucovorin. ExpHematol 1994; 22:1261–1263.

26. Sawabe Y, Kikuno K, Iseki T, Iida S, Tabata Y, Yonemitsu H.Changes in serum erythropoietin and the reticulocyte countduring chemotherapy for leukemias. Eur J Haematol 1996;57:384–388.

27. Lee SJ, Kwon JH, Jung CW. Erythropoietin response is inade-quate in cancer patients receiving chemotherapy. IntJ ematol 2001; 74:416–420.

28. Lin AC, Goldwasser E, Bernard EM, Chapman SW. Amphoter-icin B blunts erythropoietin response to anemia. J Infect Dis1990; 161:348–351.

29. Smith RE, Tchekmedyian S, Richards D, Klarnet J, Fleisch-man A, Gayko U, Hetherington A, Glaspy JA. Darbepoietinalfa effectively ameliorates anemia in patients with chronicanemia of cancer: efficacy and pharmacokinetic results of adose escalation study [Abstr]. J Clin Oncol 2002; 21:367a.

30. Wurnig C, Windhager R, Schwameis E, Kotz R, Zoubek A,Stockenhuber F, Kurz RW. Prevention of chemotherapy-induced anemia by the use of erythropoietin in patients withprimary malignant bone tumors (a double-blind, randomized,phase III study). Transfusion 1996; 36:155–159.

31. North M, Dallalio G, Donath AS, Melink R, Means RT. Serumtransferrin receptor levels in patients undergoing evaluationof iron stores: correlation with other parameters, and observedversus predicted results. Clin Lab Haemat 1997; 19:93–97.

32. Means RT, Allen J, Sears DA, Schuster SJ. Serum solubletransferrin receptor and the prediction of marrow aspirateresults in a heterogeneous group of patients. Clin Lab Haemat1999; 21:161–167.

33. Nemeth E, Valore EV, Territo M, Schiller G, Lichtenstein A,Ganz T. Hepcidin, a putative mediator of anemia of inflam-


mation, is a type II acute phase protein. Blood 2003; 101:2461–2463.

34. Dallalio G, Fleury T, Means RT. Serum hepcidin in clinicalspecimens. Br J Haematol 2003; 122:996–1000.

35. Oster W, Herrmann F, Gamm H, Zeile G, Lindemann A,Muller G, Brune T, Kraemer H-P, Mertelsmann R. Erythro-poietin for the treatment of anemia of malignancy associatedwith bone marrow infiltration. J Clin Oncol 1990; 8:956–962.

36. Ludwig H, Sundal E, Pecherstorfer M, Leitgeb C, BauernhoferT, Beinhauer A, Samonigg H, Kappeler AW, Fritz E. Recombi-nant human erythropoietin for the correction of cancer asso-ciated anemia with and without concomitant cytotoxicchemotherapy. Cancer 1995; 76:2319–2329.

37. Cazzola M, Messinger D, Battistel V, Bron D, Cimino R,Enller-Ziegler L, Essers U, Greil R, Grossi A, Jaeger G, LeMe-vel A, Najman A, Silingardi V, Spriano M, Van HA, Ehmer B.Recombinant human erythropoietin in the anemia associatedwith multiple myeloma or non-Hodgkin’s lymphoma: dose find-ing and identification of predictors of response. Blood 1995;86:4446–4453.

38. Anand A, Anand A, Anand N. Recombinant human erythro-poietin for the correction of cancer associated anemia withand without concomitant cytotoxic chemotherapy. Cancer1996; 77:1960.

39. Quirt I, Robeson C, Lau CY, Kovacs M, Burdette-Radoux S,Dolan S, Tang SC, McKenzie M, Couture F. Canadian EprexOncology Study Group. Epoetin alfa therapy increases hemo-globin levels and improves quality of life in patients withcancer-related anemia who are not receiving chemotherapyand patients with anemia who are receiving chemotherapy.J Clin Oncol 2001; 19:4126–4134.

40. Egrie JC, Browne JK. Development and characterization ofnovel erythropoiesis stimulating protein (NESP). Nephrol DialTranspl 2001; 16(suppl 3):3–13.

41. Smith RE, Jaiyesima IA, Meza LA, Tchekmediyan NS, Chan D,Griffith H, Brosman S, Bukowski R, Murdoch M, Rarick M,Saven A, Colowick AB, Fleischman A, Gayko U, Glaspy J.Novel erythropoieisis stimulating protein (NESP) for the


treatment of the anemia of chronic disease associated withcancer. Br J Cancer 2001; 84(suppl 1):24–30.

42. Glaspy J, Jadeja JS, Justice G, Kessler J, Richards D,Schwartzberg L, Rigas J, Kuter D, Harmon D, Prow D,Demetri G, Gordon D, Arseneau J, Saven A, Hynes H, BocciaaR, O’Byrne J, Colowick AB. A dose-finding and safety studyof novel erythropoiesis stimulating protein (NESP) for thetreatment of anaemia in patients receiving multicylechemotherapy. Br J Cancer 2001; 84(suppl 1):17–23.

43. Charu V, Belani CP, Gill AN, Bhatt M, Ben-Jacob A, Tomita D,Katz D, Colowick A. Every 2-week (Q2W) dosing of darbepoe-tin alpha in patients with anamia of cancer (AOC): interimanalysis of a randomized, controlled study [Abstr]. Blood2003; 102(suppl):499a.

44. Schwartzberg LS, Yee LK, Senecal FM, Charu V, Yao B,Mendes E, Rossi G. Early results of a head-to-head comparisonof darbepoietin alfa 200mcg given every 2weeks (Q2W) andepoetin alfa 40,000 U given weekly (QW) [abstract]. Blood2003; 102(suppl):515a.

45. Schwartzberg L, Shiffman R, Tomita D, Stolshek B, Rossi G,Adamson R. A multicenter retrospective cohort study of prac-tice patterns and clinical outcomes of the use of darbepoetinalfa and epoetin alfa for chemotherapy-induced anemia. ClinTher 2003; 25:2781–2796.

46. Means RT. Recent developments in the anemia of chronicdisease. Curr Hematol Rep 2003; 2:116–121.

47. Nordstrom D, Lindroth Y, Marsal L, Hafstrom I, Heinrich C,Rantapaa-Dalqvist S, Engstrom-Laurent A, Friman C. Avail-ability of iron and degree of inflammation modifies theresponse to recombinant human erythropoietin when treatinganemia of chronic disease in patients with rheumatoid arthri-tis. Rheumatol Int 1997; 17:67–73.

48. Ballard H, Rana J, Ackerman A, Merino R, Rosenoff S,Kastritsis C, Trout R, Chaudhaury M, Auerbach M. Total doseinfusion (TDI) of iron dextran (ID) optimizes erythropoietin(EPO) responsiveness in the anemia of cancer (CA). ProcASCO 1999; 18:A2245.


21

Anemia in Cancer PatientsUndergoing Surgery

D. OFNER

Department of Surgery, Division of General andTransplant Surgery, Innsbruck University

Medical School, Innsbruck, Austria

INTRODUCTION

Anemia occurring during cancer, especially in colorectaladenocarcinoma, may be the result of several causes in particu-lar: chronic disease, chemo-=radiotherapy, and blood loss. Inlast years, anemia has been recognized as commonly asso-ciated with cancer. Moreover randomized, controlled clinicaltrials and large cohort studies have demonstrated thatincreased hemoglobin levels are linked with increased qualityof life (1–4). Quite extensive literature, more or less individualstudies, has been published documenting that anemia is aprognostic factor in lung, cervix, head and neck, and prostate

607

(5–10). According to a meta-analysis (9), the overall adjustedrelative risk for death in patients with anemia increased 1.6-fold(95% CI: 1.54–1.77) when compared with patients presentingregular hemoglobin levels at time of diagnosis. In sum, one-thirdof patients were diagnosed anemic. With regard to therapy,some authors have suggested that maintenance of hemoglobinlevels during chemo- and=or radiotherapy may improveeffectiveness of the procedure accordingly (5–7,11–15).

With regard to anemia in patients suffering from colorectaladenocarcinoma, two studies were reported to be of prognosticsignificance (16,17). Nevertheless, the number of patients inthese studies is small and adjusted hazard rate rations are notnoted. The aim of the present study was to evaluate the prognos-tic value of hemoglobin levels in a multivariate approach withina large sample size of patients with colorectal cancer over a timeperiod of more than 8 years.

PATIENTS AND METHODS

Between 1992 and 1999, a total of 848 patients suffering fromcolorectal cancer have been treated at the Department ofSurgery, University Hospital Innsbruck. In this time period,806 patients have been operated due to primary colorectalmalignancy. Out of them, 19 patients had a severe per analbleeding. These patients were excluded from the study. Inthe remaining group in 787, the preoperative blood hemoglo-bin level was available. Six hundred and ninety-threepatients were operated with curative, the remaining 94patients with palliative intent.

RESULTS AND DISCUSSION

Blood hemoglobin concentrations ranged from 5.7 to 18.1 g=dL(mean 12.4 g=dL). According to publications in other tumorpatients in our study, one-third of all colorectal cancer patientsinvestigated were anemic (hemoglobin levels, female:�12 g=dL; male �13 g=dL). Two hundred and twelve out of787 (27%) patients showed severe anemia (hemoglobin levels

608 Ofner

�11 g=dL) prior to operation. Hemoglobin values were statisti-cally significantly related to gender (P< 0.0001), tumor site(P¼0.0001), pT-stage (P< 0.01), and tumor stage according toUICC (P< 0.001; see Table 1), whereas pN-stage, M-stage,tumor grade, R-classification, surgeon volume, adjuvantchemo- and=or radiotherapy, and hospital stay were not asso-ciated with hemoglobin levels of respective patients. Anemiawas more frequently diagnosed in female (33%) than in male(22%) patients (female: mean 12.0, SD 2.04 vs. male: mean13.2, SD 6.62; P< 0.01). Fifty-eight out of 140 (41%) patientswith tumors of the right hemi colon were anemic, whereasthe remaining patients showed anemia in only 21% (74 out of345), which is enough well known by clinicians. This differencewas statistically highly significant (right hemi colon: mean11.5, SD 2.27; left hemi colon: mean 12.6, SD 2.08; rectum:mean 13.0, SD 2.13; right vs. left hemi colon: P¼ 0.0001, right

Table 1 Relationship Between Hemoglobin Values and VariousClinico-Pathological Parameters in Colorectal Cancer Patients

Mean SD P

GenderFemale 12.0 2.04 < 0.01Male 13.2 6.62

Age�69 years 13.1 6.67 < 0.01> 69years 12.1 2.14

Tumor siteRight hemicolon 11.5 2.27 > 0.000Left hemicolon 12.6 2.08Rectum 13.0 2.13

pT stagepT1 13.1 2.01 > 0.000pT2 12.4 1.92pT3 12.0 2.06pT4 11.6 2.45

UICC stageUICC I 13.2 1.98 < 0.01UICC II 12.2 2.23UICC III 12.4 2.29UICC IV 12.2 2.21

Anemia in Cancer Patients Undergoing Surgery 609

hemi colon vs. rectum: P¼ 0.0001; left hemi colon vs. rectum:N.S.). Furthermore, the proportion of anemic patientsincreased with pT stage. Ten percent of pT1, 21% of pT2, 29%of pT3, and 38% of pT4 tumors showed anemia (see Fig. 1).Thirty percent of patients were found anemic in tumor stagesII, III, and IV according to UICC. In contrary, only 14% wereanemic in tumor stage I. Moreover, anemia was associated withshorter survival times in colorectal cancer patients (P< 0.001)and adjusted Hazard rate ratios showed that anemia increasedmortality by 27% (95% CI: 11–64%). Looking at patientswho had resection of their tumors with curative intent, the

Figure 1 Boxplots of hemoglobin values with regard to pT-stages.

610 Ofner

prognostic values of hemoglobin levels were even more pro-nounced increasing mortality by 53% (95% CI 15–103%).Patients with regular hemoglobin levels had 5-year survivalprobabilities of 64%, patients with severe anemia 50% (seeFig. 2).

The results of the present study strongly support the prog-nostic value of anemia in patients suffering from colorectaladenocarcinoma. The findings of this study with a sample sizeof more than 600 patients, which is the first in colorectal, are inaccordance with other three large-sized studies in humanmalignancies. Nevertheless, further studies are required toevaluate the impact of anemia treatment on survival.

Figure 2 Kaplan–Meier survival curves of patients with regard tohemoglobin values at time of operation, who underwent resection oftheir tumors with curative intent. Continuous line: regular hemo-globin values; dashed line: severe anemia (�11 g=dL). Pointed line:female patients >11 and �12 g=dL and male patients with >11 and�13 g=dL.


REFERENCES

1. Abels R. Erythropoietin for anaemia in cancer patients. Eur JCancer 1993; 29A(suppl 2):S2–S8.

2. Glaspy J, Bukowski R, Steinberg D, Taylor C, Tchekmedyian S,Vadhan-Raj S. Impact of therapy with epoetin alfa on clinicaloutcomes in patients with nonmyeloid malignancies duringcancer chemotherapy in community oncology practice. ProcritStudy Group. J Clin Oncol 1997; 15:1218–1234.

3. Demetri GD, Kris M, Wade J, Degos L, Cella D. Quality-of-lifebenefit in chemotherapy patients treated with epoetin alfa isindependent of disease response or tumour type: results froma prospective community oncology study. Procrit Study Group.J Clin Oncol 1998; 16:3412–3425.

4. Gabrilove JL, Cleeland CS, Livingston RB, Sarokhan B, WinerE, Einhorn LH. Clinical evaluation of once-weekly dosing ofepoetin alfa in chemotherapy patients: improvements in hemo-globin and quality of life are similar to three-times-weeklydosing. J Clin Oncol 2001; 19:2875–2882.

5. Wagner W, Hermann R, Hartlapp J, Esser E, Christoph B,Muller MK, Krech R, Koch O. Prognostic value of hemoglobinconcentrations in patients with advanced head and neck can-cer treated with combined radio-chemotherapy and surgery.Strahlenther Onkol 2000; 176:73–80.

6. van Acht MJ, Hermans J, Boks DE, Leer JW. The prognosticvalue of hemoglobin and a decrease in hemoglobin duringradiotherapy in laryngeal carcinoma. Radiother Oncol 1992;23:229–235.

7. Tarnawski R, Skladowski K, Maciejewski B. Prognostic valueof hemoglobin concentration in radiotherapy for cancer ofsupraglottic larynx. Int J Radiat Oncol Biol Phys 1997; 38:1007–1011.

8. Girinski T, Pejovic-Lenfant MH, Bourhis J, Campana F, CossetJM, Petit C, Malaise EP, Haie C, Gerbaulet A, Chassagne D.Prognostic value of hemoglobin concentrations and blood trans-fusions in advanced carcinoma of the cervix treated by radia-tion therapy: results of a retrospective study of 386 patients.Int J Radiat Oncol Biol Phys 1989; 16:37–42.

612 Ofner

9. Caro JJ, Salas M, Ward A, Goss G. Anemia as an independentprognostic factor for survival in patients with cancer: asystemic, quantitative review. Cancer 2001; 91:2214–2221.

10. Glaser CM, Millesi W, Kornek GV, Lang S, Schull B, Watzinger F,Selzer E, Lavey RS. Impact of hemoglobin level and use of recom-binant erythropoietin on efficacy of preoperative chemoradiationtherapy for squamous cell carcinoma of the oral cavity andoropharynx. Int J Radiat Oncol Biol Phys 2001; 50:705–715.

11. Dietz A, Rudat V, Conradt C, Vanselow B, Wollensack P, Staar S,Eckel H, Volling P, Schroder M, Wannenmacher M, Muller RP,Weidauer H.. [Prognostic value of hemoglobin level for primaryradiochemotherapy of head-neck carcinomas]. Hno 2000; 48:655–664.

12. Grogan M, Thomas GM, Melamed I, Wong FL, Pearcey RG,Joseph PK, Portelance L, Crook J, Jones KD. The importanceof hemoglobin levels during radiotherapy for carcinoma of thecervix. Cancer 1999; 86:1528–1536.

13. Harrison LB, Shasha D, Homel P. Prevalence of anemia incancer patients undergoing radiotherapy: prognostic signifi-cance and treatment. Oncology 2002; 63(suppl 2):11–18.

14. Henke M. Correction of cancer anemia—impact on diseasecourse, prognosis and treatment efficacy, particularly forpatients undergoing radiotherapy. Onkologie 2001; 24:450–454.

15. Overgaard J, Hansen HS, Jorgensen K, Hjelm Hansen M.Primary radiotherapy of larynx and pharynx carcinoma—ananalysis of some factors influencing local control and survival.Int J Radiat Oncol Biol Phys 1986; 12:515–521.

16. Edler L, Heim ME, Quintero C, Brummer T, Queisser W. Prog-nostic factors of advanced colorectal cancer patients. EurJ Cancer Clin Oncol 1986; 22:1231–1237.

17. Graf W, Glimelius B, Pahlman L, Bergstrom R. Determinantsof prognosis in advanced colorectal cancer. Eur J Cancer 1991;27:1119–1123.


22

Iron Status, Anemia of ChronicDisease, and Infection

JOHAN R. BOELAERT

Unit of Renal and Infectious Diseases,Algemeen Ziekenhuis St-Jan,

Brugge, Belgium

There is a bidirectional interplay between iron status andinfection, such that the iron status may modify the risk andoutcome of some infections. Reciprocally, infection influencesiron status and may result in the anemia of chronic disease(1–6), also known as the ‘‘anemia of inflammation.’’ This chap-ter on the bidirectional interplay between iron and infectionwill focus on two infections, tuberculosis and HIV, whichare recognized internationally as leading health problems.When a few years ago, the United Nations for the first timedevoted a full session to world health issues and decided tocombat major infectious diseases worldwide, they prioritizedHIV infection, tuberculosis, andmalaria. This decision resulted

615

in the creation of the United Nations Global Fund to combatthese infections.

TUBERCULOSIS

It is estimated that Koch’s bacillus infects more than one-thirdof the world’s population. Even if ‘‘only’’ some 10% of thoseinfected develop clinical disease, tuberculosis is still a majorhealth problem. Presently, tuberculosis accounts for (2–3)million deaths annually. The health burden of tuberculosis isaggravated by the expanding epidemic of HIV infection, andthe persistence of war and famine in many parts of the world,all of which ‘‘fuel’’ tuberculosis (7). Furthermore, multidrugresistance on the part of the bacillus complicates therapy.

Mycobacterium tuberculosis is a facultative intracellularmicro-organism that resides within the phagosomes of macro-phages. That iron is an obligate cofactor for at least 40enzymes encoded by the M. tuberculosis genome underlinesthe absolute microbial requirement for iron. The particularintramacrophagic niche that M. tuberculosis inhabits necessi-tates unique bacillary iron acquisition mechanisms. A num-ber of observations indicate that M. tuberculosis takes upiron from the host cell’s transferrin-receptor-mediated path-way to acquire iron from transferrin. (i) IntraphagosomalM. tuberculosis takes up iron from exogenous holotransferrin(8) and concentrates the metal within the phagosome (9). (ii)The micro-organism’s dual ‘‘mycobactin T’’ siderophore sys-tem (10) enables it to acquire iron from holotransferrin (11).(iii) The M. tuberculosis-containing phagosome continuouslyinteracts with the early endosomes whereto holotransferrinspecifically traffics (12). (iv) The micro-organism has a strategyto arrest phagosome maturation precisely at the iron-rich earlyendosome stage, and it is likely that this strategy evolvedat least in part to facilitate iron acquisition. (v) Finally, themacrophage that is infected with M. avium counteracts thebacterial acquisition of iron by decreasing transferrin-receptormRNA expression, but it is unclear whether this mechanismapplies to M. tuberculosis (13).

616 Boelaert

Since iron acquisition is very important forM. tuberculosis,modifications in iron availability have been reported to influ-ence the growth of M. Tuberculosis, either in vitro or in vivo.As shown in Table 1 (14–27), adding iron experimentallyhas, with one exception (18), always resulted in enhancedmicrobial growth in vitro or in vivo (9,14–16,20–22). On thecontrary, the following treatments have resulted in decreased

Table 1 Altering the Iron Status Influences the Growth ofM. tuberculosis Both In Vitro and In Vivo

M. tuberculosis growth

Enhanced(Refs.)

Inhibited(Refs.)

I. In vitro growthIA. Growth in broth or serum

Iron salts 9, 14, 15Serum transferrin 16Desferrioxamine 9

IB. Growth in macrophagesIB1. Murine macrophages

Lactoferrin 9Anti-Tf-R antibody 9Siderophore-deficient M.TB mutant 17

IB2. Human macrophagesIron salts 18Gallium-transferrin 19

II. In vivo growthIIA. Murine infection

Iron enrichment of drinking water 9Parenteral iron 16, 20–22Desferrioxamine 20

b2-microglobulin knockout mice 9b32-microglobulin knockout mice treated with

(compared to untreated knockout mice)Intranasal lactoferrin 9Desferrioxamine 9

IIB. Clinical infection in humansHIV infection associated with increased bonemarrow iron

25

African iron overload 26, 27

Iron Status, Anemia of Chronic Disease, and Infection 617

M. tuberculosisgrowth: thepharmacologic iron chelator desfer-rioxamine (9,20), the endogenous iron chelators apo-transferrin(16) or apolactoferrin (9), a transferrin-receptor antibody thatappears to decrease cellular iron influx via the transferrin=transferrin-receptor pathway (9), and gallium salts or gallium-transferrin that enter the mammalian cell (19).

Some genetic manipulations are also in agreement withthe concept that iron favors the growth of M. tuberculosis.First, a bacterial mutant, defective in the synthesis of myco-bactin siderophores, has impaired growth in macrophages(17). Second, b2-microglobulin knockout mice that may serveas a model for human hereditary hemochromatosis are moresusceptible to tuberculosis than control mice (9). However,the latter mouse model is of questionable relevance for studieson tuberculosis, as there are not increased but decreasedlevels of iron in reticulo-endothelial organs such as the spleenof b2-microglobulin knockout mice, in agreement with theobservation that macrophages are relatively iron depleted inHFE hemochromatosis (28).

More importantly, what is the clinical evidence that anexcessive iron burden increases the susceptibility to or aggra-vates the outcome of tuberculosis? In the 19th century,Armand Trousseau, the Paris-based professor who made thefirst description of hemochromatosis (‘‘bronze diabetes withcirrhosis’’), also reported that patients recovering from activetuberculosis tended to relapse if they received iron-rich sup-plements and warned against the use of such supplementsin the setting of tuberculosis (23). One hundred years later,the Murray family observed three episodes of recrudescenttuberculosis among 71 predominantly adult rural Somaliansubjects given iron supplementation for 30 days, while no epi-sodes occurred among 66 subjects given placebo (24). Sincethese anecdotal observations, two retrospective studies andone prospective study have been performed with properstatistical analysis.

One retrospective study of bone marrow macrophage irongrades in HIV-seropositive patients in the United Statesfound that a history of tuberculosis was significantly morecommon in patients with high macrophage iron grades than

618 Boelaert

in patients with low or normal iron grades (25). Another retro-spective study analyzed necropsies performed on adult blacksfrom southern Africa in which splenic iron levels were mea-sured semiquantitatively in 604 subjects, and iron grades1–5 were assigned. Splenic iron, contained principally inmacrophages, was the variable most significantly associatedwith death from tuberculosis. The odds of death from tubercu-losis with the highest grade of splenic iron were 16.9 times theodds with the two lowest grades. This finding suggests thatreticulo-endothelial iron overload in black Africans may be arisk factor for death from tuberculosis (26). More recently,the same group studied prospectively the possible relation-ship between pulmonary tuberculosis and increased dietaryiron by comparing 98 patients with proven pulmonary tuber-culosis and the same number of matched controls in ruralZimbabwe (27). The consumption of traditional beer brewedin nongalvanized steel containers was used as the indicatorof exposure to increased dietary iron. Logistic regression mod-eling indicated that, after adjustment for HIV status and liverfunction, increased dietary iron was associated with a 3.5-foldincrease in the estimated odds of developing pulmonary tuber-culosis (P¼ 0.009) and with a nonsignificant trend towardhigher mortality in the patients with pulmonary tuberculosis.In summary, there is growing evidence from the above experi-mental and clinical data indicating that iron acquisition is ofmajor importance for the M. tuberculosis growth and thatclinical situations resulting in iron excess, particularly in thereticulo-endothelial system, increase the risk and may worsenthe outcome of tuberculosis.

The last study discussed above provides detailed data onthe hematologic and iron status in 98 patients with pulmonarytuberculosis and the same number of matched healthy controlsubjects from rural Zimbabwe (27). As shown in Table 2, themean hemoglobin concentration among the tuberculosispatients at the onset of antituberculous treatment was 4.9g=dLlower than in the healthy controls (P< 0.001) and 2g=dLlower in the subset of HIV-positive compared to HIV-negativetuberculosis patients (P< 0.005). At the end of the tuberculosistherapy that lasted for 6 months, hemoglobin concentrations


had normalized in theHIV-negative patients and had increasedmarkedly in the HIV-positive patients. At the same time, meanserum ferritin concentrations decreased from 411 to 87ng=mLin the HIV-negative and from 514 to 47ng=mL in HIV-positivetuberculosis patients, demonstrating that the tuberculosistherapy had decreased the inflammatory response. These dataillustrate that the anemic response resulting from tuberculosisis aggravated by HIV=tuberculosis coinfection.

HIV INFECTION

HIV infection is typically a chronic infection having, inthe absence of therapy, a prolonged course that is mostlyclinically asymptomatic until the CD4þ lymphocyte counthas dropped significantly. Following the onset of a low CD4þ

Table 2 Hemoglobin (g=dL) and Hematocrit (%) Values inPulmonary Tuberculosis and in HIV-Infection (MeanþSD)

Hemoglobin

Pulmonary TBpatients Control patients

All HIV-pos. HIV-neg.

Tuberculosis (Zimbabwe)a

At weeks 1–3 oftherapy

9.4þ 2.1 8.8 � 1.9 10.8 � 2.1

At 7–9 months 14.3 � 1.8 12.7þ 1.9 13.9þ 2.1HIV infection (Belgium)

Hemoglobin HematocritNo AIDS, no antiviraltherapy

14.8 � 1.3 43.4 � 3.8

AIDS 33.1 � 6.5Prior to ‘‘HAART’’ 11.3 � 2.1 40.8þ 3.6At latest follow-up(>6 months oftherapy)

13.9þ 1.3

aModified from Ref. 27 with permission of the authors.

620 Boelaert

count, there is an accelerated AIDS phase characterized byopportunistic infections and eventually death, usually 10–12years after acquisition of the infection. Paradoxically, HIVinfection is also characterized by inappropriate immunologi-cal stimulation (29,30). As expected, HIV infection may resultin anemia of chronic disease (31). Table 2 shows hemoglobinvalues among HIV-infected patients followed at the infectiousdiseases unit of Algemeen Ziekenhuis St-Jan in Brugge,Belgium. Two extremes are shown: on one hand a cohort of 18asymptomatic patients with CD4þ cell counts >350=mm3

and not requiring antiviral treatment, and on the other handa cohort of 28 patients with AIDS, for whom hematologicalvalues are shown both prior to the institution of HAART(‘‘highly active antiretroviral therapy,’’ consisting of at leastthree antiretroviral compounds) and at a later follow-up. Ane-mia is more important in the subgroup of patients withadvanced disease (AIDS) than in those not having reachedthe AIDS phase of the disease, with a 3 g=dL difference inthe mean hemoglobin concentration (P< 0.001). In thepatients with AIDS, at least 6 months of HAART resulted ina good suppression of viral replication (HIV-1 plasma levels<50 copies=mL). Comparing the hemoglobin concentrationsof the AIDS patients at the time of initiation of HAART andat the latest available follow-up after six months of HAARTindicates that a good suppression of viral replication partiallyabrogates the anemia, with the mean hemoglobin concentra-tion increasing by 2.6 g=dL (P< 0.001).

The anemic response of HIV infection, particularly in itsmore advanced stage, is accompanied by iron accumulation inmacrophages, microglia, endothelial cells, and myocytes,resulting in an increased iron burden in the bone marrow,brain, muscle, liver, and spleen (31). This iron retention inthe reticulo-endothelial system is mainly due to the chronicinflammatory process, and transfusion of packed red bloodcells is also contributory (31). This progressive iron accumula-tion may have at least four hazardous consequences (31).First, an excessive amount of iron may increase oxidativestress, activate nuclear factor kB, and directly enhance HIVtranscription (32). Second, iron excess results in decreased


activity of polymorphonuclear granulocytes (33), macrophages(34), and impaired cell-mediated immunity (35,36), with a pos-sible shift from a T helper 1 to a T helper 2 immune response(37). Third, iron excess may increase the risk of developingHIV-related tumors, such as Kaposi’s sarcoma. Indeed, spin-dle cells from Kaposi sarcoma lesions brought into culturegrow better in the presence of iron and are growth inhibitedin the presence of iron chelators, suggesting that iron maybe a cofactor in the pathogenesis of Kaposi’s sarcoma (38).However, this has not been substantiated by in vivo studiesin mice (39). Finally, iron loading stimulates the growth ofmany micro-organisms (31), including AIDS opportunistssuch as M. tubercidsis (as discussed above), M. avium (40),Cryptococcus neoformans (41), Candida albicans (37), andPenicillium marneffei (42). This enhanced growth of micro-organisms may result in the vicious circle that is illustratedin Fig. 1: increased reticulo-endothelial iron deposition inc-reases microbial growth that upregulates HIV transcription,

Figure 1 HIV infection: the vicious circle between HIV infectionand reticulo-endothelial iron deposition.

622 Boelaert

further decreasing the CD4þ lymphocyte cell count, and aug-menting the inflammatory response with associated anemiaand reticulo-endothelial iron deposition.

Although malaria was previously not considered to be anAIDS-opportunistic infection, it has recently become clearthat the incidence and the severity of malaria are increasedin HIV-infected pregnant and nonpregnant adults (43). Thishas been well documented in a large cohort of pregnantwomen in western Kenya, where malaria infection at deliverywas present in 29.9% of HIV-positive vs. 19.5% of HIV-nega-tive women (44). In the same study, HIV=malaria coinfectiondecreased the mean postpartum hemoglobin concentration by1.3 g=dL, whereas either HIV infection or malaria alonedecreased the hemoglobin level by 0.4 g=dL. It is temptingto speculate that the enhanced anemic response in thedual-infected patients may be pathophysiologically related toanemia of chronic disease. However, it is also possible that thedocumented higher parasite loads in HIV-infected patientsmay have caused more hemolysis.

Several clinical studies supporting the hypothesis thathigh iron stores may adversely influence the outcome of HIVinfection were recently reviewed and will not be discussed herein detail (45).

HIV=TUBERCULOSIS COINFECTION

The case of HIV=tuberculosis coinfection merits particularattention, in view of the increasing incidence of this dualinfection, particularly in the Third World, and its majorthreat to world health (43). Even in the absence of HIV ortuberculosis, developing world populations present signs ofimmune activation compared to populations from the North(29,30). This issue is complicated by frequently coexistingmalnutrition, which impairs Th 1 immunity and decreases1,25-dihydroxyvitamin D3 levels, both factors being ofimportance in protection against tuberculosis (7,46). As indi-cated above, HIV infection is characterized by an inappropri-ate immune activation (29,30). Similarly, tuberculosis is


accompanied by generalized immune activation (47). If theyare real ‘‘partners in crime’’ (48), HIV and M. tuberculosisshould stimulate each other, and this indeed occurs (49).On one hand, the dysfunctional immune response in HIVinfection makes HIV-1-infected individuals less efficient inactivating M. tuberculosis-infected macrophages in the lung,resulting in a weaker granulomatous immune response,higher rates of latent tuberculosis reactivation, more frequentprogression to active tuberculosis, and a higher frequencyof extrapulmonary tuberculosis (49). On the other hand,tuberculosis stimulates HIV-1 replication, partially dueto increased levels of TNF-a that activate nuclear factor-KBfor which the virus has two binding sites or even more bindingsites in the case of HIV-1 subtype C that is highly prevalent insouthern Africa. This enhanced HIV replication during activetuberculosis coinfection has not only been demonstrated invitro but also in vivo, with highly increased HIV-1 RNAplasma levels during tuberculosis that revert to the originallevels after cure of the tuberculosis (50). Furthermore, thisdual infection results in increased HIV-1 heterogeneity,another factor in HIV disease progression (49). This reciprocalactivation and the ensuing vicious circle, depicted in Fig. 1,explain why the anemia of chronic disease that exists ineither tuberculosis or HIV infection is dramatically enhancedin HIV=tuberculosis coinfection (Table 2).

HEPCIDIN AS KEY MEDIATOR OF INFECTION-RELATED ANEMIA OF CHRONIC DISEASE

The pathophysiology of the anemia of chronic disease isreviewed in detail in other chapters of this book. Recently, avery interesting work has shed light on a new mechanismwhereby systemic infection may induce a signal that resultsin inhibition of iron uptake by the small intestine and in inhi-bition of iron recycling from senescent red cells in macro-phages, both mechanisms that may restrict the availabilityof iron to invading micro-organisms and that are operational

624 Boelaert

in the anemia of chronic disease. The signal is likely to be hep-cidin, a small peptide (20–25 amino acids long), cleaved froma larger precursor synthesized by the liver. Mice that loseexpression of hepcidin have the same phenotype as Hfeknockout mice, with hepatic iron overload, increased circulat-ing iron, apparently increased intestinal iron absorption, anddecreased iron in tissue macrophages (51,52). Fleming andSly (53) therefore reasoned that the reverse situation (dec-reased intestinal absorption, decreased circulating iron, andincreased reticulo-endothelial iron), as found in the anemia ofchronic disease, might result from increased hepcidin expres-sion due to inflammation=infection.

What may be the relationships between infection andhepcidin? First, hepcidin (from ‘‘hepatic bactericidal protein’’)has bactericidal and antifungal properties, like other structu-rally related antimicrobial peptides such as defensins that areinvolved in innate immunity. Second, hepatic expression ofhepcidin is increased after administration of the inflamma-tion inducer turpentine (54) or of lipopolysaccharide to mice(55) and after bacterial infection in a fish model (56). Third,hepcidin has been isolated from human plasma ultrafiltrateand urine, and patients with anemia of inflammation havevery high levels of urinary hepcidin as compared to normalthat may normalize after cure of an underlying infection(57,58). Fourth, patients have been reported with largebenign hepatic adenomas that inappropriately express hepci-din. These patients have anemia characterized by impairedintestinal iron absorption and impaired macrophage ironrelease (59). We are presently hampered by the unavailabilityof an assay to monitor hepcidin plasma levels in humans,and it is to be expected that such an assay will increase ourknowledge of the potential role of hepcidin in the anemia ofchronic disease. Furthermore, a better understanding ofthe molecular events leading to its stimulus, its secretion,and its targets should unravel whether hepcidin is indeedthe crucial mediator of the anemia of chronic disease. Ifso, pharmacological intervention aimed at impairing theproduction of hepcidin or its effects may become of clinicalrelevance.


REFERENCES

1. Jurado RL. Iron, infections, and anaemia of inflammation. ClinInfect Dis 1997; 25:888–895.

2. Weiss G. Iron and anemia of chronic disease. Kidney Intern1999; 55(suppl 69):S12–S17.

3. Weinberg ED. Iron and infection. Microbiol Rev 1978; 42:45–66.

4. Weinberg ED. Iron loading and disease surveillance. EmergingInfect Dis 1999; 5:346–352.

5. Weinberg ED. Modulation of intramacrophage iron meta-bolism during microbial cell invasion. Microbes Infect 2000; 2:85–89.

6. Brock JH. Benefits and dangers of iron during inflammation.Curr Opin Clin Nutr Metab Care 1999; 2:507–510.

7. Boelaert JR, Gordeuk VR. Protein energy malnutrition andrisk of tuberculosis infection. Lancet 2002; 360:1102.

8. Russell DG, Sturgill-Koszycki S, Vanheyningen T, Collins HL,Schaible UE. Why intracellular parasitism need not to be adegrading experience for Mycobacterium. Philos Trans R SocLondon B Biol Sci 1997; 352:1303–1310.

9. Schaible UE, Collins HL, Priem F, Kaufmann SHE. Correctionof the iron overload defect in b-2-microglobulin knockout miceby Iactoferrin abolishes their increased susceptibility to tuber-culosis. J Exp Med 2002; 196:1507–1513.

10. De Voss JJ, Rutter K, Schroeder BG, Barry CE. Iron acquisi-tion and metabolism by mycobacteria. J Bacteriol 1999; 181:4443–4451.

11. Gobin J, Horwitz MA. Exochelins of Mycobacterium tuberculo-sis remove iron from human iron-binding proteins and donateiron to mycobactins in the M. tuberculosis cell wall. J Exp Med1996; 183:1527–1532.

12. Clemens DL, Horwitz MA. The Mycobacterium tuberculosisphagosome interacts with early endosomes and is accessibleto exogenously administered transferrin. J Exp Med 1996;184:1349–1355.

626 Boelaert

13. Zhong W, Lafuse WP, Zwilling BS. Infection with Mycobacter-ium avium differentially regulates the expression of irontransport protein mRNA in murine peritoneal macrophages.Infect Immun 2001; 69:6618–6624.

14. Youdim S. In vitro effect of iron salts and chelating agents onserum tuberculostasis. Am Rev Resp Dis 1969; 99:925–931.

15. Raghu B, Raghupati S, Venkatesan P. Effect of iron on thegrowth and siderophore production of mycobacteria. BiochemMol Biol Int 1993; 31:341–348.

16. Kochan I. The role of iron in bacterial infections, with specialconsideration of hot-tubercle bacillus interaction. Curr TopicsMicrobiol Immunol 1973; 60:1–30.

17. De Voss JJ, Rutter K, Schroeder BJ, Su H, Zhu Y, Barry CE.The salicylate-derived mycobactin siderophores of Mycobacter-ium tuberculosis are essential for growth in macrophages. ProcNatl Acad Sci USA 2000; 97:1252–1257.

18. Byrd TF. Tumor necrosis factor a (TNFa) promotes growth ofvirulent Mycobacterium tuberculosis in human monocytes:iron-mediated growth suppression is correlated with decreasedrelease of TNFa from iron-treated monocytes. J Clin Invest1997; 99:2518–2529.

19. Olakanmi O, Britigan BE, Schlesinger LS. Gallium disruptsiron metabolism of mycobacteria residing within humanmacrophages. Infect Immun 2000; 68:5619–5627.

20. Lounis N, Maslo C, Boelaert JR, Bonnafous P, Truffot-PernotC, Baohong J, Grosset J. Impact of iron loading on murinetuberculosis. Clin Microbiol Infect 1999; 5:687–692.

21. Lounis N, Truffot-Pemot C, Grosset J, Gordeuk VR, Boelaert JR.Iron and Mycobacterium tuberculosis infection. J Clin Virol2001; 20:123–126.

22. Lounis N, Maslo C, Truffot-Pernot C, Grosset J, Boelaert JR.Impact of iron loading on the activity of isoniazid or ethambu-tol in the treatment of murine tuberculosis. Int J TuberculosisLung Dis 2003. In press.

23. Trousseau A. Lecture LXXXVII: true and false chlorosis. In:Lectures in Clinical Medicine. London, UK, 1872:95–117.


24. Murray MJ, Murray AB, Murray MB, Murray CJ. The adverseeffect of iron repletion on the course of certain infections.Br Med J 1978; 2:1113–1115.

25. De Monye C, Karcher DS, Boelaert JR, Gordeuk VR.Bone marrow macrophage iron grade and survival of HIV-seropositive patients. AIDS 1999; 13:375–380.

26. Gordeuk VR, McLaren CE, MacPhail AP, Deichsel G, BothwellTH. Associations of iron overload in Africa with hepatocellularcarcinoma and tuberculosis: Strachan’s 1929 thesis revisited.Blood 1996; 87:3470–3476.

27. Gangaidzo IT, Moyo VM, Mvundura E, Aggrey G, MurphreeNL, Khumalo H, Saugweme T, Kasvosve I, Gomo ZAR,Rouault T, Boelaert JR, Gordeuk VR. Iron stores, HIV anpulmonary tuberculosis. J Infect Dis 2001; 184:936–939.

28. Santos MM, de Sousa M, Rademakers LHPM, Clevers H,Marx JJM, Schilham MW. Iron overload and heart fibrosisin mice deficient for both b32-microglobulin and Ragl. Am JPathol 2000; 157:1883–1892.

29. Clerici M, Butto S, Lukwiya M, SareseHa M, Declich S,Trabattoni D, Pastori C, Piconi S, Fracasso C, Fabiani M,Ferrante P, Rizzardini G. Lopalco L for the Italian-UgandanAIDS project. Immune activation in Africa is environmentallydriven and is associated with upregulation of CCR5. AIDS2000; 14:2083–2092.

30. Bentwich Z, Maartens G, Torten D, Lai AA, Lai RB. Concur-rent infections and HIV pathogenesis. AIDS 2000; 14:2071–2081.

31. Boelaert JR, Weinberg GA, Weinberg ED. Altered iron meta-bolism in HIV infection: mechanisms, possible consequences,and proposals for management. Infect Agents Dis 1996;5:36–46.

32. Sappey C, Boelaert JR, Legrand-Poels S, Forceille C, Favier A,Piette J. Iron chelation decreases NF-kB and HIV-1 activationdue to oxidative stress. AIDS Res Hum Retroviruses 1995;11:1049–1061.

33. Cantinieaux B, Boelaert JR, De Meuleneire J, Kerrels V,Fondu P. Neutrophils from patients with secondary

628 Boelaert

haemosiderosis contain excessive amounts of autotoxic iron.Eur J Haematol 1993; 51:161–165.

34. Van Asbeck BS, Verbrugh HA, Van Oost BA, Marx JJM,Imhof HW, Verhoef J. Listeria monocytogenes meningitis anddecreased phagocytosis associated with iron overload. Br MedJ 1982; 284:542–544.

35. Weiss G, Wachter H, Fuchs D. Linkage of cell-mediated immu-nity to iron metabolism. Immunol Today 1995; 16:495–500.

36. De Sousa M. Immune cell functions in iron overload. Clin ExpImmunol 1989; 75:1–6.

37. Mencacci A, Cenci E, Boelaert JR, Bucci P, Mosci P, d’Ostiani F,Bistoni F, Romani L. Iron overload alters T helper cell responsesto Candida albicans in mice. J Infect Dis 1997; 175:1467–1476.

38. Simonart T, Noel J-C, Andrei G, Parent D, Van Vooren J-P,Hermans P, Lunardi-Yskandar Y, Lambert C, Dieye T, FarberC-M, Liesnard C, Snoeck R, Heenen M, Boelaert JR. Iron asa potential co-factor in the pathogenesis of Kaposi’s sarcoma.Lit J Cancer 1998; 78:720–726.

39. Simonart T, Boelaert JR, Andrei G, van den Oord JJ, Degraef C,Herlmans P, Noel J-C, Van Vooren J-P, HeenenM, De Clercq E,Snoeck R. Desferrioxamine enhances AIDS-associated Kaposi’ssarcoma tumor development in a xenograft model. Int J Cancer2002; 100:140–143.

40. Gomes MS, Boelaert JR, Appelberg R. Role of iron in experi-mental Mycobacterium avium infection. J Clin Virol 2001;20:117–122.

41. Barluzzi R, Saleppico S, Nocentini A, Boelaert JR, Bistoni F,Blasi E. Iron overload exacerbates experimental meningoence-phalitis by Cryptococcus neoformans. J Neuroimmunol 2002;132:140–146.

42. Taramelli D, Brambilla S, Sala G, Bruccoleri A, Tognazioli C,Riviera-Uzielli L, Boelaert JR. Effects of iron on extracellularand intracellular growth of Penicillium marneffei. InfectImmum 2000; 68:1724–1726.

43. Corbett EL, Steketee RW, ter Kuile FO, Latif AS, Kamali A,Hayes RJ. HIV-1=AIDS and the control of other infectiousdiseases in Africa. Lancet 2002; 359:2177–2187.


44. Ayisi JG, van Eijk AM, ter Kuile FO, Kolczak MS, Otieno JA,Misore AO, Kager PA, Steketee RW, Nahlen BL. The effect ofdual infection with HIV and malaria on pregnancy outcome inWestern Kenya. AIDS 2003; 17:585–594.

45. Gordeuk VR, Delanghe JR, Langlois MR, Boelaert JR. Ironstatus and the outcome of HIV infection: an overview. J ClinVirol 2001; 20:1111–1115.

46. Rockett KA, Brookes R, Udalova V, Vidal V, Hill AVS,Kwiatowski D. 1,25-dihydroxyvitamin D3 induces nitric oxidesynthase and suppresses growth of Mycobacterium tuberculo-sis in a human macrophage-like cell line. Infect Immun1998; 66:5314–5321.

47. Vanham G, Edmonds K, Qing L, Horn D, Toossi Z, Jones B,Daley CL, Huebner B, Kestens L, Gigase P, Ellner JJ.Generalized immune activation in pulmonary tuberculosis:co-activation with HIV infection. Clin Exp Immunol 1996;103:30–34.

48. Pennycook A, Openshaw P, Hussell T. Partners in crime: co-infections in the developing world. Clin Exp Immunol 2000;122:296–299.

49. Collins KR, Quinones-Mateu ME, Toossi Z, Arts EJ. Impactof tuberculosis on HIV-1 replication, diversity, and diseaseprogression. AIDS Rev 2002; 4:165–176.

50. Goletti D, Weissman D, Jackson RW, Graham NM, Vlahov D,Klein RS, Munsiff SS, Ortona L, Vauda R, Fauci AS. Effect ofMycobacterium tuberculosis on HIV replication. Role ofimmune activation. J Immunol 1996; 157:1271–1278.

51. Nicolas G, Bennoun M, Devaux I, Beaumont C, GrandchampB, Kahn A, Vaulont S. Lack of hepcidin gene expression andsevere tissue iron overload in upstream stimulatory 2 (USF2)knockout mice. Proc Natl Acad Sci USA 2001; 98:8780–8785.

52. Nicolas G, Bennoun M, Porteu A, Mativet S, Beaumont C,Grandchamp B, Sirito M, Sawadogo M, Kahn A, Vaulont S.Severe iron deficiency anemia in transgenic mice expressingliver hepcidin. Proc Natl Acad Sci USA 2002; 99:4596–4601.

53. Fleming RE, Sly WS. Hepcidin: a putative iron-regulatoryhormone relevant to hereditary hemochromatosis and the

630 Boelaert

anemia of chronic disease. Proc Natl Acad Sci USA 2001; 98:8160–8162.


55. Pigeon C, Ilyin G, Courselaud B, Leroyer P, Turlin B, BrissotP, Loreal O. A new mouse liver-specific gene, encoding a pro-tein homologous to human antimicrobial peptide hepcidin, isoverexpressed during iron overload. J Biol Chem 2001;276:7811–7819.

56. Shike H, Lauth X, Westerman ME, Ostland VE, Carlberg JM,Van JC I, Shimizu C, Bulet P, Bums JC. Bass hepcidin is anovel antimicrobial peptide induced by bacterial challenge.Eur J Bioch 2002; 480:147–150.

57. Krause A, Neitz S, Magert H-J, Schulz A, Forssmann W-G,Schulz-Knappe P, Adermann K. LEAP-1, a novel highlydisulfide-bonded human peptide, exhibits antimicrobial activ-ity. FEBS Lett 2000; 480:147–150.


59. Weinstein DA, Roy CN, Fleming MD, Loda MF, Wolfsdorf JI,Andrews NC. Inappropriate expression of hepcidin is asso-ciated with iron refractory anemia: implications for theanemia of chronic disease. Blood 2002; 100:3776–3781.


23

ACD in Inflammatory RheumaticDiseases

J.P. KALTWASSER and U. ARNDT

Abteilung Rheumatologie, Medizinische Klinik II,Zentrum der Inneren Medizin, der J.W. Goethe—

Universitat, Frankfurt, Germany

INTRODUCTION AND PATHOGENESIS

Anemia is a frequent concomitant feature in chronicinflammatory rheumatic diseases. Weakness, pallor, and fati-gue have been recognized as hallmarks of these diseases, andhemoglobin concentrations ranging from 9.0 to 12.0 g=dLhave often been considered ‘‘normal’’ for patients with chronicrheumatic illnesses, e.g., rheumatoid arthritis (1). In rheuma-toid arthritis, anemia is the most frequent extra-articularmanifestation of the disease (2–5). Anemia is also commonin all forms of juvenile chronic arthritis, systemic connective

633

tissue diseases [e.g., systemic lupus erythematosus (SLE),polymyositis, dermatomyositis, and scleroderma], vasculi-tides, as well as inflammatory spondyloarthropathies (6–9).Anemia is also seen in infection-related rheumatic diseases,e.g., septic arthritis, acute rheumatic fever, Lyme disease,and viral arthritis. The anemia in rheumatic diseases usuallydevelops slowly and nonprogressively. Its severity is to someextent correlated with the inflammatory activity of the rheu-matic disease (5). Different pathogenic mechanisms contri-bute to the development of anemia in rheumatic diseases.Although ACD is the most prominent among the varioustypes of anemia in these patients, one must take into accountother causes of anemia, e.g., blood losses, hemolysis, nutri-tional deficiencies, and toxic side effects of drug treatment(10). As also demonstrated in Chapter 3, proinflammatorycytokines such as TNF-a, IL-1b, and IL-6 play an importantrole in the pathophysiology of ACD (11). These cytokines

Figure 1 Proinflammatory and anti-inflammatory cytokines andcellular players of synovial inflammation in rheumatoid arthritis.

634 Kaltwasser and Arndt

are also central players in the still growing network of knowncytokine hematopoietic growth factors and other mediatorsinvolved in the pathogenesis of chronic inflammatory rheu-matic diseases such as rheumatoid arthritis (Fig. 1). Insuffi-cient erythropoietin synthesis in comparison to the degreeof anemia is an important pathogenic factor among the differ-ent pathways that contribute to ACD, and had already beendescribed for patients with rheumatoid arthritis in 1969 (12)(see also Fig. 2).

In this chapter, the heterogeneous group of inflammatoryrheumatic autoimmune disorders will be analyzed withrespect to the incidence and differential diagnosis of ACD.The laboratory requirements for diagnosis of ACD in this

Figure 2 Correlation of serum erythropoietin and hemoglobinconcentration in 23 anemic patients with chronic inflammatoryrheumatic diseases (�) in comparison to 365 measurements of 24healthy male controls (�) undergoing repeated phlebotomies untilthe complete exhaustion of body iron stores and development of amild iron deficiency anemia. Note the blunted erythropoietin reac-tion of most of the patients regarding the individual hemoglobinconcentration.

ACD in Inflammatory Rheumatic Diseases 635

group of diseases will be addressed as well as the treatmentmodalities presently in use. Specific attention will be drawnto erythropoietin treatment with regard to its effectivenessin correcting anemia, iron requirements during this therapy,and changes of health-related quality of life (HRQoL) in thisspecific setting of anemic patients with chronic inflammatoryrheumatic diseases.

INCIDENCE OF ANEMIA

The published frequency of anemia in patients with inflam-matory rheumatic diseases ranges from 15% up to over80%, depending on the underlying clinical condition and itsinflammatory activity. In rheumatoid arthritis, the publishedprevalence of anemia ranges between 20% and 70% (2–5). Forexample, in a longitudinal review of the records of 225 newlydiagnosed RA patients, anemia was found in 64%. Anemicpatients had higher disease activity, including more erosivejoint damage. Notably, anemia developed soon after the onsetof the disease. Ninety-three percent of the patients had ane-mia within the first year of follow-up (13). A case–controlstudy comparing female and male patients with ankylosingspondylitis detected anemia in 32% of the female and 8% ofthe male patients (9). In a cohort of 90 patients with giant cellarteritis or polymyalgia rheumatica at the time of diagnosis,more than 60% were anemic. In 28% of these patients, hemo-globin concentration was in the range below 10g=dL (7). Thepublished prevalence of anemia in patients with SLE variesbetween 15% (in a cohort with a disease duration >10years)(14) and 68% (6). It increases up to 84% in a subgroup of SLEpatients with special Fcg-receptor polymorphism, showinggenetic risk factors to influence disease severity includingthe frequency of anemia (15).

DIFFERENTIAL DIAGNOSIS OF ANEMIA

Different pathogenic conditions may contribute to the anemicstate in patients with inflammatory rheumatic diseases.


The most commonly occurring among the various forms ofanemia in these patients are the anemia of chronic disease(ACD), anemia due to iron deficiency (ID)—including chronicintestinal bleeding, vitamin B12, or folic acid deficiency,autoimmune hemolytic anemia (AHA), drug-induced hypopro-liferative anemia, and renal anemia. Several causes of anemiamay be present simultaneously. Vreugdenhil and colleaguesinvestigated 25 anemic RA patients and defined more thanone type of anemia in 68% of the patients. Forty-eight percentwere identified as having ACD only. Iron deficiency, asassessed by stainable bone marrow iron, was present in52%, folic acid deficiency in 21%, and vitamin B12 deficiencyin 29% (10).

Autoimmune hemolytic anemia plays a special role inSLE. Among 345 prospectively investigated consecutive SLEpatients, 38.3% were anemic, 37.1% of them classified ashaving ACD, 35.6% with ID, and 14.4% with AHA (16). In arecently published SLE study (17), the prevalence of severeAHA with hemoglobin concentrations below 8g=dL was even9.8% of all patients. Severe AHA was associated with anticar-diolipin antibodies and renal involvement in this study.

In a cohort of 180 patients with systemic sclerosis (PSS),the prevalence of anemia was 25% (8). In addition to ACD,anemia in these patients was attributed to bleeding of muco-sal telangiectases, microangiopathic hemolysis, and intestinalmalabsorption, indicating disease-specific pathogenic condi-tions being important in developing anemia in the differentrheumatic disorders, as seen here in the case of PSS.

As hereby demonstrated, anemia in inflammatory diseasesis a frequent but multicausal condition, which needs carefuldiagnostic investigation to select the optimal treatment forthe individual patient.

LABORATORY DIAGNOSIS

Suspicion of the presence of ACD arises in an anemic patientwith a chronic rheumatic disease reflected by clinical andlaboratory signs of chronic inflammation, when no other appar-ent causes of anemia can be detected. Diagnosis of ACD in


inflammatory rheumatic diseases is mostly made by exclusion,sincemany other factors influencing the development of anemiasuch as blood losses, hemolysis, vitamin deficiencies, and mye-lotoxic drug effects may be present concomitantly in thesepatients (see above paragraph). In general, laboratory indica-tors of ACD in inflammatory rheumatic diseases are not differ-ent from those of ACD in other diseases, e.g., malignancy,infection, or inflammatory bowel diseases (see Chapter 4).

Serologic nonspecific markers of disease activity ininflammatory rheumatic diseases are erythrocyte sedimenta-tion rate (ESR) and C-reactive protein (CRP). Most patientswith inflammatory rheumatic diseases have elevated ESRand=or CRP. Although there is a correlation between diseaseactivity and presence and severity of ACD (5), a considerableproportion of patients suffer from severe rheumatic disease,e.g., RA with radiographic progressive course, despite normalESR (18). In active systemic lupus erythematosus, CRP isusually not increased, and elevated CRP levels in patientswith SLE have to be regarded as an indicator of infectionrather than inflammation (19). Serologic testing for specificautoantibodies such as rheumatoid factor, antinuclear antibo-dies (ANA), DNA antibodies, antibodies against extractablenuclear antigens (ENAs), or antineutrophil cytoplasmic anti-bodies (ANCA) is of limited value for the diagnosis of ACDand for the discrimination from other types of anemia ininflammatory rheumatic diseases (20).

The most important problem in differential diagnosis ofACD in RA and other autoimmune rheumatic diseases isthe presence of concomitant ID, which is present in about ahalf of the anemic patients with RA (10). A set of iron- anderythropoiesis-related parameters has been evaluated con-cerning its value in detecting ID in patients with chronicrheumatic diseases, among them being serum ferritin, trans-ferrin iron saturation, reticulocytes, hypochromic eryth-rocytes, hemoglobin concentration of reticulocytes, solubletransferrin receptor, and zinc protoporphyrin (21–25). Never-theless, until now, there exist no clear cut-off levels for theseparameters to distinguish between ACD with or withoutconcomitant ID.


Serum iron is below normal (<14 mmol=L) in most ACDpatients and does not distinguish between ACD and ID. Inaddition, it underlies circadian variability influencing trans-ferrin saturation also, which is calculated from transferrinand serum iron (26). Transferrin and transferrin iron satura-tion are mostly lower than normal, while serum ferritin israised or is within the normal range (15–300mg=L). Since fer-ritin behaves as an acute phase reactant, chronic inflamma-tion can result in elevation of serum ferritin values forpatients with true ID into the normal range (27). With activeinflammation, plasma levels of ferritin can rise sharply inthe absence of any change in total body iron stores, placinglimitations on the use of serum ferritin as a marker of bodyiron stores. In addition to absolute ID with depleted or com-pletely exhausted iron stores, iron-deficient erythropoiesismay also occur despite adequate or even increased iron stores,then called functional iron deficiency (FID) (28). The percen-tage of erythrocytes with MHC <28pg (hypochromic erythro-cytes) is a useful indicator for iron-deficient erythropoiesis,provided an appropriate blood cell counter is available (24).A percentage of hypochromic erythrocytes greater than 10%is an accepted cut-off level for iron-deficient erythropoiesisin renal anemia. But although the red cells in ACD accompa-nying inflammatory rheumatic diseases are generally normo-chromic and normocytic, hypochromia and microcytosis (MCV<80fl) may be present even without concomitant ID, as hasbeen shown, e.g., in juvenile rheumatoid arthritis (29). Ery-throcyte ferritin has been shown to be a reliable parameterfor the detection of ID, but it seems to be of little value inthe detection of ACD (30). Reticulocyte counts may be slightlyincreased in ACD but are not helpful in discriminating ACDfrom ID. A significantly increased reticulocyte count in untre-ated patients with anemia in rheumatic diseases may ratherbe an indicator of the presence of concomitant hemolysisthan ACD.

The content of stainable iron (Prussian blue stain) inbone marrow aspirates may be a reliable alternative in differ-entiating ACD and ID. Since bone marrow aspiration is, how-ever, an invasive method and measurement of stainable iron


is a rather semiquantitative way to determine storage iron, ithas lost its importance as a golden standard for the discrimi-nation of ACD and ID. More recent indicators of body ironstatus and erythropoietic activity, e.g., the soluble transferrinreceptor (sTfR), erythrocyte zinc protoporphyrin (ZnPP), andthe hemoglobin concentration of reticulocytes, have also beenevaluated in different RA studies (21–25). But there exist nogenerally accepted cut-off levels for these parameters todetect iron-deficient erythropoiesis. In addition, the inflam-matory process itself may influence these parameters. Whilesome authors described an elevated sTfR as a reliable para-meter to detect ID in ACD (22), others observed elevated sTfRalso in patients with high amounts of storage iron (23). It hasto be noted that the expression of sTfR is regulated by cyto-kines, and thus its levels in ACD patients may be affectedby proinflammatory cytokines (11).

CLINICAL IMPACT OF ACD IN RHEUMATICDISORDERS

As with cancer patients, fatigue is also reported as one of themajor complaints of patients with different inflammatoryrheumatic diseases, including rheumatoid arthritis, ankylos-ing spondylitis, SLS, and Sjoegren’s syndrome (32). In 573patients with early RA, fatigue was identified as the conse-quence of disease that differentiated best on a series of qualityof life aspects such as disability, psychological well being,social support, and general well being (33). In patients withankylosing spondylitis, the main symptoms of an inflamma-tory flare are pain, immobility, and fatigue. In these patients,fatigue is associated with the level of disease activity, func-tional ability, global well being, and mental health status(34). Fatigue appears to be multifactorial and multidimen-sional. Physiological, psychological, social, and personal fac-tors influence the presence and experience of fatigue and itsnegative effect on quality of life. Whilst it is acknowledgedthat severe anemia causes fatigue, it is more difficult to showevidence of a causal relationship between fatigue and mild


chronic anemia. Glaus and Mueller (35) investigated 444cancer patients and demonstrated the grade of anemiacorrelating with the intensity of fatigue. Comparable investi-gations have not been undertaken so far with patients withinflammatory rheumatic diseases. Although fatigue is iden-tified as a very important factor influencing HRQoL of pat-ients with rheumatic diseases, there exist only rare dataconcerning the correlation of anemia with fatigue and theeffect of therapeutic interventions on fatigue. In differentclinical trials, correction of anemia in rheumatoid arthritisby application of human recombinant erythropoietin has beendemonstrated to increase energy, vitality, muscle strength,and general well being, and reduced fatigue and patients’need for daily rest (31,36–40) (see the section on Treatmentof ACD). This improvement of HRQoL via treating anemiawas observed not only for patients with severe anemia, butalso even for those with only slightly reduced levels of hemo-globin concentration. This underlines the clinical importanceof anemia of chronic disease and identifies its treatment asone relevant aspect of reaching the main treatment goal forevery chronically ill patient: increasing HRQoL.

TREATMENT OF ACD

As in all chronic disorders, treatment of the underlying diseaseis the first and main opportunity for a successful treatment ofACD. For example, treatment of inflammatory active RA witha chimeric antibody to TNFa (cA2¼Remicade�) not onlyresulted in significant improvements in disease activity mea-sures, but also in a dose-dependent increase of hemoglobinconcentration (5). In giant cell arteritis and polymyalgia rheu-matica, effective steroid therapy led to a fast correction ofanemia in nearly all patients. The same antianemic effect ofcorticosteroid treatment was observed in RA patients followinghigh-dose oral steroid therapy (41).

However, sufficient control of the underlying disease ispresently still not achievable in all patients. In rheumatoidarthritis for example, response to treatment is classified by


special response criteria, the ACR-response criteria (42) of theAmerican College of Rheumatology or, for example, the Pau-lus response criteria (43), both including the number of swol-len and tender joints and other clinical symptoms as well asparameters of systemic inflammation (CRP and ESR). Com-plete remission of the inflammatory disease is rarelyachieved. By applying new TNFa- and IL-1-inhibiting agents,e.g., Remicade�, Enbrel�, and Kineret, or by combining dis-ease modifying antirheumatic drugs (DMARDs), e.g., metho-trexate, sulfasalazine, and hydroxychloroquine, many patients(up to 60–80%) fulfill the ACR-20=50 response criteria(44–46), which include a reduction of the number of swollenand tender joints by 20=50%, but most of them still havesigns of disease activity. Until cure or even complete long-time remission of inflammatory rheumatic diseases is notfeasible, we need special therapeutic options for treating anemiaaccompanying inflammatory rheumatic diseases.

Blood transfusion offers an option for fast correction ofsevere anemia. Despite modern laboratory tests to ensurethe safety of blood transfusion, there still remains a risk oftransmitting infectious diseases or inducing side effects likeanaphylactic reactions. Possible impacts of blood transfusionon the immune system are not fully understood in autoim-mune rheumatic diseases. The supply of donor blood isdiscussed to induce energy, which can be referred to down-regulation of Th-1 and stimulation of Th-2 cell-mediatedimmune effector pathways (47). One Dutch study (48)has investigated the effect of partially HLA-matched bloodtransfusion on the disease activity of RA, showing someimprovement of RA according to the ACR response criteriaafter blood transfusion compared to placebo infusion. Thisimprovement was slight and reached significance only at1month after transfusion (one out of four examined timepoints ranging from 2weeks to 6months). Further studieswould be necessary to understand possible anti-inflammatoryor immunosuppressive effects of blood transfusion in RApatients, made even more difficult by the need of removingwhite cells from the transfused blood in many countries,due to the theoretical risk of transmitting diseases, e.g.,


Creutzfeld-Jacob disease. The risk of iron overload—with allits negative consequences—by repeated transfusions isanother reason for avoiding frequent transfusions. The clini-cal value of blood transfusion in acute anemia is evident,but it seems not to be an adequate therapy in chronic and=ormild anemia.

While iron administration is effective in correctinganemia caused by pure ID, its value as single drug treatmentin ACD is rather limited. Intravenous iron dextran evenresulted in flares of rheumatoid arthritis in some former stu-dies (49–51). In addition, intra-articular ferritin-bound iron inRA patients is increased, in synovial fluid as well as in syno-vial membrane (52). Since iron is regarded as a potentialproinflammatory agent in RA, iron chelators such as desfer-rioxamine (DFO) and deferiprone (L1) have been tested aspossible anti-inflammatory drugs in rheumatoid arthritis inseveral small studies. Desferrioxamine was administered sub-cutaneously in 10 patients for 14 days (53) and in five patientsfor 4weeks (54) at different dosages (500 and 1000mg daily).Both studies observed an increase of hemoglobin concentra-tion, which seems to be dose dependent, as well as a moderateimprovement of clinical disease activity measured, e.g., bypain intensity. The background for this is not known so farbut may relate to the fact that DFO increases EPO expressionpossibly by interfering with HIF-1 activity. However, the useof DFO in the treatment of ACD is limited not only because ofits uncomfortable continuous subcutaneous application, butalso due to a lot of possible toxic side effects of this drug. Asknown from treatment of thalassemia and other iron-overloaddiseases, the oral iron-chelator L1 has comparable efficacy inmobilization of iron stores, it is easier to administer by theoral route. However, possible side effects such as agranulocy-tosis warrant careful and close supervision (55). L1 short-timetherapy has been tested in 10 RA patients with ACD andresulted in an increase of hemoglobin concentration in sevenpatients, accompanied by decreased ferritin- and increasederythropoietin-levels without obvious signs of toxicity after 1week of treatment (56). To our knowledge, no data existconcerning possible anti-inflammatory effects of L1 in RA,


but some side effects of this drug, e.g., joint pain and swelling,SLE, and vasculitis (57–59) have been observed.

Erythropoietin because of its known role in the patho-genesis of ACD was already discussed to be an interestingtherapeutic option before it was available as an antianemicdrug. Allan J. Erslev wrote in 1983: ‘‘Erythropoietin, if avail-able in a therapeutic form, would undoubtedly increasethe rate of red cell production, but it could aggravate the hypo-ferremia and thereby be of very questionable therapeuticbenefit’’ (1).

The blunted erythropoietin production (see Fig. 2) inACD patients, which has already been described years beforerHu-Epo was available (12,60), was the pathogenic back-ground for the first successful treatment of two anemic RApatients with recombinant human erythropoietin (rHu-Epo) in1989 (61). Until now, rHu-Epo has proved to be successful inimproving ACD in different clinical RA trials (31,36–39,62–66).Prior to elective operative intervention, e.g., synovectomy oreven joint replacement, treatment with rHu-Epo can increasepreoperative hemoglobin concentration and thereby reducethe pre- and postoperative risks (67). In addition, this ther-apy is used to assist presurgical autologous blood donationin patients undergoing major orthopedic surgery with exp-ected relevant blood loss (68,69).

While in earlier studies rHu-Epo was supplemented intra-venously (36,61–63,68), this route of administration was repla-ced since 1992 by subcutaneous application (31,37–39,64–66)because of the more stabilized Epo plasma levels (70,71). Dueto plasma half-life of 48–72hr, rHu-Epo should be injectedtwo or three times a week (72). The dosages of rHu-Epo inthe different RA trials were higher than those in renal ane-mia, ranging from 300 to 1250 IE per kg body weight weekly.Despite this use of high rHu-Epo dosages, rates of non-responders up to 85.4% of the treated patients (66) havebeen observed. As already predicted by Erslev in 1983, dis-turbances of iron metabolism might be of special importanceduring erythropoietin substitution. Insufficient iron supplywith development of FID is discussed to be one major reasonfor limited responsiveness to rHu-Epo. In contrast to absolute


ID with depleted or completely exhausted iron stores, FIDdefines a state of iron-deficient erythropoiesis despiteadequate or even increased iron stores (28). Functional irondeficiency may occur in a situation of increased erythropoi-esis, e.g., by application of rHu-Epo and=or by limitation ofiron release from the reticulo-endothelial system due toinflammatory processes. In renal anemia, i.v. iron supply incase of FID led to reduction of the mean rHu-Epo require-ment by up to 70% (73). A recently published uncontrolledRA study investigated the development of FID during rHu-Epo therapy (150 IE=kg twice weekly) and supplementedi.v. iron-sucrose (200mg weekly) in case of FID. Eighty-threepercent of the patients in this study received i.v. iron due topre-existing or developing FID. Treatment with rHu-Epo� i.v. iron was effective—without nonresponders to treat-ment—and well tolerated in all patients (31), indicating i.v.iron-sucrose substitution being important, effective, andwithout proinflammatory activity in this special situation ofrHu-Epo-induced increased erythropoiesis.

In addition to correction of anemia, the impact of rHu-Epotherapy on HRQoL was investigated in different RA studies.An increase of general well being was observed (36,37,39) aswell as significant improvements of several fatigue-relatedparameters, e.g., an increase of energy, measured by the useof the Nottingham health profile score (39) or higher levels ofthe vitality subscale of SF-36 (31) (Fig. 3). These positiveeffects on HRQoL were observed even in patients with onlyslightly reduced hemoglobin levels. Possible effects of rHu-Epoon disease activity were not investigated systematically for along time, although Gudbjornsson et al. (37) had alreadydescribed a reduced number of painful joints in 7 out of 10RA patients during rHu-Epo therapy in 1992. In a larger pla-cebo-controlled, double-blind study, Peeters et al. (65) investi-gated short- and long-term effects of rHu-Epo not only onhemoglobin but also on disease activity. Compared to the pla-cebo group, significant improvements in the Epo group wereobserved regarding the number of swollen and tender joints,patients’ global assessment of disease activity, pain score,and response to treatment according to Paulus 20% response


(8%vs. 32%, see Fig. 4). This unexpected decrease of RA dis-ease activity was confirmed in the already mentioned, uncon-trolled trial with rHu-Epo � i.v. iron therapy (31).

The observed anti-inflammatory effect of erythropoietinneeds further confirmation. Preliminary data indicate a possibleeffect of erythropoietin on synovial fibroblasts via the ery-thropoietin receptor and a reduction of prostaglandin E2-production in the inflamed joint by reduced arachidonic acidrelease (74).

Due to the chronic inflammatory process in rheumaticdisease, anemia mostly relapses within weeks, independentof whether anemia was corrected by blood transfusion, ironchelators, or rHu-Epo therapy. As demonstrated by Peeterset al. (65), maintenance therapy with a reduced dosage ofrHu-Epo following correction phase can stabilize normalhemoglobin concentration as well as the increased HRQoL.

Figure 3 Changes of percentage of the vitality scale of SF 36(hatched bars) and course of hemoglobin concentration (filled cir-cles) during 3 months of treatment with rHu-Epo (150U=kg tws.c.) and within 3months after termination of erythropoietin treat-ment in 28 patients with ACD and rheumatoid arthritis. In case ofthe development of functional iron deficiency, rHu-Epo was com-bined with i.v. administration of 200mg of iron sucrose per week.(From Ref. 31.)


Further clinical trials are needed to define the optimaldosages of rHu-Epo and requirements of iron supplementa-tion ensuring an effective, safe, and economical treatmentschedule for correcting ACD in patients with RA and otherinflammatory rheumatic diseases.

SUMMARY AND POSSIBLE THERAPEUTICFUTURE OPTIONS

There is convincing clinical evidence, that ACD is one ofthe major causes of anemia in inflammatory autoimmunerheumatic diseases.

A great deal of knowledge has been gathered so farconcerning pathogenesis, diagnosis, and treatment of ACD.Effects of the proinflammatory cytokine network, resultingin a restricted iron release from the storage department and

Figure 4 Percentage of patients in a placebo-controlled study ful-filling the criteria for a Paulus 20% response. Comparison of 34 RApatients treated with 240 U=kg rHu-Epo s.c. three times a week(drawn line) for 52 weeks, with 36 patients receiving placebo (dottedline). �p< 0.001; ��p< 0.01. (According to Ref. 65.)


a blunted response of erythropoietin production, are identifiedfeatures of the complex pathogenetic process underlying theerythropoietic hypoproliferation in ACD in various rheumaticdiseases. Inhibition of pro-inflammatory cytokines, e.g., TNFaby monoclonal antibodies or TNF receptors has proven to beeffective not only in controlling disease activity but also in cor-recting the anemia of rheumatoid arthritis and other inflam-matory rheumatic diseases. Anemia in chronic rheumaticdiseases is, however, not always and not only caused by theeffects of chronic inflammation but may be caused by ironand other deficiencies, hemolysis and toxic drug effects onbone marrow proliferation, or by a combination of some ofthese factors. The diagnosis of ACD therefore has to be madeby exclusion of other causes of anemia. Iron deficiency is themost important differential diagnosis for ACD and has to beexcluded by demonstrating normal or increased iron stores.Functional iron deficiency is another common feature inpatients with ACD and may be detected best by measurementof the proportion of hypochromic red cells in the peripheralblood or by combining serum ferritin determination withpara-meters like soluble transferrin receptor and red cell zincprotoporphyrin as indicators of an iron-deficient erythropoi-esis.

Due to its impact on quality of life, anemia is presentlyregarded much more as an important feature for patients withinflammatory rheumatic diseases than it has been done in thepast. If treatment of the underlying disease is not achievable,ACD may be treated by blood transfusion. Substitution of ironin pure ACD and even in combination with ID is of limitedvalue. Among other treatment modalities, the treatment withrHu-Epo has been shown to be the most suitable and effectiveway to treat ACD. Functional iron deficiency, induced by denovo erythropoiesis due to stimulation by erythropoietin, istherefore a common feature that can limit the therapeuticeffect of rHu-Epo. Substitution of iron is hence a frequentrequirement in the treatment of ACD with erythropoietin. Inthe induction phase of erythropoietin treatment of ACD, i.v.iron application as in renal anemia is more effective thanthe oral route. It has been shown that iron supplementation


does not increase inflammatory activity, e.g., in rheumatoidarthritis, when given in combination with erythropoietin.

Quality of life can be increased substantially by correct-ing ACD with erythropoietin and even disease activity maybe reduced in inflammatory active rheumatic diseases. Thereis no clear definition so far of the most effective dose of ery-thropoietin in ACD treatment in patients with rheumaticdiseases. Newly developed derivatives of recombinant ery-thropoietin, e.g., darbepoetin (for details please refer to thechapters ‘‘Erythropoietin’’ and ‘‘Iron and erythropoietin’’)are promising with respect to simplifying its therapeuticapplication but have to be tested for an appropriate use inACD. Other indications for erythropoietin treatment of ACDin rheumatic diseases, e.g., perisurgical correction of anemiamay also be considered.

Therefore, further clinical trials with rHu-Epo and espe-cially with the new erythropoietic protein darbepoetin areneeded to define the optimal effective, safe, and economictreatment schedule for the correction of ACD in patients withRA and other inflammatory rheumatic diseases.

REFERENCES

1. Erslev AJ. Erythrocyte disorders—anemias related tounknown or multiple mechanisms. In: Williams WJ, Beutler E,Erselev AJ, Lichtman MA, eds. Hematology. 3rd ed. McGraw-Hill Book Company, 1983:522–528.

2. Mowat AG. Haematologic abnormalities in rheumatoid arthri-tis. Sem Arhritis Rheum 1971; 1:195–219.

3. Bennett RM. Haematological changes in rheumatoid disease.Clin Rheum Dis 1977; 3:433–465.

4. Remacha AF, Rodriguez-De La Serna A, Garcia-Die F, Geli C,Diaz C, Gimferrer E. Erythroid abnormalities in rheumatoidarthritis: the role of erythropoietin. J Rheumatol 1992; 19:1687–1691.

5. Davis D, Charles PJ, Potter A, Feldmann M, Maini N, ElliottMJ. Anaemia of chronic disease in rheumatoid arthritis: in vivo


effects of tumor necrosis a blockade. Br J Rheumatol 1997;36:50–56.

6. Vila LM, Mayor AM, Valentin AH, Garcia-Soberal M, Vila S.Clinical and immunological manifestations in 134 PuertoRican patients with systemic lupus erythematosus. Lupus1999; 8:279–286.

7. Rivero Puente A, Berasategui Calderon JI, Murie Carrillo deAlbornoz JM, Rivero Marcotegui M, Garcia-Bragado Acin F.Giant cell arteritis and polymyalgia rheumatica: diagnosisand evolution of 90 cases. An Med Interna 2001; 18:191–194.

8. Frayha RA, Shulman LE, Stevens MB. Hematological abnorm-alities in scleroderma. A study of 180 cases. Acta Haematol1980; 64:25–30.

9. Marks SH, Barnett M, Calin A. Ankylosing spondylitis inwomen and men: a case–control study. J Rheumatol 1983;10:624–628.

10. Vreugdenhil G, Wognum AW, van Eijk HG, Swaak AJG. Anae-mia in rheumatoid arthritis: the role of iron, vitamin B12, andfolic acid deficiency, and erythropoietin responsiveness. AnnRheum Dis 1990; 49:93–98.

11. Weiss G. Pathogenesis and treatment of anaemia of chronicdiseases. Blood Rev 2002; 16:87–96.

12. Ward HP, Gordon B, Pisarczyk JC. Serum levels of erythro-poietin in anemia associated with chronic infection, malig-nancy, and primary hematopoietic disease. J Clin Invest1971; 50:332–335.

13. Peeters HR, Jongen-Lavrencic M, Raja AN, Ramdin HS,Vreugdenhil G, Breedveld FC. Course and characteristics ofanemia in patients with RA of recent onset. Ann Rheum Dis1996; 55:162–168.

14. Swaak AJG, van den Brink HG, Smeenk RJ, Manger K,Kaleden JR, Tosi S, Domljan Z, Rozman B, Logar D, Pokorny G,Kovacs L, Vlachoyiannopoulos PG, Moutsopoulos HM,Chwalinska-Sadowska H, Kiss E, Cikes N, Anic B, Schneider M,Fischer R, Bombardieri S, Mosca M, Graninger W, Smolen JS.Systemic lupus erythematosus. Disease outcome in patients with


a disease duration at least 10years: second evaluation. Lupus2001; 10:51–58.

15. Manger K, Repp R, Jansen M, Geisselbrecht M, Wassmuth R,Westerdaal NAC, Pfahlberg A, Manger B, Kalden JR, van deWinkel JGJ. Ann Rheum Dis 2002; 61:786–792.

16. Voulgarelis M, Kokori SIG, Ioannidis JPA, Tzioufas AG,Kyriaki D, Moutsopoulos HM. Anaemia in systemic lupuserythematosus: aetiological profile and role of erythropoietin.Ann Rheum Dis 2000; 59:217–222.

17. Sultan SM, Begum S, Isenberg DA. Prevalence, patternsof disease and outcome in patients with systemic lupuserythematosus who develop severe haematological problems.Rheumatology 2003; 42:230–234.

18. Wolfe F, Michaud K. The clinical and research significance ofthe erythrocyte sedimentation-rate. J Rheumatol 1994; 21:1227–1237.

19. Pereira DA, Silva JA, Elkon KB, Hughes GRV. C-reactive pro-tein levels in systemic lupus erythematosus: a classficationcriterion? Arthritis Rheum 1980; 23:770–774.

20. PincusT. Laboratory tests in rheumatic disorders. In:Klippel JH,Duppe PA, eds. Rheumatology. 2nd ed. London, Philadelphia,St. Louis, Sydney, Tokyo: Mosby, 1998 [chapter 10].

21. Baumann Kurer S, Seifert B, Michel B, Ruegg R, Fuhr J. Pre-diction of iron deficiency in chronic inflammatory rheumaticdisease anaemia. Br J Haematol 1995; 91:820–826.

22. Suominen P, Mottonen T, Rajamaki A, Irjala K. Single valuesof serum transferrin receptor and transferrin receptor ferritinindex can be used to detect true and functional iron deficiencyin rheumatoid arthritis patients with anaemia. ArthritisRheum 2000; 43:1016–1020.

23. Feelders RA, Vreugdenhil G, Manger B, van Eijk HG,Swaak AJG. Serum transferrin receptor levels in anaemia ofrheumatoid arthritis and effects of iron chelation treatment.Eur J Haematol 1994; 52:61–62.

24. Hastka J, Lasserre JJ, SchwarzbeckA, StrauchM,HehlmannR.Zinc protoporphyrin in anemia of chronic disorders. Blood 1993;81:1200–1204.


25. Garrett J, Worwood M. Zinc protoporphyrin and iron-deficienterythropoiesis. Acta Haematol 1994; 91:21–25.


27. Lipschitz DA, Cook JD, Finch CA. A clinical evaluation ofserum ferritin. N Engl J Med 1974; 290:1213–1216.

28. Eschbach JW,Egrie JC,DowningMR, Browne JK, Adamson JW.Correction of anaemia of end-stage renal disease with recombi-nant human erythropoietin. Results of a combined phase I andII clinical trial. N Engl J Med 1987; 316:73–78.

29. Koerper MA, Stempel DA, Dallman PR. Anemia in patientswith juvenile arthritis. J Pediatr 1978; 92:930–933.

30. Davidson A, van derv Weyden MN, Fong H, Breidahl MJ,Ryan PFJ. Red cell ferritin content: a re-evaluation of indicesfor iron deficiency in the anaemia of rheumatoid arthritis. BrMed J 1984; 284:648–650.

31. Kaltwasser JP, Kessler U, Gottschalk R, Stucki G, Moller B.Effect of recombinant human erythropoietin and intravenousiron on anemia and disease activity in rheumatoid arthritis.J Rheumatol 2001; 28:2430–2436.

32. Lwin CT, Bishay M, Platts RG, Booth DA, Bowman SJ. Theassessment of fatigue in primary Sjogren’s syndrome. ScandJ Rheumatol 2003; 32:33–37.

33. Suurmeijer TP, Waltz M, Moum T, Guillemin F, van SonderenFL, Briancon S, Sanderman R, van den Heuvel WJ. Quality oflife profiles in the first years of rheumatoid arthritis: resultsfrom the EURIDISS longitudinal study. Arthritis Rheum2001; 45:111–121.

34. Brophy S, Cali A. Definition of disease flare in ankylosingspondylitis: the patient’s perspective. J Rheumatol 2002; 29:954–958.

35. Glaus A, Muller S. Haemoglobin and fatigue in cancer patients:inseparable twins? Schwei Med Wochenschr 2000; 130:471–477.

36. Salvarani C, Lasagni D, Casali B, Macchioni P, Boiardi L,Rossi F, Rivasi P, Portioli I. Recombinant human erythropoie-


tin therapy in patients with the anemia of chronic disease.J Rheumatol 1991; 18:1168–1171.

37. Gudbjornsson B, Hallgren R, Wide L, Birgegard G. Responseof anaemia in rheumatoid arthritis to treatment with subcuta-neous recombinant human erythropoietin. Ann Rheum Dis1992; 51:747–752.

38. Petterson T, Rosenloef K, Friman C, Mickos A, Teppo AM,Fyhrquist F. Successful treatment of the anemia of rheuma-toid arthritis with subcutaneously administered recombinanthuman erythropoietin. Slower response in patients with moresevere inflammation. Scand J Rheumatol 1993; 22:188–193.

39. Murphy EA, Bell AL, Wojtulewski J, Brzeski M, Madhok R,Capell HA. Study of erythropoietin in treatment of anaemiain patients with rheumatoid arthritis. Br J Med 1994; 309:1337–1338.

40. Peeters HR, Jongen-Lavrencic M, Bakker CH, Vreugdenhil G,Breedveld FC, Swaak AJG. Recombinant human erythropoie-tin improves health-related quality of life in patients withrheumatoid arthritis and anaemia of chronic disease; utilitymeasures correlate strongly with disease activity measures.Rheumatol Int 1999; 18:201–206.

41. Kendall R, Wasti A, Harvey A, Hill J, Chapman C, Norfolk DR,Pullar T. The relationship of hemoglobin to serum erythropoietinconcentrations in the anaemia of rheumatoid arthritis: the effectof oral prednisolone. Br J Rheumatol 1993; 32:204–208.

42. Felson DT, Anderson JJ, Boers M. American College of Rheu-matology preliminary definition of improvement in rheuma-toid arthritis. Arthritis Rheum 1995; 38:727–735.

43. Paulus HE, Egger MJ, Ward JR, Williams HJ. CSSRD: analy-sis of improvement in individual rheumatoid arthritis patientstreated with disease-modifying antirheumatic drugs, based onthe findings in patients treated with placebo. Arthritis Rheum1990; 33:477–484.

44. Maini RN, Breedveld FC, Kalden JR, et al. The therapeuticefficacy of multiple intravenous infusions of anti-tumor necro-sis factor monoclonal antibody combined with low-dose weeklymethotrexate in rheumatoid arthritis. Arthritis Rheum 1998;41:1552–1563.


45. Moreland LM, Cohen SB, Baumgartner S, et al. Long-term useof etanercept in patients with DMARD-refractory rheumatoidarthritis [abstr]. Arthritis Rheum 1999; 42:S401.

46. O’Dell JR, Haire CE, Erikson N, Drymalski W, Palmer W,Eckhoff PJ, Garwood V, Maloley L, Klassen LW, Wees S,Klein H, Moore GF. Treatment of rheumatoid arthritis withmethotrexate alone, sulfasalazine and hydroxychloroquine,or a combination of all three medications. N Engl J Med1996; 334:1287–1291.

47. Eleftherios C, Vamvakas EC, Morris R, Blajchman A. Deleter-ious clinical effects of transfusion-associated immunomodula-tion: fact or fiction? Blood 2001; 97:1180–1195.

48. van der Horst-Bruinsma IE, Huizinga TWJ, Lagaay EM, HazesJMW, Breevfeld FC, Schreuder GMT, Tomson TAS, Zwinder-man AH, van Rood JJ, de Vries RRP. The influence of partiallyHLA-matched blood transfusion on the disease activity of rheu-matoid arthritis. Rheumatology 1999; 38:53–58.

49. Reddy PS, Lewis M. The adverse effect of intravenousiron-dextran in rheumatoid arthritis. Arthritis Rheum 1969;12:454–457.

50. Lloyd KN, Williams P. Reactions to total dose infusion of irondextran in rheumatoid arthritis. Br J Med 1970; 2:323–325.

51. Blake DR, Lunec J, Ahern M, Ring EFJ, Bradfield J,Gutteridge JMC. Effect of intravenous iron dextran on rheu-matoid synovitis. Ann Rheum Dis 1985; 44:183–188.

52. Biemond P, Swaak AJG, van Eijk HG, Koster JF. Intraarticu-lar ferritin-bound iron in rheumatoid arthritis. A factor thatincreases oxygen free radical-induced tissue destruction.Arthritis Rheum 1986; 29:1187–1193.

53. Giordano N, Sancasciani S, Borghi C, Fioravanti A,Marcolongo R. Antianemic and potential anti-inflammatoryactivity of desferrioxamine: possible usefulness in rheumatoidarthritis. Clin Exp Rheumatol 1986; 4:25–29.

54. Salvarani C, Baricchi R, Lasagni D, Boiardi L, Piccinini R,Brunati C, Macchioni P, Portioli I. Effects of desferrioxaminetherapy on chronic disease anemia associated with rheumatoidarthritis. Rheumatol Int 1996; 16:45–48.


55. Taher A, Sheikh-Taha M, Koussa S, Inati A, Neeman R,Mourad F. Comparison between deferoxamine and deferiprone(L1) in iron-loaded thalassemia patients. Eur J Hematol 2001;67:30–34.

56. Vreugdenhil G, Kontoghiorghes GJ, van Eijk HG, Swaak AJG.Impaired erythropoietin responsiveness to the anaemia inrheumatoid arthritis: a possible inverse relationship with ironstores and effects of the oral iron chelator 1,2-dimethyl-3-hydroxypyrid-4-one. Clin Exp Rheum 1992; 9:35–40.

57. Castriota-Scanderberg A, Sacco M. Agranulocytosis, arthritisand systemic vasculitis in a patient receiving the oral ironchelator L1 (deferiprone). Br J Hematol 1997; 96:254–255.

58. Berkovitch M, Laxer RM, Inman R, Koren G, Pritzker KP,Fritzler MJ, Olivieri NF. Arthropathy in thalassaemiapatients receiving deferiprone. Lancet 1994; 343:1471–1472.

59. Mehta J, Singhal S, Revankar R, Walvalkar A, Chablani A,Mehta BC. Fatal systemic lupus erythematosus in patientstaking oral iron chelator L1. Lancet 1991; 337:298.

60. Pavlovic-Kentera V, Ruvidic R, Milenkovic P, Marinkovic D.Erythropoietin in patients with anaemia in rheumatoidarthritis. Scand J Haematol 1979; 23:141–145.

61. Means RT, Olsen NJ, Krantz SB, Dessypris EN, Graber SE,Stone WJ, O’Neil VL, Pincus T. Treatment of the anemia ofrheumatoid arthritis with recombinant human erythropoietin:clinical and in vitro studies. Arthritis Rheum 1989; 32:638–642.

62. Pincus T, Olsen NJ, Russell J, Wolfe F, Harris ER, Schnitzer TJ,Boccagno JA, Krantz SB. Multicenter study of recombinanthuman erythropoietin in correction of anemia in rheumatoidarthritis. Am J Med 1990; 89:161–168.

63. Takashina N, Kondo H, Kashiwaszaki S. Suppressed serumerythropoietin response to anemia and the efficacy of recombi-nant erythropoietin in the anemia of rheumatoid arthritis.J Rheumatol 1990; 17:885–887.

64. Kato Y, Takagi C, Tanaka J, Masaki Y, Furuya H. Effect ofdaily subcutaneous administration of recombinant erythro-poietin on chronic anemia in rheumatoid arthritis. Int Med1994; 33:193–197.


65. Peeters HRM, Jongen-Lavrencic M, Vreugdenhil G, SwaakAJG. Effect of recombinant human erythropoietin on anaemiaand disease activity in patients with rheumatoid arthritis andanaemia of chronic disease: a randomised placebo controlleddouble blind 52 weeks clinical trial. Ann Rheum Dis 1996;55:739–744.

66. Nordstroem D, Lindroth Y, Marsal L, Hafstroem I, Henrich C,Rantapaa-Dahlqvist S, Engstroem-Laurent A, Fyhrquist F,Friman C. Availability of iron and degree of inflammationmodifies the response to recombinant human erythropoietinwhen treating anemia of chronic disease in patients with rheu-matoid arthritis. Rheumatol Int 1997; 17:67–73.

67. Nierman E, Zakrzewski K. Recognition and managementof preoperative risk. Rheum Dis Clin North Am 1999; 25:585–622.

68. Mercuriali F, Gualtieri G, Singaglia L, Inghilleri G, Biffi E,Vinci A, Colotti MT, Barosi G, Lambertenghi Deliliers G. Useof recombinant human erythropoietin to assist autologous blooddonation by anemic rheumatoid arthritis patients undergoingmajor orthopedic surgery. Transfusion 1994; 34:501–506.

69. Matsui H, Shiraishi N, Yasuda T, Nezuka T. The effects ofrecombinant human erythropoietin on autologous blood dona-tion in rheumatoid arthritis patients with anaemia. Clin ExpRheumatol 1999; 17:69–74.

70. Bommer J, Ritz E, Weinreich T, Bommer G, Ziegler T. Subcu-taneous erythropoietin. Lancet 1988; II:406.

71. Besareb A, Flaharty KK, Erslev AJ, McCrea JB, Vlasses PH,Medina F. Clinical pharmacology and economics of recombi-nant human erythropoietin in end-stage renal disease: thecase for subcutaneous administration. J Am Soc Nephrol1992; 2:1405–1416.

72. Kampf D, Kahl A, Passlick J, Pustelnik J, Eckardt KU, Ehmer B,Jacobs C, Baumelou A, Grabensee B, Gahl GM. Single-dosekinetics of recombinant human erythropoietin after intravenous,subcutaneous and intraperitoneal administration. ContribNephrol 1989; 76:106–111.


73. Sunder-Plassmann G, Hoerl WH. Importance of iron supplyfor erythropoietin therapy. Nephrol Dial Transplant 1995;10:2070–2076.

74. Kessler U, Betschel S, Woelfer S, Kaltwasser JP, Moeller B,Kaszkin M. Erythropoietin inhibits IL-1b stimulated arachido-nic acid release from human synovial fibroblasts [abstr].Arthritis Rheum 2002; 46:S259.


24

Anemia in Intensive Care Patients

ALBERT VAN DE WIEL

Department of Internal Medicine,Meander Medical Center, Amersfoort,

The Netherlands

INTRODUCTION

Acute events in medicine often lead to anemia. Many patientsbecome anemic not only after surgery but also during sepsisand during their stay in an intensive care unit (ICU). After3 days on ICU, about 95% of patients have hemoglobin con-centrations below normal (1). At first sight, this seems theresult of blood loss including frequent blood drawing, invasiveprocedures, and gastrointestinal blood loss. But apart fromhypoferremia, the acute event-related anemia, as observedin ICU patients, bears no resemblance to iron deficiency (2).With high ferritin and low-to-normal transferrin and serumtransferrin receptor levels, it shares characteristics with theanemia of chronic disease as found in patients with infectious

659

diseases, inflammatory disorders, and malignancies. This sug-gests other mechanisms to be operative as well such as ablunted erythropoietin (EPO) response and suppression oferythropoiesis by inflammatory cytokines.

No general guidelines exist for the optimal hemoglobinconcentration or hematocrit in ICU patients. A low hemoglo-bin concentration impairs, among others, oxygen availabilityto the tissues, whereas a high hematocrit may be unfavorablefrom a rheologic point of view. Because of disadvantagesrelated to blood transfusions and growing insights into thepathogenesis of this type of anemia, new strategies such asthe application of recombinant human EPO (rh-EPO) havebeen introduced into intensive care medicine.

This review focuses on the possible mechanisms involvedin the anemia of the ICU patient and new developments forits treatment.

MECHANISMS

Blood Loss

In ICU, patients blood loss is not only related to the reason ofadmission such as trauma or surgery, but also to intensivecare related procedures flatteringly described as medical‘‘vampirism’’ (3). Compared to 1.1 times a day in patients ingeneral wards, patients in ICU are phlebotomized a mean of3.4 times a day for a mean volume of 41.5mL of blood (4).Patients who have arterial lines have more blood drawn andmore often than patients who do not have such lines. Phlebot-omy policy accounts for almost 50% of the variation in theamount of red blood cell transfusions on ICU (5). Based onthe experience that laboratories usually collect more bloodthan is needed for specific determinations, it has been calcu-lated that, for routine collections, a mean of 45 times therequired volume of specimen is obtained (6). This indicatesthat part of the anemia on ICU is iatrogenic and can bemanaged by a more restrictive policy in blood collections.

Gastrointestinal bleeding can be a source of blood loss andmay be caused by gastric tube lesions, ulcers, and stress-induced

660 van de Wiel

mucosa ulceration. In a study by Brown et al. (7), the uppergastrointestinal tractwas the site of bleeding in34.8%ofpatientswhose bleeds commenced on ICU, and accounted for 22% of allsites. Risk factors for bleeding after ICU admission includemechanical ventilation, nutritional factors, acute renal failure,and anticoagulants. In some of these situations prophylaxisagainst stress ulcers may be considered (8). Coagulation disor-ders may aggravate blood loss and include disorders in plateletfunction, decreased platelet count, (hyper) fibrinolysis, loss ofplasma coagulation factors, and endothelium-related coagula-tion disorders (9,10).

Erythropoietin Response

Under normal conditions, the body responds to a state ofanemia with an increased production of EPO. This 34 kDaglycoprotein is mainly produced in the peritubular cells ofthe cortex and outer medulla of the kidney. The primary sti-mulus for EPO production is hypoxia, as caused by a reducedred cell mass. Hemoglobin concentration has to fall below10.5 g=dL before a significant rise in EPO concentration canbe observed (11). The hormone EPO stimulates the productionof red cells in the bone marrow by acting on erythroid progeni-tor cells, the erythroid burst forming unit (BFU-E), and theerythroid colony-forming unit (CFU-E) (12,13).

Increasing data suggest that EPO concentrations areinappropriately low in critically ill patients. Krafte-Jacobs etal. (14) compared EPO concentrations in critically ill acutelyanemic, and acutely hypoxemic pediatric patients with thosein control groups including chronically anemic children. Sig-nificantly lower EPO concentrations were measured in the cri-tically anemic group with only one out of five children with alevel above normal. Such a blunted EPO response was alsofound by Rogiers et al. (15) in critically ill, nonhypoxemicadults who stayed more than 7 days in ICU and by Hobisch-Hagen et al. (16) inmultiply traumatized patients, who neededadmission on ICU. Inappropriately low endogenous EPO con-centrations were observed in prospective studies on the useof rhEPO in ICU patients (17,18). Various proinflammatory

Anemia in Intensive Care Patients 661

cytokines have been suggested and tested to be responsible forthis blunted EPO response. Inhibition of renal EPO geneexpression in rat kidneys has been demonstrated for interleu-kin 1 (IL-1) and tumor necrosis factor-alpha (TNF-a ) (19,20).

However, during the first days after sepsis or a septicshock, increased EPO levels have been found together withhigh levels of IL-6 (21,22). This high EPO level is irrespectiveof hemoglobin concentration and resembles that of acutephase proteins; its clinical significance remains unknown.

Because EPO is primarily produced in the kidney, onemight expect the lowest EPO levels in critically ill patientswith renal failure. However, all critically ill patients showextremely low values with only a relatively preserved EPOresponse in patients without sepsis and without renal failure(15). Acute renal failure, often seen in ICU, leads to equallylow EPO levels as in chronic renal failure (23). In thosepatients, EPO levels remain depressed even during therecovery phase.

Inflammation

The majority of patients on ICU will have elevated levels ofvarious inflammatory cytokines either because of their origi-nal disorder or in reaction to complications necessitating theirstay on this unit. Apart from blunting an EPO response by thekidneys, inflammatory cytokines are also able to inhibit redcell production by the bone marrow. This inhibition mayresult from a direct effect, such as described for IL-1 andTNF-a (24), or indirectly by the induction of changes in ironmetabolism (2).

Iron is essential for the number of hemoglobin moleculesproduced, and striking changes in iron metabolism areobserved in ICU patients with the characteristics of func-tional iron deficiency, hypoferremia in the presence ofadequate iron stores (18). The actions of cytokines on ironmetabolism are complex with some of them playing an inter-mediate role rather than causing a direct effect. IL-1b andTNF-a induce hypoferremia by increasing iron uptake into

662 van de Wiel

monocytes and macrophages in vitro and in mouse models(25,26). TNF-a also hampers iron release from the mononuc-lear phagocytic system (26,27). IL-1b is able to increase ferri-tin H-subunit production and not L-subunits, which aretranslationally controlled by iron-responsive protein, withenhanced intracellular entrapment of iron (28). The cytokineIL-6 is demonstrable within the first 48 hr after surgery andin critically ill patients depending on the duration of insult. Itincreases ferritin synthesis either by enhanced total proteinsynthesis or by redistribution of intracellular iron throughthe IRE=IRP (iron responsive elements=iron responsiveprotein) system (29). This additional ferretin may act as atrap for iron (30). IL-6 also induces direct uptake of iron inhepatocytes (31).

Apart from hypoferremia and high levels of ferritin, lowplasma levels of the iron transporter transferrin and a lowtransferrin saturation are observed in critically ill patients(18,32). Serum transferrin receptor levels are found to be lowto normal (18). Iron absorption from the gastrointestinal tractdecreases in cases of inflammation (33).

An impaired increase of erythropoiesis may also becaused by defective iron incorporation in the erythron.Although this has not been studied in ICU patients, a veryrapid iron utilization has been demonstrated in patients aftertotal hip surgery, a situation also characterized by temporaryhigh cytokines levels (34). It is therefore likely, as reflected byan adequate production of reticulocytes after treatment withrhEPO and iron (17,18), that the bone marrow of criticallyill patients can incorporate iron properly.

This leaves stimulation of iron uptake and impairment ofiron release from the macrophage system and hepatocytes asthe most plausible mechanism responsible for the impairmentof hemoglobin synthesis contributing to the anemia in ICUpatients.

Other Mechanisms

Apart from impairment of red cell production by functional irondeficiency and inappropriately low EPO levels, shortening of


red cell survival may contribute to the anemia of certain ICUpatients. In cardiac surgery, using cardiopulmonary bypassor in valve replacement, hemolysis influences erythrocyte’s lifespan (35,36). Some procedures especially those involving ische-mia–reperfusion injury and tissue damage induce oxidativestress. Oxygen radicals are able to enhance lipid peroxidation,which can affect red cell membrane integrity contributing tothe removal of erythrocytes from the circulation (37). Finally,the inflammatory state of surgery and sepsis may have a nega-tive influence on erythrocyte survival although this phenom-enon has only been reported in chronic inflammatoryconditions like rheumatoid arthritis (38,39). A study on ery-throcyte survival in total hip surgery showed the same half lifeof 51Cr-labeled cells before and after the operation (40). MostICU patients have normal vitamin B12 and folic acid concentra-tions, while their bone marrow shows normal cellularity with-out megaloblastic signs (18). This indicates that in mostcircumstances red cells of ICU patients will have a normal sur-vival unless there is ongoing contact with foreign material orcontinuation of inflammation.

THERAPY

Transfusion of red cells and supplementation with ironand=or vitamins are the usual therapies if the hemoglobinconcentration falls. However, there is still a debate aboutthe optimal hemoglobin concentration in critically ill patientsand maybe this optimum varies between patients and situa-tions. This may explain why many blood transfusions appearto be administered because of an arbitrary transfusionthreshold rather than a physiologic need for blood (5). A num-ber of factors such as age, cardiac performance, expectedblood loss, and sometimes even religion influence the decisionto transfuse a patient (41).

Nowadays a more restrictive transfusion policy can benoticed because of clinically relevant disadvantages relatedto transfusions. There is continuous concern about transmis-sion of infectious agents such as viruses and prions. Blood

664 van de Wiel

transfusions have effects on the immune system, which maypersist for a long period apart from more or less acute hemo-lytic and nonhemolytic reactions (42). The use of relativelylong stored blood may be associated with adverse effects incritically ill patients and fails to improve tissue oxygenation(43,44). Hebert et al. (45) have examined whether a restrictivestrategy of red cell transfusion would produce equivalentresults in critically ill patients compared to a liberal strategy.Such a restrictive policy proved to be equally effective if notsuperior, with the possible exception of patients with acutemyocardial infarction and unstable angina.

The introduction of rh-EPO has made it possible toachieve higher hemoglobin concentrations without exposureto the negative effects of transfused blood. Prospective trialshave shown that the bone marrow of critically ill patients isable to respond to the administration of exogenous EPO(17,18). Although this may result in reduction of the numberof units of red blood cells required, establishing the real impactof this therapy as well as of the optimal dosage needs furtherresearch (46).

SUMMARY

The anemia of critically ill patients shows a high degree ofanalogy with the anemia of chronic disease. Several mechan-isms may be involved, with a central role for inflammation.This induces functional iron deficiency, and blunts theresponsiveness and action of EPO. In some cases, a shorterlife span of erythrocytes may aggravate the anemia. Thereare strong indications supporting the role of the mononuclearphagocytic system in enhanced iron entrapment and a tem-porary inhibition of iron release while the bone marrowremains capable to incorporate iron and to react to exogenousEPO. The ICU patients respond to rh-EPO treatment withincreased erythropoiesis. Duration of anemia is related to con-tinued inflammation and suggests that this type of anemia isan acute variant of the anemia of chronic disease. A better


understanding of the transient character of the mechanismsinvolved as well as a more restrictive transfusion policymay contribute to a better and more appropriate treatmentof patients in those circumstances.

REFERENCES

1. Corwin HL, Krantz SB. Anemia of the critically ill:‘‘acute’’ anemia of chronic disease. Crit Care Med 2000; 28:3098–3099.

2. van Iperen CE, van de Wiel A, Marx JJM. Acute event-relatedanemia. Br J Haemat 2001; 115:739–743.

3. Burnum JF. Medical vampires. N Engl J Med 1986; 314:1250–1251.

4. Smoller BR, Kruskall MS. Phlebotomy for diagnostic labora-tory tests in adults. Pattern of use and effect on transfusionrequirements. N Engl J Med 1986; 314:1233–1235.

5. Corwin HL, Parsonnet KC, Gettinger A. RBC transfusion inthe ICU. Is there a reason? Chest 1995; 108:767–771.

6. Dale JC, Pruett SK. Phlebotomy—a minimalist approach.Mayo Clin Proc 1993; 68:249–255.

7. Brown RB, Klar J, Teres D, Lemeshow S, Sands M. Prospec-tive study of clinical bleeding in intensive care unit patients.Crit Care Med 1988; 16:1171–1176.

8. Cook DJ, Fuller HD, Guyatt GH, Marshall JC, Leasa D,Hall R, Winton TL, Rutledge F, Todd TJ, Roy P. Risk factorsfor gastrointestinal bleeding in critically ill patients. Canadiancritical care trials group. N Engl J Med 1994; 330:377–381.

9. Bone RC. Sepsis and coagulation. An important link. Chest1992; 102:594–595.

10. Boldt J. Coagulation abnormalities in the OR and on the ICU.Acta Anaesth Scand Suppl 1996; 109:75–76.

11. Kickler TS, Spivak JL. Effect of repeated whole blooddonations on serum immunoreactive eryhthropoietin levelsin autologous donors. JAMA 1988; 260:65–67.

666 van de Wiel

12. Spivak JL. The mechanism of action of erythropoietin. Int JCell Cloning 1986; 4:139–166.

13. Krantz SB. Erythropoietin. Blood 1991; 77:419–434.

14. Krafte-Jacobs B, Lvetown ML, Bray GL, Ruttemann UE,Pollack MM. Erythropoietin response to critical illness. CritCare Med 1994; 22:821–826.

15. Rogiers P, Zhang H, Leeman M, Nagler J, Neels H, Melot C,Vincent JL. Erythropoietin response is blunted in criticallyill patients. Intensive Care Med 1997; 23:159–162.

16. Hobisch-Hagen P, Wiedermann F, Mayr A, Fries D, JelkmannW, Fuchs D, Hasibeder W, Mutz N, Klingler A, SchobersbergerW. Blunted erythropoietic response to anemia in multiplytraumatized patients. Crit Care Med 2001; 29:743–747.

17. Corwin HL, Gettinger A, Rodriguez RM, Pearl RG, Gubler KD,Enny C, Colton T, Corwin MJ. Efficacy of recombinant humanerythropoietin in the critically ill patient: a randomized,double-blind, placebo-controlled trial. Crit Care Med 1999;27:2346–2350.

18. van Iperen CE, Gaillard CAJM, Kraaijenhagen RJ, BraamBG, Marx JJM, van de Wiel A. Response of erythropoiesisand iron metabolism to recombinant human erythropoietinin intensive care unit patients. Crit Care Med 2000;28:2773–2778.

19. Frede S, Fandrey J, Pagel H, Hellwig T, Jelkmann W. Ery-thropoietin gene expression is suppressed after lipopolysac-charide or interleukin-1 beta injections in rat. AM J Physiol1997; 273:1067–1071.

20. Jelkmann W. Proinflammatory cytokines lowering erythro-poietin production. J Interferon Cytokine Res 1998; 18:555–559.

21. Krafte-Jacobs B, Bock GH. Circulating erythropoietin andinterleukin-6 concentrations increase in critically ill childrenwith sepsis and septic shock. Crit Care Med 1996; 24:1455–1459.

22. Abel J, Spannbrucker N, Fandrey J, Jelkmann W. Serum ery-thropoietin levels with sepsis and septic shock. Eu J Haemat1996; 57:359–363.


23. Nielsen OJ, Thaysen JH. Erythropoietin deficiency in acutetubular necrosis. J Int Med 1990; 227:373–380.

24. Means RT, Krantz SB. Progress in understanding the patho-genesis of the anemia of chronic disease. Blood 1992; 80:1639–1647.

25. Gordeuk VR, Prithviraj P, Dolinar T, Brittenham GM. Inter-leukin 1 administration in mice produces hypoferremia despiteneutropenia. J Clin Invest 1988; 82:1934–1938.

26. Tsuji Y, Miller LL, Miller SC, Torti SV, Torti FM. Tumornecrosis factor-alpha and interleukin 1-alpha regulate trans-ferrin receptor in human diploid fibroblasts. J Biol Chem1991; 66:7257–7261.

27. Moldawer LL, Marano MA, Wei H, Fong Y, Silen ML, Kuo G,Manogue KR, Vlassura H, Cohen H, Cerami A. Cachec-tin=tumor necrosis factor-a alters red blood kinetics andinduces anemia in vivo. FASEB J 1989; 3:1637–1643.

28. Rogers JT, Bridges KB, Durmowicz G, Glass J, Auron PE,Munro HN. Translational control during the acute phaseresponse. J Biol Chem 1990; 265:14572–14578.

29. Rogers JT. Ferritin translation by interleukin-1 andinterleukin-6: the role of sequences upstream of the startcodons of the heavy and light subunit genes. Blood 1996;87:2525–2537.

30. Konijn AM, Hershko C. Ferretin synthesis in inflammation. I.Pathogenesis of impaired iron release. Br J Haemat 1977;37:7–16.

31. Kobune M, Kohgo Y, Kato J, Miyazaki E, Nitsu Y. Interleukin-6 enhances hepatic transferring uptake and ferritin expressionin rats. Hepatology 1994; 19:1468–1475.

32. von Ahsen N, Muller C, Serke S, Frei U, Eckhardt KU. Impor-tant role of nondiagnostic blood loss and blunted erythropoieticresponse in the anemia of medical intensive care patients. CritCare Med 1999; 27:2630–2639.

33. Weber J, Were JM, Julius HW, Marx JJM. Decreased ironabsorption in patients with active rheumatoid arthritis, withandwithout iron deficiency. Ann Rheumat Dis 1988; 47:404–409.

668 van de Wiel

34. van Iperen CE, van de Wiel A, van Leeuwen W, van der HorstFAL, Marx JJM. Postoperative iron utilisation and ferroki-netics. Neth J Med 2001; 58:A14.

35. Roeser WHP, Powell LW, O’Brien MF. Red cell survival afterheterograft valve surgery. Br Med J 1968; 28:806–808.

36. Tabak C, Eugene J, Stemmer EA. Erythrocyte survival follow-ing extracorporeal circulation. A question of membrane versusbubble oxygenator. J Thoracic Cardiovasc Surg 1981; 81:30–33.

37. Gutteridge JMC. Lipid peroxidation and antioxidants as bio-markers of tissue damage. Clin Chem 1995; 41:1819–1828.

38. Dinant HJ, de Maat CEM. Erythropoiesis and mean red-celllifespan in normal subjects and in patients with the anemiaof active rheumatoid arthritis. Br J Haematol 1978; 39:437–444.

39. Salvarani C, Casali B, Salvo D, Brunati C, Macchioni PL,Massai G, Lasagni D, Rivasi P, Portioli L. The role of interleu-kin-1, erythropoietin and red cell bound immunoglobulins inthe anaemia of rheumatoid arthritis. Clin Exp Rheumatol1991; 9:241–246.

40. van Iperen CE, van de Wiel A, de Bruin M, Marx JJM. Totalhip surgery does not influence red cell survival. Transfusion2000; 40:1235–1238.

41. Hebert PC, Wells G, Martin C, Tweeddale M, Marshall J,Blajchman M, Paglianello G, Schweitzer I, Calder L. A Cana-dian survey of transfusion practices in critically ill patients.Crit Care Med 1998; 26:482–487.

42. Mickler TA, Longnecker DE. The immunosuppressive aspectsof blood transfusion. J Intensive Care Med 1992; 7:176–188.

43. Marik PE, Sibbaid WJ. Effect of stored-blood transfusion onoxygen delivery in patients with sepsis. JAMA 1993; 269:3024–3029.

44. Fitzgerald RD, Martin CM, Dietz GE, Doig SG, Potter RF,Sibbald WJ. Transfusing red blood cells stored in citrate phos-phate dextrose adenine-1 for 28 days fail to improve tissue oxy-genation in rats. Crit Care Med 1997; 25:726–732.


45. Hebert PC, Wells G, Blichman MA, Marshall J, Martin C,Pagliarello G, Tweeddale M, Schweitzer I, Yetisir E. A multi-center, randomized, controlled clinical trial of transfusionrequirements in critical care. Transfusion requirements incritical care investigators, Canadian critical care trials group.N Engl J Med 1999; 340:409–417.

46. Darveau M, Notebaert E, Denault AY, Belisle S. Recombinanthuman erythropoietin use in intensive care. Ann Pharmac-other 2002; 36:1068–1074.

670 van de Wiel

25

Anemia in Renal Disease

JORGE LUIS AJURIA andPAUL L. KIMMEL

Department of Medicine, Division ofRenal Diseases and Hypertension,

George Washington University MedicalCenter, Washington, D.C., U.S.A.

ROBERT S. SIEGEL

Department of Medicine, Division ofHematology and Oncology, George

Washington University MedicalCenter, Washington, D.C., U.S.A.

INTRODUCTION

Anemia has long been recognized as a significant comorbidityin patients with chronic renal failure (CRF). With growingnumber of Americans developing chronic renal insufficiency(CRI) and progression to end-stage renal disease (ESRD),financial resources dedicated to treating the anemia of renalfailure with recombinant human erythropoietin (rHuEPO)have been further taxed. There is significant patient variabil-ity in rHuEPO dose response. Approximately 90% of patientstreated with rHuEPO respond with correction of anemia,whereas approximately 10% fail to attain target hemoglobinlevels despite the use of very high doses (1,2). A relative

671

deficiency of erythropoietin (EPO) is considered the majorcontributor to the anemia of CRF, however, there are a num-ber of other potentially modifiable conditions that contributeto anemia (Table 1) which can alter the therapeutic responseto rHuEPO therapy. Iron deficiency, chronic inflammatory

Table 1 Factors that May Contribute to Anemia of Renal Disease

Decreased erythropoietin synthesisDecreased production of EPO mRNAFewer EPO producing cells in chronic kidney failureEPO producing cells are refractory to stimulation in acute renal failureUremic toxins

Decreased stimulation of EPO mRNAChronic metabolic acidosisUremic toxins

Impaired erythropoiesisDecreased erythroid colony proliferationElevated PTHElevated cytokinesAcute and chronic inflammationUremic toxins

Decreased bone marrow response to EPOOsteitis fibrosaVitamin deficiencyIron deficiencyB12 deficiencyFolate deficiencyUremic toxins

Alterations in iron metabolism in chronic inflammatory statesIron sequestration

Shortened erythrocyte survival timeHemolysis associated with acute systemic illness with renal diseaseHemolysis associated with chronic renal failureDefect in erythrocyte hexose monophosphate shunt pathwayContaminants in dialysate

Splenic sequestrationErythrocyte membrane alterations in uremia

Abnormal hemostasisBlood lossGastrointestinalPhlebotomySurgery

672 Ajuria et al.

conditions, and under-dosing are common reasons for poorresponse to rHuEPO therapy. Blood loss, hemolysis, hyper-parathyroidism, vitamin B12 or folate deficiency, therapywith angiotensin converting enzyme (ACE) inhibitors, carni-tine deficiency, hemoglobinopathies, or antibodies directedagainst rHuEPO are other less common reasons for EPOresistance (3). This chapter will review the epidemiology ofrenal disease and anemia in the United States, the uniquepathophysiology of anemia associated with renal failure andthe clinical benefits of and trends in the use of rHuEPO sinceits introduction in 1989.

EPIDEMIOLOGY OF RENAL DISEASE IN THEUNITED STATES

Data from the Third National Health and Nutrition Examina-tion Survey (NHANES III), collected from 1988 through 1994,suggest approximately 5.6 millions persons in the UnitedStates, or 3% of the population, have a serum creatinine levelgreater than or equal to 1.6mg=dL in men and 1.4mg=dL inwomen (4). Seventy-seven percent of these individuals withan elevated serum creatinine were age 60 years or older.The prevalence of renal disease among different ethnic groupsvaries. Data from NHANES III suggest that CRI was presentin up to 19.3% of non-Hispanic Black Americans, 11.9% ofnon-Hispanic White Americans, and 8.1% of Mexican-Americans (4).

Each year more patients with CRI reach ESRD, requiringrenal replacement therapy (RRT). According to the UnitedStates Renal Data System (USRDS) 2001 Annual Report, in1999 there were over 340,000 Americans with ESRD. Approxi-mately 88,000 patients were initiated on RRT that year (5). Ofthe 340,000 patients, approximately 64% were treated withchronic hemodialysis (HD), 7% were treated with peritonealdialysis (PD), and the remaining 29% had a functioning renaltransplant. Over the last decade, higher prescribed doses ofHD, improved dialyzer design and membrane materials aswell as advances in the management of calcium=phosphate

Anemia in Renal Disease 673

imbalances and the introduction of rHuEPO have contributedto increased well being and survival among the ESRDpopulation (6–8).

Among the 340,000 Americans receiving RRT, approxi-mately 20% die yearly (5). Most ESRD patients have multiplecomorbidities that impact on mortality, frequency of hospita-lizations, and overall perceptions of quality of life (9,10).Known predictors of morbidity and mortality in patients withESRD include age, gender, race, primary diagnosis of dia-betes, low serum albumin concentration, and elevated levelsof circulating proinflammatory cytokines and C-reactiveprotein (CRP) (11–16). The leading cause of death in patientswith ESRD treated with chronic HD is cardiovasculardisease, followed by infectious complications (5,17).

ANEMIA AND CHRONIC RENAL DISEASE

The anemia of chronic renal disease is characterized asnormochromic and normocytic. Relative to the degree of ane-mia, the reticulocyte count and circulating EPO levels arelow. Erythroid precursors in the bone marrow are hypoplasticwith normal appearance and retain capacity for normalmaturation. Leukopoiesis and megakaryocytopoiesis are typi-cally not altered in chronic renal disease (18,19).

With progression of renal disease and a decrease in glo-merular filtration rate (GFR), the prevalence and severity ofanemia increases (20). In a recent analysis of the NHANESIII database involving approximately 15,000 adults, the pre-valence of serum hemoglobin less than 12 g=dL was 1.8% inpersons with a calculated GFR 90 mL=min per 1.73m2 orhigher, 5.2% in persons with a GFR between 30 and 59mL=min per 1.73m2, and 44.1% in persons with a GFRbetween 15 and 29mL=min per 1.73m2. Lower GFR was asso-ciated with a lower serum hemoglobin level (20). Men tend tohave a lower hematocrit than women at the same creatinineclearance (21). Before the introduction of rHuEPO approxi-mately one-quarter of patients requiring chronic HD requiredintermittent or regular red-cell transfusions for symptomatic

674 Ajuria et al.

anemia (22). Androgen therapy was commonly used in thispopulation for anemia correction with limited success(23–25).

PATHOPHYSIOLOGY

The anemia of CRF can be attributed to several mechanisms(26): (1) deficiency of EPO production relative to the degree ofanemia (27–30); (2) retention of metabolites that impair ery-thropoiesis (31–35); (3) bone marrow failure (36),(4) blood loss(37); and (5) shortened erythrocyte survival and hemolysis(38,39).

ERYTHROPOIETIN AND THE KIDNEY

The association of anemia with renal disease was described byRobert Bright in the early 19th century. The observation thatpersons living at high altitudes had elevated red blood cell(RBC) counts led to the understanding that hypoxia contribu-ted to RBC production. Carnot and Deflandre at the Sorbonnea century ago first demonstrated that sera drawn from ane-mic rabbits and injected into normal rabbits stimulatedRBC production (40). This ‘‘hemopoietin’’ factor was lateridentified as erythropoietin by Erslev in the early 1950s usingsimilar experiments with sera from rabbits with severe ane-mia (41). Erythropoietin stimulates the generation of erythro-cytes by regulating the differentiation of hematopoieticprogenitor cells in the bone marrow. Jacobson et al. (42) in1957 identified the kidney as the primary site of EPO produc-tion by demonstrating that nephrectomized rats did notincrease hemoglobin levels in response to hypoxic stimuli.Miyake et al. (43) purified EPO from the urine of patientswith aplastic anemia. This development led to the eventualisolation and cloning of the human EPO gene on chromosomeseven by Lin et al. and Jacobs et al. (44,45).

Erythropoietin is a sialylglycoprotein produced by theliver in the fetus (46), then by the kidney in late gestation.EPO has a molecular weight of approximately 34kDa and is


composed of 166 amino acid residues with two disulfide bonds(47). The carbohydrate moiety is rich in sialic acid, and isneeded to maintain EPO reactivity in vivo (48). Under normalconditions, serum EPO concentration ranges between 8 and18 mU=mL, and serum levels can increase a thousand-foldin persons with normally functioning kidneys (49). Undernonhypoxic conditions, approximately 90% of EPO is pro-duced by the kidneys in adults, with the remaining 10% byliver hepatocytes or Ito cells located within the space of Disse(50–52). During hypoxic stimulation, hepatic EPO mRNA canaccount for 25–50% of the total EPO mRNA synthesized (53).Erythropoietin gene expression, though not of the samemagnitude, has been demonstrated in macrophages (54) andtissue other than kidney or liver (55,56). The estimatedhalf-life of EPO in the circulation is approximately 5.2 hr innormal subjects (57).

In situ hybridization experiments localized the site ofEPO production within the kidney to fibroblast-like cells inthe peri-tubular interstitium of the cortex and outer medulla(58–62). Levels of urine and plasma EPO increase as thehematocrit is lowered by phlebotomy and decrease aftertransfusion of RBCs (63). Erythropoietin production is regu-lated both by the kidney and liver through control of EPOmRNA transcription (64). There is a rapid increase in the rateof transcription of EPO mRNA by the kidney and liver inresponse to a decrease in tissue oxygen tension either becauseof anemia or under hypoxic conditions (53,64–68).

Goldberg et al. (69) proposed that the oxygen-sensingmechanism mediating the hypoxic signal-transduction path-way for EPO mRNA synthesis was dependent upon whetherthe heme protein conformation was in the oxygenated or thedeoxygenated state. The oxygenated conformation of theheme protein does not stimulate EPO production, while thedeoxygenated protein conformation stimulates EPO mRNAtranscription (70). There has been a variety of data to supportthe Goldberg model. The divalent cations cobalt and nickelcan substitute for ferrous iron in the heme protein, and stimu-late EPO mRNA production (69,71,72). Binding of carbonmonoxide to the heme protein will lock the heme protein in

676 Ajuria et al.

the oxygenated conformation, reducing hypoxia-induced EPOproduction (69).

Interference with mitochondrial oxidative phosphoryla-tion was another mechanism proposed for the oxygen-regu-lated control of EPO mRNA transcription. However, hypoxiccompromise of mitochondrial oxidative phosphorylation inrat kidneys perfused with cyanide, antimycin, and oligomycindid not stimulate EPO mRNA production (73). Molecular stu-dies reported in 1991 identified a hypoxia-inducible enhancerregion located 30 to the EPO gene that activated tissue-specificEPO gene transcription in response to hypoxia (74,75). Wangand Semenza (56) identified a hypoxia-inducible factor 1(HIF-1) that bound to this 30 enhancer region and stimulatedEPO mRNA transcription. Molecular studies by Pugh et al.used cobalt and cyanide to measure stimulation of EPOmRNAproduction via this enhancer region. Cobalt stimulated EPOmRNA production, while cyanide did not (75). The inabilityof cyanide to stimulate EPO mRNA production disprovedthe theory that oxygen-sensing regulation of EPO mRNAproduction was coupled to oxidative metabolism.

ERYTHROPOIETIN PRODUCTION IN RENALDISEASE

Erythropoietin serum concentrations in persons with renaldisease tend to be normal or elevated (27,28,76). Patients withrenal disease, however, have a relative deficiency of serumEPO in response to the degree of anemia present (77).Decreased synthesis of EPO by a kidney with advanced renalfailure is attributed to scarring and loss of the renal cellsinvolved in oxygen sensing and EPO production (26). Cur-iously, patients with polycystic kidney disease often do notdevelop anemia to the same degree as other patients withrenal disease despite progression to ESRD (77). The differ-ence in severity of anemia is attributed to autonomous pro-duction of EPO by interstitial cells associated with proximaltubular cysts (78). Autonomous production of EPO by renalcell carcinomas complicated by polycythemia has been


reported in persons with or without renal disease, includingESRD patients with acquired cystic kidney disease treatedwith HD (79–81).

The metabolic acidosis associated with renal disease mayinhibit hypoxia-induced EPO production. Metabolic acidosisin patients with CRF results from decrease in net acid excre-tion and bicarbonate loss by the kidney (82,83). Serum pHand acidosis have been shown to modify hypoxia-inducedEPO production (84–87). In a series of experiments in mice,Eckardt et al. (88) were able to demonstrate a 30–40% reduc-tion in the rate of EPO production in response to hypoxia dur-ing metabolic and respiratory acidosis, but not duringalkalosis (86). The chronic metabolic acidosis of renal failuremay decrease erythrocyte intracellular 2,3-DPG levels andinfluence heme protein oxygen affinity (89). Oxygen affinityof the heme protein is influenced by pH, pCO2 and intracellu-lar 2,3-diphosphoglycerate (2,3-DPG) levels (89–92). At lowpH, high pCO2, and elevated intracellular 2,3-DPG levels,the heme protein has decreased oxygen affinity. Astrup (89)proposed that sustained chronic acidosis could influence gly-colytic enzyme activity within erythrocytes and ultimatelylower erythrocyte intracellular 2,3-DPG concentrations. Lowerythrocyte intracellular 2,3-DPG levels in blood frompatients with chronic metabolic acidosis and renal failurewould favor stabilization of the oxygenated heme protein.More oxygenated hemoglobin present in the circulation couldpotentially stimulate less EPO mRNA production.

ERYTHROPOIETIN PRODUCTION IN ACUTERENAL FAILURE

Anemia is present in most patients with acute renal failure(ARF) (93–95). During hospitalization, blood loss from surgi-cal procedures, phlebotomy from frequent blood draws, conco-mitant infection and gastrointestinal bleeding are common(95). Critically ill patients in the intensive care unit with orwithout renal failure are at increased risk for developmentof anemia because circulating EPO levels mimic circulating

678 Ajuria et al.

levels of acute-phase reactants (96). Elliot et al. were able todemonstrate a significant initial elevation of circulatingEPO levels in critically ill patients with or without ARF dur-ing the first 48hr of admission in an intensive care unit. Cir-culating serum EPO levels dropped to normal or belownormal along with serum hemoglobin levels in all criticallyill patients by the third day of hospitalization. Some of thepatients in the study had some degree of anemia at the timeof admission (96).

Serum EPO levels can be suppressed during ARF in non-critically ill patients (97). Studies by Maxwell et al. (49) inacute renal injury induced in mice demonstrated an alteredresponse of EPO mRNA expression by EPO synthesizingrenal interstitial fibroblasts. The fibroblasts became refrac-tory to stimulation by anemia or hyperbaric hypoxia afterinjury. Serum EPO levels return to baseline levels after recov-ery of renal function and restoration of the GFR (94).

UREMIA AND ERYTHROPOIESIS

The term uremia was first coined in 1847 by Piorry to describethe contamination of blood by urine in patients with renalfailure (98). In uremia, waste products usually excreted bythe kidneys accumulate in blood and tissue, resulting in dele-terious metabolic abnormalities affecting nearly all bodysystems. Clinical manifestations of the uremic syndromeinclude neurologic, cardiovascular, pulmonary, gastrointest-inal, dermatologic, hematologic, endocrine, and ophthalmolo-gic abnormalities (99). Abnormal metabolism of traceminerals and a deficiency in membrane ion transport systemscontribute to the clinical sequelae of the uremic syndrome.Although many compounds have been implicated in uremictoxicity, the identity of specific uremic toxins and the patho-physiology of the uremic syndrome remain uncertain. Manyof the toxins that accumulate in uremia have been associatedwith the anemia of renal failure (35).

Dietary protein restriction is known to decrease theseverity of uremic symptoms (100). Urea is the principal


end-product of protein catabolism and nitrogen metabolism.Despite the association of a protein-restricted diet withdecreased severity of uremic symptoms, urea has not consis-tently been shown to be a mediator of uremic toxicity. Resultswere equivocal in two studies where chronic HD patientswere dialyzed using a urea-supplemented dialysate andsymptoms were assessed (101,102). Nitrogen, another productof protein catabolism, can be incorporated into organic basesknown as guanidine compounds that accumulate in tissueand sera of uremic patients (103–105). Guanidine compoundsare strong organic bases that have been implicated aspotential mediators of the uremic syndrome and inhibitorsof erythropoiesis (35,99).

ERYTHROPOIESIS IN UREMIA

A decreased bone marrow response to circulating EPO in per-sons with CRF contributes to anemia. Erythrokinetic studiesby Erslev and Besarab (76) compared erythrocyte productionand destruction in stable ESRD patients treated with HD andin normal, nonanemic individuals without renal disease. Therate of erythrocyte production in the ESRD patients wasapproximately half the rate of erythrocyte production in thenormal individuals, irrespective of circulating EPO levels,which were equal or higher in the ESRD patients (76).

Increases noted in the hematocrit of some ESRD patientsafter initiation of HD suggest theremay be clearance of uremictoxins that could potentially impair bone marrow erythropoi-esis (31,106,107). In vitro studies have demonstrated impairedcell growth of both erythroid burst-forming units and erythroidcolony-forming units in bone marrow cultures incubated withuremic sera (32–34). Potential uremic toxins investigatedincluded the polyamines spermine, spermidine, and putricine(108–112). Arsenic, ribonuclease, vitamin A, polar lipids, andparathyroid hormone (PTH) were also investigated as possibleuremic inhibitors (113–117). Recent studies have shown thatthe inhibitory effects of these potential uremic toxins can beovercome with rHuEPO therapy (115,118–121).

680 Ajuria et al.

ERYTHROPOIESIS AND INFLAMMATION

Iron metabolism is altered in patients with CRF (122). Acuteand chronic inflammation influences iron metabolism anderythropoiesis. End-stage renal disease patients treated withHD have higher levels of circulating proinflammatory cyto-kines when compared to normal controls (12,123–125). Wewere able to show that higher levels of circulating proinflam-matory cytokines were associated with increased mortality inESRD patients treated with HD, independent of medical riskfactors (12).

Elevated inflammatory cytokine levels contribute to theanemia of chronic inflammatory disorders by decreasing theamount of iron available for hemoglobin synthesis andimpairing the erythropoietic response to EPO. Whereas ironabsorption by intestinal mucosa is increased with inflamma-tion (126), the iron absorbed is sequestered and unavailablefor erythropoiesis. Inflammation-mediated alterations in ironstorage and availability include decreased iron transport(127), increased content and synthesis of ferritin withinmononuclear phagocytic cells (128), and impaired iron releasefrom macrophage storage (129,130). The inflammatory cyto-kines tumor necrosis factor-alpha (TNF-a), interleukin-1(IL-1), interleukin-6 (IL-6), and interferon gamma (IFNg)can inhibit erythropoiesis and contribute to alterations in ironmetabolism (96,127,130–134). Various proinflammatorycytokines have been shown to impair proliferation of erythro-cyte progenitor cells (133,135), suppress EPO production(136,137), and blunt the bone marrow response to EPO(134,138).

Proinflammatory cytokines in the sera of uremic patientstreated with chronic HD can blunt the effect of rHuEPO onerythropoiesis and contribute to EPO resistance. In a seriesof in vitro experiments to evaluate rHuEPO response ininflammatory conditions, Allen et al. (139) used sera fromnonuremic controls and uremic sera from HD patients withor without concurrent infection. The nonuremic sera andthe uremic sera were added to normal human bone marrowcultures. After rHuEPO was added, significantly fewer


erythroid colony-forming units (CFU-E) formed in the cul-tures incubated in uremic sera compared to cultures incu-bated in sera from normal controls. Fewer CFU-E formed inthe uremic sera drawn from patients with infection, than inuremic sera drawn from patients without infection. Additionof anti-IFNg and anti-TNF-a antibodies to the culturesreversed uremic inhibition of CFU-E formation. These resultssuggest that a concomitant acute inflammatory process in auremic patient may suppress erythropoiesis significantlymore than uremia alone. C-reactive protein is an acute-phaseprotein generated by the liver during acute inflammation.Elevated serum levels of CRP during acute inflammatoryevents have been correlated with increased rHuEPOrequirements in ESRD patients treated with HD (1,140).

The urea reduction ratio (URR) is one marker used tomeasure adequacy of HD. A URR less than 65% has beenassociated with decreased hematocrit and increased rHuEPOdosing (141). In a prospective study of 135 randomly selectedESRD patients, Ifudu et al.(141) compared URR and serumhemoglobin concentration. Patients were studied for 4 weeks.Patients in the study had ESRD treated with HD for at least 3months, were greater than 20 years of age, and had receivedrHuEPO for at least 3 months. There was a direct correlationbetween URR and hematocrit, such that the higher the URR,the more likely patients were to have a hematocrit greaterthan 30%. More studies, however, are needed to clarify theassociation of HD adequacy and anemia.

RENAL FAILURE AND MYELOFIBROSIS

Most patients with renal disease have secondary hyperpar-athyroidism. The etiology of hyperparathyroidism in renaldisease is multifactorial. 1,25-Dihydroxyvitamin D3 is criticalfor suppression of PTH secretion (142,143). Renal secretion of1,25-dihydroxyvitamin D3 decreases with progression of renaldisease. The decline in serum 1,25-dihydroxyvitamin D3 con-centration in patients with CRI correlates inversely with arise in serum PTH levels (144). Hyperphosphatemia has a

682 Ajuria et al.

direct stimulatory role on the parathyroid gland independentof 1,25-dihydroxyvitamin D3 concentration. Serum phosphateconcentration directly correlates with serum PTH levels inpatients with mild to moderate CRF (145). Osteitis fibrosa isa type of renal osteodystrophy associated with severe hyper-parathyroidism, characterized by fibrosis of the bone marrow(146). The anemia of hyperparathyroidism is attributed todecreased bone marrow volume and responsiveness as aresult of fibrosis (115,147–149). Independent of myelofibrosis,in vitro studies have demonstrated direct suppression of ery-throid colony proliferation by PTH (115,150). ESRD patientstreated with rHuEPO often require higher dosing to achievegoal hematocrit depending on the severity of hyperparathyr-oidism and marrow fibrosis (36,151). ESRD patients treatedwith HD who fail medical therapy with vitamin D havedemonstrated a significant increase in hemoglobin, increasein reticulocyte count, and decrease in rHuEPO requirements3 months after subtotal parathyroidectomy (146,149,152).

SHORTENED ERYTHROCYTE LIFE SPAN

In most patients with renal disease, erythrocyte life span isshorter than in people without renal disease. Hemolysis ispresent in a variety of conditions associated with renal failureand may contribute to shortening of erythrocyte survivaltime. Causes of ARF that can present with hemolysis include:Wilson’s disease (153); thrombotic thrombocytopenic purpura(TTP); hemolytic uremic syndrome (HUS); microangiopathichemolytic anemia from vasculitis or malignant hypertension;infectious endocarditis; transfusion reactions; malarial orclostridial infections; genetic disorders such as glucose-6-phosphate deficiency (G6PD) or paroxysmal nocturnal hemo-globinuria; and venomous bites from some spiders or snakes(154).

During HD, municipal water is purified through anelaborate and regulated water treatment system before it iscombined with concentrated dialysate and circulated throughthe dialyzer where it is in direct contact with blood (155).


Thermal injury during HD, or contamination of dialysatewater with copper, nitrate or chloramines may cause varyingdegrees of hemolysis in ESRD patients treated with HD(156–159). Rare reports of zinc (160) or aluminum (161) toxi-city from contaminated dialysate water were associated withacute drops in hemoglobin concentrations in ESRD patientstreated with HD.

Eschbach et al. used 14C cyanate to tag erythrocytes fromuremic and nonuremic patients and quantified erythrocytelife span. The erythrocyte survival time in the nonuremicpatients averaged approximately 115 days, while erythrocytesurvival time in the uremic patients averaged approximately73 days (39). Shortened erythrocyte survival time may beattributed to retention of an unknown solute in uremia, nor-mally cleared by the kidney, that alters erythrocyte biology.Joske et al.(162) demonstrated, almost 50 years ago, that ery-throcytes from uremic donors transfused into normal recipi-ents had normal erythrocyte survival time. Modifications toerythrocytes in uremia that may contribute to hemolysisand premature erythrocyte destruction include: alteration inthe erythrocyte hexose monophosphate shunt (HMPS) path-way; lipid peroxidation of the erythrocyte membrane inducedby oxidant stress; and impaired erythrocyte deformabilitywith subsequent splenic sequestration.

The HMPS in erythrocytes generates NADPH. NADPHfunctions to protect RBC from oxidative degeneration andpremature destruction by the generation of reduced glu-tathione. Reduced glutathione is needed to detoxify hydrogenglutathione through a reaction catalyzed by glutathioneperoxidase. In vitro studies on NADPH production by uremicerythrocytes exposed to oxidant stress demonstrate areduction in NADPH production by 50% or more. In theseexperiments, reduction in erythrocyte NADPH productioncorrelated with decreased erythrocyte survival time (163).Patients with G6PD have an inherited defect in the HMPSpathway resulting in erythrocyte hemolysis when exposed tooxidant compounds such as primaquine or sulfonamides. Ure-mic patients have evidence of a defect in the erythrocyteHMPS comparable to persons with G6PD. Jacob et al.

684 Ajuria et al.

screened for HMPS activity in 200 chronically hemodialyzeduremic patients using the ascorbate-cyanide test. Approxi-mately three quarters of the uremic patients had test resultsin the same range or higher than patients with G6PD (164).Incubation of erythrocytes from normal individuals in uremicsera reproduces defects in the HMPS pathway.

The metabolic derangement in the HMPS pathway ofuremic erythrocytes may involve defective recycling of glucosewith intracellular accumulation of glycolytic intermediates(163,165). Transketolase is an enzyme required for the meta-bolism of glucose by the HMPS pathway. Guanidino com-pounds such as methylguanidine, guanidinopropionic acid,guanidinoacetic acid, or gamma-guanadinobutyric acid accu-mulate in the sera and tissue of uremic patients and caninhibit erythrocyte transketolase enzyme activity (105,166).Hemodialysis lowers circulating levels of guanidino com-pounds and significantly increases erythrocyte transketolaseactivity (105). In vitro studies with methylguanidine, guanidi-nopropionic acid, guanidinoacetic acid, or gamma-guanadino-butyric acid have been shown to cause autohemolysis of RBC,as well as in vitro inhibition of erythrocyte glucose-6-phosphate dehydrogenase (99,104,167,168).

Patients with acute or chronic renal failure occasionallydevelop iatrogenic hypophosphatemia during continuous RRTor as a result of aggressive use of phosphate binders. Hypopho-sphatemia depletes erythrocytes of ATP necessary to maintainbiconcavity and viability in the circulation. Increased rigidityof erythrocytes depleted of ATP may result in hemolysis andsplenic removal from the circulation (169–171).

MEMBRANE ALTERATIONS IN UREMICERYTHROCYTES

Erythrocytes from uremic patients have various membraneand intracellular alterations. Echinocytes (Burr cells) arecommonly seen in the peripheral blood smears of patientswith elevated serum blood urea nitrogen concentration. Thetransformation to echinocytes is attributed to changes in the


erythrocyte membrane resulting from the presence of uremictoxins in the sera. The presence of echinocytes is not asso-ciated with severity of anemia (172).

Membrane alterations of uremic erythrocytes may influ-ence survival time. The aminophospholipid phosphatidylser-ine (PS) is normally located on the inner leaflet of thephospholipid bilayer of the RBC membrane. Phosphatidylser-ine expression on the outer surface of erythrocytes maypromote adherence to vascular endothelium, macrophagerecognition, and splenic clearance of aged or abnormal ery-throcytes (173–175). There is a significant increase in theamount of PS expressed on the outer leaflet of erythrocytemembranes from patients with renal disease when comparedto erythrocytes from normal controls (176). In vitro studieshave demonstrated a significant correlation between PS-expressing erythrocytes and erythrophagocytosis by macro-phages in uremic patients (177). Alterations in erythrocytemembrane deformability and fluidity are present in erythro-cytes from uremic patients and may be involved in splenicsequestration (178,179). Prior to the introduction of rHuEPOtherapy, splenectomy in ESRD patients with severe anemiawas occasionally used with some benefit in patientswho developed hypersplenism as a consequence of multipletransfusions (180).

Markers of oxidant stress are higher in patients withchronic renal disease when compared to normal controls(181,182). Oxidant stress has been implicated in acceleratedatherogenesis of patients with chronic renal disease and otherclinical manifestations of uremia (183). Himmelfarbet al.(183) proposed that patients with chronic renal diseasehave an increased oxidative burden, and with progression ofrenal disease, there is increased oxidation of proteins, lipids,and carbohydrates. Erythrocytes are free radical scavengersand provide important antioxidant defense (184). Carrelet al.(185) proposed that the accumulation of free oxygen radi-cals may contribute to hemolysis by direct oxidation of ery-throcyte membranes. Results of studies designed to comparethe degree of erythrocyte membrane oxidation in ESRDpatients treated with HD and normal controls have been

686 Ajuria et al.

equivocal (186–188). Malonyldialdehyde (MDA) is a shortchain aldehyde produced by the oxidation of erythrocytemembrane polyunsaturated fatty acids. While the erythrocyteMDA membrane content from ESRD patients treated withHD has been shown to be elevated when compared to controlsin some studies (186), other investigators have failed to showsignificant differences (187).

Leukocytes in ESRD patients undergoing HD come intodirect contact with the HD membrane. Leukocytes maybecome activated during HD and produce oxidative radicalspecies (189,190). Vitamin E and glutathione have antioxidantproperties (191). ESRD patients treated with oral vitamin Esupplementation and intravenous glutathione infusion duringHD had reduced presence of oxidant markers and reducedseverity of anemia (192,193). Dialysis with vitamin E-bondeddialyzer membranes has been shown to significantly increaseerythrocyte survival (193). Oxidant stress can be measured bythe rate of superoxide release from polymorphonuclear leuko-cytes. Release of superoxide from polymorphonuclear leuko-cytes obtained from ESRD patients treated with PD is dec-reased during rHuEPO therapy (194). Correction of anemiawith rHuEPO in patients with ESRD treated with HD hasnot been shown to correct erythrocyte HMPS activity, mem-brane MDA levels, erythrocyte deformability, or evidence ofsplenic sequestration (195).

Changes in ion transport systems are present in erythro-cyte membranes of nondialyzed uremic patients. Alterationsin cell membrane electrical potential gradients can result inchanges of intracellular ion concentrations. In erythrocytesfrom uremic patients, the intracellular sodium and calciumconcentration are higher than in erythrocytes from nonure-mic patients or stable adequately dialyzed patients withESRD (196). In erythrocytes from uremic patients, theamount and activity of magnesium-dependent Na=K ATPasetransporters on the cell membrane are reduced (197,198).These changes can be reproduced in erythrocytes from normalsubjects incubated in the sera of uremic patients, and are cor-rected after adequate dialysis (199). Similar changes arefound in erythrocyte membrane NaK=2Cl cotransporters


(200) and erythrocyte membrane Ca ATPase enzyme activity(201). The higher intracellular concentrations of sodium andcalcium in erythrocytes of uremic patients may be attributedto alterations in these ion transport mechanisms. Alterationsin erythrocyte intracellular calcium and sodium concentra-tions have not been shown to shorten erythrocyte survivaltime.

ABNORMAL HEMOSTASIS

Patients with acute and chronic renal failure have a platelet-mediated hemorrhagic tendency that contributes to theanemia of renal failure. The hemostatic abnormality in renalfailure is characterized by a prolonged bleeding time withnormal coagulation studies. Ecchymosis, gastrointestinalblood loss, epistaxis, and bleeding complications from surgicalprocedures are common manifestations of uremic coagulopa-thy (202–204). The hemostatic abnormality in renal failureis attributed to qualitative defects in platelet function. Theplatelet count is typically in the normal range unless the etiol-ogy of the renal failure is a microangiopathic hemolyticanemia, a hypotensive-ischemic injury associated with gramnegative septicemia, or disseminated intravascular coagulo-pathy. A significantly increased incidence of mild thrombocy-topenia (less than 150 � 109=L) has recently been described inESRD patients with hepatitis C infection treated with HD orPD (205).

Qualitative platelet abnormalities in uremia are multi-factorial and involve alterations in platelet function, as wellas platelet–platelet and platelet–endothelial interactions.The platelet granule count as well as intragranule ADP andserotonin stores are reduced in uremia (206). A functionaldefect in platelet granule secretion may be present in uremia(207). It is attributed to a decrease in platelet ATP release inresponse to stimuli. Alterations in platelet–platelet interac-tions include impaired platelet aggregation attributed todecreased production of thromboxane B2, a potent platelet-aggregating agent (208,209). Increased production of

688 Ajuria et al.

prostacyclin (PGI2), another inhibitor of platelet aggregation,may also contribute to uremic bleeding (207,210,211).

Defects in uremic platelet–endothelial interactions mayinvolve nitric oxide (NO). NO is a vasoactive compound thathas been shown to inhibit platelet function. Excessive forma-tion of platelet-derived and endothelium-derived NO inuremia has been associated with impaired platelet adhesionand aggregation (212). Administration of n-monomethyl-l-arginine (l-NMMA), a specific inhibitor of NO formation, nor-malized bleeding time when given to uremic rats (212,213).Platelet adhesion to the endothelium is also reduced in ure-mia (214,215). Defects in platelet–endothelial interactionsmay be attributed to functional abnormalities of the plateletglycoprotein (GP) IIb–IIIa receptor (216) and factor VIII:vonWillebrand factor (VIII=vWF) complex (217–219).

Adequate dialysis may sometimes partly correct qualita-tive platelet abnormalities and decrease bleeding risk inuremic patients (220,221). Acute uremic bleeding can be cor-rected with deamino-8-d-arginine vasopressin (DDAVP)administered intravenously, subcutaneously (217) or intrana-sally. DDAVP releases FVIII:vWF complex from storage sitesinto plasma (222). Tachyphylaxis to effects of DDAVP occursafter a few doses secondary to depletion of FVIII:vWF stores.Conjugated estrogens can produce sustained correction ofuremic bleeding time with multiple administrations. Patientsdo not develop tachyphylaxis to the effects of estrogen(223,224). A single intravenous dose of conjugated estrogenis effective in shorting uremic bleeding time, attaining amaximum effect at 24hr, and ending at 72hr. Four or fiveconsecutive administrations of conjugated estrogen shortenthe uremic bleeding time for up to 25 days in some uremicpatients (223). Infusion of cryoprecipitate can transientlyshorten uremic bleeding times. The bleeding time may correctbetween 1 and 12 hr after infusion of cryoprecipitate, andreturn to the pretransfusion bleeding time 24–48 hr later(225).

Prolonged bleeding time is influenced by serum hemoglo-bin concentration. Prolonged bleeding times may be morecommon in patients with profound anemia. Uremic patients


transfused with washed packed RBCs had shortening of thebleeding time (226). The mechanism underlying the hemo-static effect of anemia correction is unknown. Therapy withRHuEPO with a subsequent increase in the hematocrit toapproximately 30% effectively normalized the bleeding timein uremic patients (227,228) and was shown to improve plate-let aggregation (229,230).

TRENDS IN rHuEPO THERAPY

Despite Food and Drug Administration (FDA) approval ofrHuEPO use in the ESRD population in 1989, and Medicarecost coverage, implementation of rHuEPO therapy increasedonly 139% after its introduction in 1989 through 1996, withonly a modest increase in the hematocrit level of the popula-tion (231). Most recent data from the Health Care FinancialAdministration (HCFA) show that the mean hematocrit atinitiation of RRT increased from 28.1% in 1995 to 29.3% in1999 (232). Recent studies evaluating anemia in ESRDpatients initiated on chronic RRT from 1995 through 2000demonstrate that 50% have serum hematocrit values lessthan 28%, and only about 20–28.1% received rHuEPO ther-apy prior to initiation of RRT (232–234).

The 2002 report of the ESRD Clinical Performance Mea-sures (CPM) Project sampled 8863 adults with ESRD treatedwith HD from October through December 2001. The averagehemoglobin concentration in these ESRD patients treatedwith HD was 11.7 g=dL. Eight percent of the ESRD patientstreated with HD had a mean hemoglobin concentration lessthan 10 g=dL (235). A higher prevalence of African-Ameri-cans, patients with ESRD treated with HD less than 6months, and patients 18–44 years of age had a hemoglobinconcentration less than 10 g=dL (235). The CPM Project alsosampled 1451 adults with ESRD treated with PD from Octo-ber 2001 through March 2002. Eight percent of ESRDpatients treated with PD had a hemoglobin concentration lessthan 10 g=dL. This is an improvement from the 1996Peritoneal Dialysis-Core Indicators Study (PD-CIS) that

690 Ajuria et al.

showed 30% of ESRD patients treated with PD had a hemato-crit value less than 30% (236).

For those patients treated with RRT, analysis of HCFAclaims submitted by outpatient providers showed that thenumber of patients receiving rHuEPO increased by 25% from1995 through 1998, as did the dose of rHuEPO per patient.Despite the increase in rHuEPO doses, the hematocrit inpatients receiving rHuEPO rose only by approximately4.1–8.0% during this interval, at an increased cost to Medi-care of 20%. The trend in increased use and dosing ofrHuEPO during this period may be due to changes in HCFAreimbursement regulations, financial incentives for outpati-ent providers, and the presence of rHuEPO resistance dueto iron deficiency or chronic inflammation (122,237).

With a growing number of people requiring RRT forESRD, the cost of rHuEPO therapy has come under consider-able scrutiny. Medicare spent approximately $868 millionon rHuEPO therapy for all patients with ESRD receivingdialysis in 1998 (237). A randomized, controlled trial of subcu-taneous vs. intravenous administration of rHuEPO at theDepartment of Veteran Affairs Medical Centers in ESRDpatients demonstrated that the hematocrit could be main-tained at target range at a lower average weekly dose whenrHuEPO was administered subcutaneously rather than intra-venously (238,239). The subcutaneous rHuEPO dose wasapproximately one-third that of the intravenous EPO doseto reach the same target hematocrit range. It is estimatedthat Medicare could save $47–142 million annually if 25–75%of patients receiving intravenous rHuEPO were switched tosubcutaneous rHuEPO administration (238,239).

MANAGEMENT OF ANEMIA

Patients with CRF and anemia should be treated withrHuEPO (240,241). There are no data to show that therapywith rHuEPO in patients with CRF will worsen or arrestthe progression of renal disease (240,242,243). Weeklysubcutaneous dosing of rHuEPO with oral iron supplementa-


tion is usually sufficient (242). Functional iron deficiency hasbeen observed in patients at doses of 15units per kg ofrHuEPO or higher (22).

According to USRDS data, approximately half of ESRDpatients treated with RRT are iron deficient (5). The presenceof iron deficiency in patients with renal disease is a strong pre-dictor of the presence of anemia (20), and patients with CRIand ESRD would benefit from more aggressive iron replace-ment therapy (244–246). ESRD patients with normal ironstores at initiation of rHuEPO therapy may still develop irondeficiency as soon as the eighth week of rHuEPO therapy(247). Intravenous iron supplementation has been shown todecrease rHuEPO dosing in ESRD patients treated withrHuEPO (248). HCFA data on outpatient HD demonstrateunderutilization of intravenous iron repletion in anemicpatients with a transferrin saturation less than or equal to20% (249). Given recent reports of increased infectious compli-cations in patients treated with intravenous iron supplemen-tation and the history of adverse allergic reactions in earlypreparations, use of intravenous iron supplementation amongESRD patients with iron deficiency remains less than optimal.

Clinical guidelines for the management of anemia inESRD patients are provided by the National Kidney Founda-tion=Dialysis Outcome Quality Initiative (NKF=DOQI) (250).Clinical practice guidelines were originally published in1997 and recently revised. Recommended target hemoglobinfor rHuEPO therapy in ESRD patients is 11–12g=dL (250).In patients on HD, subcutaneous rHuEPO can maintain thehematocrit at a lower average weekly dose (average 32%lower) thanwith intravenous administration (239). It has beensuggested that iron overload and hemoglobin correction abovethe recommended target range may contribute to increasedcardiovascular events and increased risk of infection (251).

The optimal hemoglobin target range remains controver-sial. In a randomized, prospective, open-label trial, Besarabet al.(252) studied 1233 patients with ischemic or congestiveheart failure and ESRD treated with HD. Risk and benefitsof a low target hematocrit of 30% were compared to a normaltarget hematocrit of 42%. After 29 months, causes of death

692 Ajuria et al.

between the two groups were the same, and difference inevent-free survival did not reach statistical significance.There were 183 deaths and 19 first event nonfatal myocardialinfarctions (MI) in the normal hematocrit group, compared to150 deaths and 14 nonfatal MIs in the low hematocrit group.The study was halted early despite no significant differencebetween groups. The higher hematocrit group required moreiron supplementation and had a decline in adequacy of dialy-sis as assessed by Kt=V. Other studies have associated higherhematocrit with reduced clearance of urea, creatinine, andphosphate during HD, as well as increased heparin require-ments (79,253–255).

Side effects of rHuEPO therapy are common. Phase IIImulticenter trials of rHuEPO were conducted in 1986 with333 chronic HD patients (256). The target hematocrit chosenwas 35%. During the trial 45% of the patients developed irondeficiency, with a 40% drop in iron levels within 6 weeks ofstarting rHuEPO therapy. Blood pressure increased in 35%of patients, 5.4% had a seizure, and 5% complained of myal-gia. Elevations in blood pressure during rHuEPO therapyare common and typically only seen in patients with renal dis-ease (22,240,256–258). Increased erythrocyte mass has notbeen associated with increase in blood pressure (259). Theincrease in blood pressure associated with rHuEPO therapyin patients with CRF has been attributed, at least in part,to increased blood viscosity. An increase in hematocrit willincrease viscosity and correct the hyperdynamic cardiovascu-lar state. Compensatory systemic vasodilatation associatedwith the hyperdynamic circulatory state typically resolvesover time and peripheral vascular resistance increases(260,261). Blood pressure elevations can be managed withantihypertensive agents and increased volume ultrafiltrationduring HD. The increased viscosity associated with rHuEPOtherapy has not been shown to increase the probability ofvascular access thrombosis (256,262).

Patients receiving rHuEPO who develop resistance totherapy should be evaluated for iron or folate deficiency,occult infections, and causes of chronic blood loss, such asgastrointestinal bleeding (263) (Table 1). Patients undergoing


surgical procedures may become transiently refractory torHuEPO therapy (242). Hyperparathyroidism with myelofi-brosis, aluminum-related bone disease, or bone marrowmalignancy may also cause bone marrow failure and reducethe response to rHuEPO (264). Inadequate HD may increasethe risk of bleeding, worsen the patient’s nutritional status,and further blunt the response to rHuEPO. There is no evi-dence to support the need for dietary folate supplementationin adequately nourished HD patients receiving rHuEPO ther-apy (265). Angiotensin converting enzyme inhibitors (ACEI)are commonly used in patients with CRI or ESRD and havebeen associated with exacerbation of anemia in persons withrenal disease (266,267). While the exact mechanisms underly-ing the effect are unclear, it has been suggested that ACEImay suppress proliferation of erythroid precursors. ACEItherapy in ESRD patients treated with HD, however, doesnot significantly reduce rHuEPO response nor result in thedevelopment of severe anemia (268). The development of anti-erythropoietin antibodies can occur in ESRD patients treatedwith HD receiving rHuEPO therapy (269). Casadevall et al.recently reported 13 ESRD patients treated with HD whodeveloped neutralizing antibodies against the protein moietyof rHuEPO. All patients were treated with subcutaneousinjections of either epoetin alfa or epoetin beta. After aninitial response to rHuEPO therapy, patients who developanti-rHuEPO antibodies may become severely anemic andtransfusion dependent. Cessation of rHuEPO therapy resultsin a slow decline in antibody titers.

CLINICAL SEQUELAE OF ANEMIA IN CHRONICRENAL DISEASE

Untreated anemia in patients with CRI and ESRD is asso-ciated with various cardiovascular complications includingincreased left ventricular hypertrophy (270–273), increasedincidence of congestive heart failure (270,274), angina (275),increased cardiac output (276), and lower exercise tolerance(277). Retrospective analysis of the HCFA ESRD registry

694 Ajuria et al.

and claims showed that ESRD patients with hematocrits lessthan 30% had a 14–30% increased risk of hospitalization and12–33% increased risk for all-cause mortality (278,279).Given the incidence of diabetes and hypertension in the Uni-ted States ESRD population, it is not surprising that the mostcommon cause of death in ESRD patients treated with HD iscardiac disease, followed by infectious complications (5). Thepresence of CRF has been shown to be an independent riskfactor for cardiovascular death (280).

ESRD patients with anemia are more likely to developdecompensated heart failure, most commonly from eitherdilated or hypertrophic cardiomyopathy (281). In a prospectivecohort study of 432 ESRD patients receiving either HD or PD,Foley et al.(270) were able to demonstrate that a 1 g=dLdecrease in hemoglobin was independently associated withthe presence of left ventricular dilation, and the developmentor recurrence of heart failure (270). Left ventricular hypertro-phy is an independent determinant of survival in ESRDpatients, and present in the majority of patients beginningchronic HD in the United States (282–284). Low serum hemo-globin concentration is an independent predictor of left ventri-cular hypertrophy and evident in patients with even mild tomoderate chronic kidney disease (272). The cardiovascularsystem in chronic anemia is characterized by a hyperdynamiccirculatory state associated with increased cardiac output,stroke volume, preload, and contractility (260,276,285). Myo-cardial oxygen supply is further compromised in patients withcoronary artery disease (CAD).

CLINICAL BENEFITS OF rHuEPO THERAPY

Significant clinical benefits are noted in patients with CRF andESRD treated with rHuEPO. In ESRD patients treated withHD and rHuEPO, there is an inverse relationship betweenhematocrit and length of hospital stay and number of hospita-lizations (278,286–288). ESRD patients with higher hemato-crits have better survival (279,287,288). Cardiovascularconsequences of rHuEPO therapy and correction of anemia


include decreased cardiac output and decreased contractilitywith no significant change of preload and afterload (289,290).Correction of anemia has also been shown to reduce exercise-induced electrocardiographic ST segment depression (291),increase symptom-limited treadmill exercise performance(275), improve exercise capacity (292), and prevent leftventricular dilation in patients without symptomatic cardiacdisease (271).

Correction of anemia with rHuEPO has been shown toimprove appetite, nutritional status (255), and health-relatedquality of life. Controlled studies in patients with chronic renaldisease before RRT demonstrate a significant improvement inassessments of energy, physical function, social activity, and cog-nitive function with rHuEPO therapy (293). Similar benefitshave been observed in ESRD patients treated with chronic HD(294–296). The Cooperative Multicenter EPO Clinical TrialGroup evaluated objective and subjective quality of lifemeasuresin 300ESRDpatients treatedwith chronicHDafter introductionof rHuEPO. Patients reported significant improvement inemployment status, functional ability and overall health status,as well as life satisfaction, sense of well being, and happiness(295). Some HD patients treated with rHuEPO had improvedsexual function, return of menses, and partial correction ofalterations in the growth hormone secretory axis (297–300).

Introduction of rHuEPO significantly decreased therequirement for blood transfusions in ESRD patients. Fewertransfusions decrease the risk of iron overload, infection,and development of cytotoxic antibodies that could precludekidney transplantation (246). rHuEPO therapy with correc-tion of anemia in patients with ESRD has been shown tocorrect cellular phagocytic function as well as the ability tomount an antibody response after vaccination (301,302). Arecent analysis of the Reduction of End Points in NIDDMwiththe Angiotensin II Receptor Antagonist Losartan (RENAAL)Study which included 1513 patients with type II diabetesand nephropathy, identified anemia as an independent riskfactor for progression to ESRD (303). Further studies to clarifythe association between anemia and progression of chronickidney disease are needed.

696 Ajuria et al.

CONCLUSION

Anemia in patients with CRI or ESRD is common. While themajor cause of the anemia associated with renal disease is arelative deficiency of EPO production, there are multiple fac-tors associated with uremia that impair erythropoiesis andshorten red-cell survival (Table 1). Introduction of rHuEPOas a therapy for the anemia of renal disease revolutionizedthe management of patients with CRF and ESRD. Improve-ments in cardiac function, increased energy, and fewer hospi-talizations because of the therapeutic use of EPO havedramatically improved the overall quality of life for patientsliving with kidney disease.

REFERENCES

1. Macdougall IC, Cooper AC. Erythropoietin resistance: the roleof inflammation and pro-inflammatory cytokines. NephrolDial Transplant 2002; 17(suppl 11):39–43.

2. Eschbach JW, Varma A, Stivelman JC. Is it time for a para-digm shift? Is erythropoietin deficiency still the main causeof renal anaemia? Nephrol Dial Transplant 2002; 17(suppl 5):2–7.

3. Remuzzi G, Minetti L. Hematologic consequences of renalfailure. In: Brenner B, ed. The Kidney, 7 th ed. Philadelphia:W.B.Saunders, 2000:2079–2102.

4. Coresh J, Wei GL, McQuillan G, Brancati F, Levey A,Jones C, Klag M. Prevalence of high blood pressure and ele-vated serum creatinine level in the United States: findingsfrom the Third National Health and Nutrition ExaminationSurvey (1988–1994). Arch Intern Med 2001; 161:1207–1216.

5. U.S. Renal Data System, USRDS 2001 Annual Data Report:Atlas of End-Stage Renal Disease in the United States.Bethesda, MD: National Institutes of Health, National Insti-tute of Diabetes and Digestive and Kidney Diseases, 2001.

6. Parker TF, Husni L, Huang W, Lew N, Lowrie EG, DallasNephrology Associates. Survival of hemodialysis patients in


the United States is improved with a greater quantity ofdialysis. Am J Kidney Dis 1994; 23:670–680.

7. Port FK, Orzol SM, Held PJ, Wolfe RA. Trends in treatmentand survival for hemodialysis patients in the United States.Am J Kidney Dis 1998; 32(suppl 4):S34–S38.

8. Ritz E, Schomig M, Bommer J. Osteodystrophy in the millen-nium. Kidney Int 1999(suppl 73):S94–S98.

9. Keane WF, Collins AJ. Influence of comorbidity on mortalityand morbidity in patients treated with hemodialysis. Am JKidney Dis 1994; 24:1010–1018.

10. Evans RW, Manninen DL, Garrison LP, Hart LG, Blagg CR,Gutman RA, Hull AR, Lowrie EG. The quality of life ofpatients with end-stage renal disease. N Engl J Med 1985;312:553–559.

11. Port FK. Morbidity and mortality in dialysis patients. KidneyInt 1994; 46:1728–1737.

12. Kimmel PL, Phillips TM, Simmens SJ, Peterson RA,Weihs KL, Alleyne S, Cruz I, Yanovski JA, Veis JH. Immuno-logic function and survival in hemodialysis patients. KidneyInt 1998; 54:236–244.

13. Kimmel PL, Varela MP, Peterson RA, Weihs KL, Simmens SJ,Alleyne S, Amarashinge A, Mishkin GJ, Cruz I, Veis JH. Inter-dialytic weight gain and survival in hemodialysis patients:effects of duration of ESRD and diabetes mellitus. Kidney Int2000; 57:1141–1151.

14. Ikizler TA, Wingard RL, Harvell J, Shyr Y, Hakim RM. Asso-ciation of mortality with markers of nutrition and inflamma-tion in chronic hemodialysis patients: a prospective study.Kidney Int 1999; 55:1945–1951.

15. Zimmermann J, Herrlinger S, Pruy A, Metzger T, Wanner C.Inflammation enhances cardiovascular risk and mortality inhemodialysis patients. Kidney Int 1999; 55:648–658.

16. Yeun JY, Levine RA, Mantadilok V, Kaysen GA. C-reactiveprotein predicts all-cause and cardiovascular mortalityin hemodialysis patients. Am J Kidney Dis 2000; 35:469–476.

698 Ajuria et al.

17. Bloembergen WE, Stannard DC, Port FK, Wolfe RA, Pugh JA,Jones CA, Greer JW, Golper TA, Held PJ. Relationship of doseof hemodialysis and cause-specific mortality. Kidney Int 1996;50:557–565.

18. Dainiak N. Hematologic complications of renal disease. In:Hoffman R, Benz EJ, Shattil SJ, Furie B, Cohen HJ, Silber-stein LE, McGlave P, eds. Hematology: Basic Principles andPractice. Philadelphia, PA: Churchill Livingstone, 2000:2357–2373.

19. Kaye M. The anemia associated with renal disease. J LabClin Med 1957; 52:83–95.

20. Astor BC, Muntner P, Levin A, Eustace JA, Coresh J. Asso-ciation of kidney function with anemia. The Third NationalHealth and Nutritional Examination Survey (1988–1994).Arch Intern Med 2002; 162:1401–1408.

21. Hsu C-Y, Bates DW, Kuperman GJ, Curhan GC. Relationshipbetween hematocrit and renal function in men and women.Kidney Int 2001; 59:725–731.

22. Eschbach JW, Egrie JC, Downing MR, Browne JK, AdamsonJW. Correction of the anemia of end-stage renal disease withrecombinant human erythropoietin. Results of combinedphase I & II clinical trials. N Engl J Med 1987; 316:73–78.

23. Shahidi NT. Androgens and erythropoiesis. N Engl J Med1973; 289:72–80.

24. Eschbach JW, Adamson JW. Improvement in the anemia ofchronic renal failure with fluoxymesterone. Ann Intern Med1973; 78:527–532.

25. Neff MS, Goldberg J, Slifkin RF, Eiser AR, Calamia V,Kaplan M, Baez A, Gupta S, Mattoo N. A comparison ofandrogens for anemia in patients on hemodialysis. N Engl JMed 1981; 304:871–875.

26. Eschbach JW. The anemia of chronic renal failure: pathophy-siology and the effects of recombinant erythropoietin. KidneyInt 1989; 35:134–148.

27. Caro J, Brown S, Miller O, Murray T, Erslev AJ. Erythropoie-tin levels in uremic nephric and anephric patients. J Lab ClinMed 1979; 93:449–458.


28. Radke HW, Claussner A, Erbes PM, Scheuermann EH,Schoeppe W, Koch KM. Serum erythropoietin concentrationin chronic renal failure: relationship to degree of anemiaand excretory renal function. Blood 1979; 54:877–884.

29. Erslev AJ, Besarab A. Erythropoietin in the pathogenesis andtreatment of the anemia of chronic renal failure. Kidney Int1997; 51:622–630.

30. Erslev AJ, Caro J, Miller O, Silver R. Plasma erythropoietinin health and disease. Am Clin Lab Sci 1980; 10:250–257.

31. MacGonigle RJS, Husserl F, Wallin JD, Fisher JW. Hemodia-lysis and continuous ambulatory peritoneal dialysis effects onerythropoiesis in renal failure. Kidney Int 1984; 25:430–436.

32. Wallner SF, Kurnick JE, Ward HP, Vautrin R, Alfrey AC.The anemia of chronic renal failure and chronic diseases: invitro studies of erythropoiesis. Blood 1976; 47:561–569.

33. Ohno Y, Rege AB, Fisher JW, Barona J. Inhibitors oferythroid colony forming cells (CFU-E and BFU-E) in seraof azotemic patients with anemia of renal disease. J Lab ClinMed 1978; 92:916–923.

34. Wallner SF, Kurnick JE, Vautrin R, Ward HP. The effect ofserum from uremic patients on erythropoietin. Am J Hematol1977; 3:45–55.

35. MacDougall IC. Role of uremic toxins in exacerbating anemiain renal failure. Kidney Int 2001; 59(suppl 78):S67–S72.

36. Rao DS, Shih M-S, Mohini R. Effect of serum parathyroidhormone and bone marrow fibrosis on the response to ery-thropoietin in uremia. N Engl J Med 1993; 328:171–175.

37. Loge JP, Lange RD, Moore CV. Characterization of theanemia associated with chronic renal insufficiency. Am JMed 1958; 24:4–18.

38. Eklund SG, Johansson SV, Strandberg O. Anemia in uremia.Hemolysis evaluated by common methods and application ofthe bilirubin turnover method. Acta Med Scand 1971;190:435–443.

700 Ajuria et al.

39. Eschbach JW, Korn D, Finch CA. 14C cyanate as a tag for redcell survival in normal uremic man. J Lab Clin Med 1977; 89:823–828.

40. Erslev AJ. Erythropoietin: from physiology to clinical trialsvia molecular biology. In: Erslev AJ, Adamson JW, EschbachJW, Winearls CG, eds. Erythropoietin. Molecular, Cellularand Clinical Biology. Baltimore: Johns Hopkins UniversityPress, 1991:3–18.

41. Erslev AJ. Humoral regulation of red cell production. Blood1953; 8:349–387.

42. Jacobson LO, Goldwasser E, Fried W, Plzak L. Role of thekidney in erythropoiesis. Nature 1957; 179:633–634.

43. Miyake T, Kung CKH, Goldwasser E. Purification of humanerythropoietin. J Biol Chem 1977; 252:5558–5564.

44. Lin FK, Suggs S, Lin CH, Browne JK, Smalling R, Egrie JC,Chen KK, Fox GM, Martin F, Stabinsky Z, Badrawi SM,Lai PH, Goldwasser E. Cloning and expression of the humanerythropoietin gene. Proc Natl Acad Sci USA 1985; 82:7580–7584.

45. Jacobs K, Shoemaker C, Rudersdorf R, Neill SD,Kaufman RJ, Mufson A, Seehra J, Jones SS, Hewick R,Fritsch EF, Kawakita M, Shimizu T, Miyake T. Isolationand characterization of genomic and cDNA clones of humanerythropoietin. Nature 1985; 313:806–810.

46. Zanjani ED, Poster J, Burlington H, Mann LI, Wasserman LR.Liver as the primary site of erythropoietin formation in thefetus. J Lab Clin Med 1977; 89:640–644.

47. Lai P-H, Everett R, Wang F-F, Arakawa T, Goldwasser E.Structural characterization of human erythropoietin. J BiolChem 1986; 261:3116–3121.

48. Spivak JL, Hogans BB. The in vivo metabolism of recombi-nant human erythropoietin in the rat. Blood 1989; 73:90–99.

49. Maxwell PH, Ferguson DJP, Nicholls LG, Iredale JP, PughCW, Johnson MH, Ratcliffe PJ. Sites of erythropoietinproduction. Kidney Int 1997; 51:393–401.


50. Erslev AJ. Renal biogenesis of erythropoietin. Am J Med1975; 58:25–30.

51. Fried W. The liver as a source of extrarenal erythropoieticproduction. Blood 1972; 40:671–677.

52. Maxwell PH, Ferguson DJP, Osmond MK, Pugh CW, HeryetA, Doe BG, Johnson MH, Radcliffe PJ. Expression of a homo-logously recombined erythropoietin-SV40 T antigen fusiongene in mouse liver: evidence for erythropoietin productionby Ito cells. Blood 1994; 84:1823–1830.

53. Tan CC, Eckardt K-U, Ratcliffe PJ. Organ distribution oferythropoietin messenger RNA in normal and uremic rats.Kidney Int 1991; 40:69–76.

54. Rich IN, Heit W, Kubanec B. Extrarenal erythropoietinproduction by macrophages. Blood 1982; 60:1007–1018.

55. Maxwell PH, Pugh CW, Ratcliffe PJ. Inducible operation ofthe erythropoietin 30 enhancer in multiple cell lines: evidencefor a widespread oxygen sensing mechanism. Proc Natl AcadSci USA 1993; 90:2423–2427.

56. Wang GL, Semenza GL. General involvement of hypoxia-inducible factor 1 in transcriptional response to hypoxia. ProcNatl Acad Sci USA 1993; 90:4304–4308.

57. Eckardt K-U, Boutellier U, Kurtz A, Schopen M, Koller EA,Bauer C. Rate of erythropoietin formation in humans inresponse to acute hyperbaric hypoxia. J Appl Physiol 1989;66:1785–1788.

58. Lacombe C, Da Silva J-L, Bruneval P, Fournier JG, WendlingF, Casadevall N, Camilleri JP, Bariety J, Varet B, TambourinP. Peritubular cells are the site of erythropoietin synthesis inthe murine hypoxic kidney. J Clin Invest 1988; 81:620–623.

59. Koury ST, Bondurant MC, Koury MJ. Localization of erythro-poietin synthesizing cells in murine kidneys by in situ hybri-dization. Blood 1988; 71:524–527.

60. Bachmann S, Le Hir M, Eckardt K-U. Co-localization of ery-thropoietin mRNA and ecto-50-nucleotidase immunoreactivityin peritubular cells of rat renal cortex indicates that fibro-blasts produce erythropoietin. J Histochem Cytochem 1993;41:335–341.

702 Ajuria et al.

61. Koury ST, Bondurant MC, Koury MJ. Localization of erythro-poietin synthesizing cells in murine kidneys by in situ hybri-dization. Blood 1988; 71:524–527.

62. Maxwell PH, Osmond MK, Pugh CW, Heryet A, Nicholls LG,Tan CC, Doe BG, Ferguson DJP, Johnson MH, Ratcliffe PJ.Identification of the renal erythropoietin-producing cellsusing transgenic mice. Kidney Int 1993; 44:1149–1162.

63. Adamson JW. The erythropoietin-hematocrit relationshipin normal and polycythemic man: implications of marrowregulation. Blood 1968; 32:597–609.

64. Bondurant MC, Koury MJ. Anemia induces accumulation oferythropoietin mRNA in the kidney and liver. Mol Cell Biol1986; 6:2731–2733.


66. Kochling J, Curtin PT, Madan A. Regulation of human ery-thropoietin gene induction by upstream flanking sequencesin transgenic mice. Br J Haematol 1998; 103:960–968.

67. Beru N, McDonald J, Lacombe C, Goldwasser E. Expressionof the erythropoietin gene. Mol Cell Biol 1986; 6:2571–2575.

68. Goldberg MA, Gaut CC, Bunn HF. Erythropoietin mRNAlevels are governed by both the rate of gene transcriptionand posttranscriptional events. Blood 1991; 77:271–277.

69. Goldberg MA, Dunning SP, Bunn HF. Regulation of the ery-thropoietin gene: evidence that the oxygen sensor is a hemeprotein. Science 1988; 242:1412–1415.

70. Huang LE, Ho V, Arany Z, Krainc D, Galson D, Tendler D,Livingston DM, Bunn HF. Erythropoietin gene regulationdepends on heme-dependent oxygen sensing and assemblyof interacting transcription factors. Kidney Int 1997; 51:548–552.

71. Goldberg MA, Glass GA, Cunningham JM, Bunn HF. Theregulated expression of erythropoietin by two human hepa-toma cell lines. Proc Natl Acad Sci USA 1987; 84:7972–7976.


72. Jacobson LO, Fried W, Plazak LF. Studies on erythropoiesis,V: the effect of cobalt on the production of erythropoietin.Blood 1958; 13:55–60.

73. Tan CC, Ratcliffe PJ. Effect of inhibitors of oxidative phos-phorylation on erythropoietin mRNA in isolated perfusedrat kidneys. Am J Physiol 1991; 261:F982–F987.

74. Semenza GL, Nejfelt MK, Chi SM, Antonarakis SE. Hypoxia-inducible nuclear factors bind to an enhancer element located30 to the human erythropoietin gene. Proc Natl Acad Sci USA1991; 88:5680–5684.

75. Pugh CW, Tan CC, Jones RW, Ratcliffe PJ. Functional analy-sis of an oxygen-regulated transcriptional enhancer lying 30

to the mouse erythropoietin gene. Proc Natl Acad Sci USA1991; 88:10553–10557.

76. Erslev AJ, Besarab A. The rate and control of baseline red cellproduction in hematologically stable patients with uremia. JLab Clin Med 1995; 126:283–286.

77. Chandra M, Miller ME, Garcia JF, Mossey RT, McVicar M.Serum immunoreactive erythropoietin levels in patients withpolycystic kidney disease as compared with other hemodialy-sis patients. Nephron 1985; 39:26–29.

78. Eckardt K-U, Mollmann M, Neumann R, Brunkhorst R,Burger HU, Lonnemann G, Scholz H, Keusch G, BuchholzB, Frei U, Bauer C, Kurtz A. Erythropoietin in polycystic kid-neys. J Clin Invest 1989; 84:1160–1166.

79. Sakamoto S, Igarashi T, Osumi N, Imamoto T, Tobe T,Kamiya M, Ito H. Erythropoietin-producing renal cell carci-noma in chronic hemodialysis patients: a report of two cases.Int J Urol 2003; 10:49–51.

80. Hanada T, Mimata H, Ohno H, Nasu N, Nakagawa M,Nomura Y. Erythropoietin-producing renal cell carcinomaarising from acquired cystic disease of the kidney. Int J Urol1998; 5:493–494.

81. Wiesener MS, Seyfarth M, Warnecke C, Jurgensen JS,Rosenberger C, Morgan NV, Maher ER, Frei U, Eckardt KU.Paraplastic erythrocytosis associated with an inactivating

704 Ajuria et al.

point mutation of the von Hippel–Lindau gene in a renal cellcarcinoma. Blood 2002; 99:3562–3565.

82. Perazella MA, Rastegar A. Disorders of potassium and acid-base metabolism in association with renal disease. In:Schrier, ed. RWDiseases of the Kidney and Urinary Tract.7th ed. Philadelphia: Lippincott Williams and Wilkins,2001:2577–2605.

83. Warnock DG. Uremic acidosis. Kidney Int 1988; 34:278–287.

84. Cohen RA, Miller ME, Garcia JF, Moccio G, Cronkite EP.Regulatory mechanism of erythropoietin production: effectsof hypoxemia and hypercarbia. Exp Hematol 1981; 9:513–521.

85. Eckardt K-U, Kurtz A, Bauer C. Regulation of erythropoietinformation is related to proximal tubular function. Am J Phy-siol 1989; 256:F942–F947.

86. Miller ME, Rorth M, Parving HH, Howard D, Reddington I,Valeri CCR, Stohlman F. pH effect on erythropoietinresponse to hypoxia. N Engl J Med 1973; 288:706–710.

87. Faura J, Gurney CW, Fried W. The effect of carbon dioxide onerythropoiesis. Ann N Y Acad Sci 1968; 149:456–461.

88. Eckardt K-U, Kurtz A, Bauer C. Triggering of erythropoietinformation by hypoxia is inhibited by respiratory and meta-bolic acidosis. Am J Physiol 1990; 258:R678–R683.

89. Astrup P. Red-cell pH and oxygen affinity of hemoglobin. NEngl J Med 1970; 283:202–203.

90. Astrup P, Engel K, Severinghaus JW, Munson E. Theinfluence of temperature and pH on the dissociation curveof oxyhemoglobin of human blood. Scand J Clin Lab Invest1965; 17:515–523.

91. Astrup P, Rorth M, Thorshauge C. Dependency on acid-basestatus of oxyhemoglobin dissociation and 2,3-diphosphoglyce-rate level in human erythrocytes. II. In vivo studies. Scand JClin Lab Invest 1970; 26:47–52.

92. Tomita S, Riggs A. Studies of the interaction of 2,3-diphospho-glycerate and carbon dioxide with hemoglobins from mouse,man, and elephant. J Biol Chem 1971; 246:547–554.


93. Hales M, Solez K, Kjellstrand C. The anemia of acute renalfailure: association with oliguria and elevated blood urea.Ren Fail 1994; 16:125–131.

94. Nielsen OJ, Thaysen JH. Erythropoietin deficiency in acutetubular necrosis. J Intern Med 1990; 227:373–380.

95. Anderson RJ, Schrier RW. Acute renal failure. In: SchrierRW, ed. Diseases of the Kidney and Urinary Tract. 7th ed.Philadelphia: Lippincott Williams and Wilkins, 2001:1093–1136.

96. Elliot JM, Virankabutra T, Jones S, Tanudsintum S,Lipkin G, Todd S, Bion J. Erythropoietin mimics the acutephase response in critical illness. Crit Care 2003; 7:R35–R40.

97. Kokot F, Wiecek A, Grzeszczak W. Plasma parathyroidhormone and erythropoietin levels in patients with nonin-flammatory acute renal failure. Int Urol Nephrol 1993;25:89–96.

98. Richet G. Early history of uremia. Kidney Int 1988; 33:1013–1015.

99. Baily JL, Mitch WE. Pathophysiology of uremia. In: BrennerBM, ed. The Kidney. 6th ed. Pennsylvania: W.B.Saunders,2000:2059–2078.

100. Cotton JR, Knochel JP. Correction of uremic cellular injurywith a protein-restricted, amino acid-supplemented diet. AmJ Kidney Dis 1985; 5:233–236.

101. Merrill JP, Legrain M, Hoigne R. Observations on the role ofurea in uremia. Am J Med 1953; 14:519.

102. Johnson WJ, Hagge WH, Wagoner RD, Dinapoli RP,Rosevear JW. Effects of urea loading in patients withfar-advanced renal failure. Mayo Clin Proc 1972; 47:21–29.

103. Kopple JD, Gordon SI, Wang M, Swendseid ME. Factorsaffecting serum and urinary guanidinosuccinic acid levelsin normal and uremic subjects. J Lab Clin Med 1977; 90:303–311.

104. Orita Y, Ando A, Tsubakihara Y, Mikami H, Kikuchi T,Nakata K, Abe H. Tissue and blood cell concentration

706 Ajuria et al.

of methylguanidine in rats and patients with chronic renalfailure. Nephron 1981; 27:35–39.

105. Pietrzak I, Baczyk K. Erythrocyte transketolase activity andguanidino compounds in hemodialysis patients. Kidney Int2001; 59(suppl 78):S97–S101.

106. Wallner SF, Vautrin RM. Evidence that inhibition of erythro-poiesis is important in the anemia of chronic renal failure. JLab Clin Med 1981; 97:170–178.

107. Radtke HW, Radke HW, Frei U, Erbes PM, Schoeppe W,Koch KM. Improving anemia by hemodialysis. Effect onserum erythropoietin. Kidney Int 1980; 17:382–387.

108. Rege AB, LaMarche MB, Dagmar B, Bartos F, Campbell RA,Fisher JW. Identification of spermine as an inhibitor of ery-thropoiesis in patients with chronic renal failure. J ClinInvest 1981; 67:1623–1629.

109. Kushner D, Beckman B, Nguyen L, Chen S, Della Santina C,Husserl F, Rice J, Fisher JW. Polyamines in the anemia ofend-stage renal disease. Kidney Int 1991; 39:725–732.

110. Saito A, Takagi T, Chung TG, Ohta K. Serum levels of poly-amines in patients with chronic renal failure. Kidney Int1983; 24(suppl 16):S234–S237.

111. Segal GM, Stueve T, Adamson JW. Spermine and spermidineare non-specific inhibitors of in vitro hematopoiesis. KidneyInt 1987; 31:72–76.

112. Cambell RA. Anemia, uremia, and polyamines. Nephron1985; 41:299–301.

113. Pershagen G, Hast R, Lins LE, Pehrsson K. Increased arsenicconcentration in the bone marrow in chronic renal failure: acontribution to anemia? Nephron 1982; 30:250–252.

114. Ono K, Waki Y, Takeda K. Hypervitaminosis A: a contribut-ing factor to anemia in regular dialysis patients. Nephron1984; 38:44–47.

115. Meytes D, Bogin E, Ma A, Dukes PP, Massry SG. Effect ofparathyroid hormone on erythropoiesis. J Clin Invest 1981;67:1263–1269.


116. Wallner SF, Vautrin RM. The anemia of chronic renal failure:studies of the affect of organic solvent extraction of theserum. J Lab Clin Med 1978; 92:363–369.

117. Freedman MH, Saunders EF, Cattran DF, Rabin EZ. Ribonu-clease inhibition of erythropoiesis in anemia of uremia. Am JKidney Dis 1983; 2:530–533.

118. Eschbach JW, Haley NR, Egrie JC, Adamson JW. A compar-ison of the responses to recombinant human erythropoietin innormal and uremic subjects. Kidney Int 1992; 42:407–416.

119. Eschbach JW, Miadenovic J, Garcia JF, Wahl PW, AndersonJW. The anemia of chronic renal failure in sheep: theresponse to erythropoietin-rich plasma in vivo. J Clin Invest1984; 74:434–441.

120. Delwiche F, Garrity MJ, Powell JS, Robertson RP, AdamsonJW. High levels of the circulating form of parathyroid hor-mone do not inhibit in vivo erythropoiesis. J Lab Clin Med1983; 102:613–620.

121. Besarab A. Anemia in renal disease. In: Schrier RW, ed. Dis-eases of the Kidney and Urinary Tract. 7th ed. Philadelphia:Lippincott Williams and Wilkins, 2001:2719–2734.

122. Ifudu O, Dawood M, Friedman EA. Relative contributions ofbody iron status and uremia severity to anemia in patientswith advanced chronic renal failure. Nephron 1997; 77:315–318.

123. Kaysen GA. The microinflammatory state in uremia: causesand potential consequences. J Am Soc Nephrol 2001;12:1549–1557.

124. Borawski J, Pawlak K, Mysliwiec. Inflammatory markersand platelet aggregation tests as predictors of hemoglobinand endogenous erythropoietin levels in hemodialysispatients. Nephron 2002; 91:671–681.

125. Kimmel PL, Chawla LS, Amarasinghe A, Peterson RA,Weihs KL, Simmens SJ, Alleyne S, Burke HB, Cruz I, VeisJH. Anthropometric measures, cytokines and survival inhemodialysis patients. Nephrol Dial Transplant 2003;18:326–332.

708 Ajuria et al.

126. Raja KB, Simpson RJ, Pippard MJ, Peters TJ. In vivo studiesof the relationship between intestinal iron (FE3þ) absorp-tion, hypoxia and erythropoiesis. Br J Haematol 1988; 68:373–378.

127. Jongen-Lavrencic M, Peeters HRM, Vreugdenhil G, SwaakAJG. Interaction of inflammatory cytokines and erythropoie-tin in iron metabolism and erythropoiesis in anemia ofchronic disease. Clin Rheum 1995; 14:519–525.

128. Konijn AM, Carmel N, Levy R, Hershko C. Ferritin synthesisin inflammation. II. Mechanism of increased ferritin synth-esis. Br J Haematol 1981; 49:361–370.

129. Fillet G, Beguin Y, Baldelli L. Model of reticuloendothelialiron metabolism in humans: abnormal behavior in idiopathichemochromatosis and in inflammation. Blood 1989; 74:844–851.

130. Baumann H, Gauldie J. Regulation of hepatic acute phaseplasma protein genes by hepatocyte stimulating factors andother mediators of inflammation. Mol Biol Med 1990; 7:147–159.

131. Johnson RA, Waddelow TA, Caro J, Oliff A, Roodman GD.Chronic exposure to tumor necrosis factor in vivo preferen-tially inhibits erythropoiesis in nude mice. Blood 1989; 74:130–138.

132. Rogers J, Lacroix L, Durmowitz G, Kasschau K, AndriotakisJ, Bridges KR. The role of cytokines in the regulation offerritin expression. Adv Exp Med Biol 1994; 356:127–132.

133. Means RT, Dessypris EN, Krantz SB. Inhibition of humanerythroid colony-forming units by interleukin-1 is mediatedby gamma interferon. J Cell Physiol 1992; 150:59–64.

134. Schooley JC, Kullgren B, Allison AC. Inhibition by interleu-kin-1 of the action of erythropoietin on erythroid precursorsand its possible role in the pathogenesis of hypoplastic anae-mias. Br J Haematol 1987; 67:11–17.

135. Mamus SW, Beck-Schroeder S, Zanjani ED. Suppression ofnormal human erythropoiesis by gamma interferon in vitro.Role of monocytes and T-lymphocytes. J Clin Invest 1985; 75:1496–1503.


136. Faquin WC, Schneider TJ, Goldberg MA. Effect of inflamma-tory cytokines on hypoxia-induced erythropoietin production.Blood 1992; 79:1987–1994.

137. Frede S, Fandrey J, Pagel H, Hellwig T, Jelkmann W. Ery-thropoietin gene expression is suppressed after lipopolysac-charide or interleukin-1 beta injections in rats. Am JPhysiol 1997; 273:R1067–R1071.

138. Means RT, Dessypris EN, Krantz SB. Inhibition of humancolony-forming-unit erythroid by tumor necrosis factorrequires accessory cells. J Clin Invest 1990; 86:538–541.

139. Allen DA, Breen C, Yaqoob MM, Macdougal IC. Inhibition ofCFU-E colony formation in uremic patients with inflamma-tory disease: role of IFN-gamma and TNF-alpha. J InvestigMed 1999; 47:204–211.

140. Barany P, Divino Filho JC, Bergstrom J. High C-reactive pro-tein is a strong predictor of resistance to erythropoietin inhemodialysis patients. Am J Kidney Dis 1997; 29:565–568.

141. Ifudu O, Feldman J, Friedman EA. The intensity of hemodia-lysis and the response to erythropoietin in patients with end-stage renal disease. N Engl J Med 1996; 334:420–425.

142. Delmez JA, Tindira C, Grooms P, Dusso A, Windus DW,Slatopolsky E. Parathyroid hormone suppression by intrave-nous 1,25-dihydroxyvitamin D: a role for increased sensitiv-ity to calcium. J Clin Invest 1989; 83:1349–1355.

143. Dunlay R, Rodriguez M, Felsenfeld AJ, Llach F. Direct inhi-bitory effect of calcitriol on parathyroid function (sigmoidalcurve) in dialysis. Kidney Int 1989; 36:1093–1098.

144. Pitts TO, Piraino BH, Mitro R, Chen TC, Segre GV, Green-berg A, Puschett JB. Hyperparathyroidism and 1,25-dihydroxyvitamin D deficiency in mild, moderate, and severerenal failure. J Clin Endocrinol Metab 1988; 67:876–881.

145. Kates DM, Sherrard DJ, Andress DL. Evidence that serumphosphate is independently associated with serum PTH inpatients with chronic renal failure. Am J Kidney Dis 1997;30:809–813.

710 Ajuria et al.

146. Zingraff J, Drueke T, Marie P, Man NK, Jungers P, Bordier P.Anemia and secondary hyperparathyroidism. Arch InternMed 1978; 138:1650–1652.

147. Falko JM, Guy JT, Smith RE, Mazzaferri EL. Primary hyper-parathyroidism and anemia. Arch Intern Med 1976; 136:887–889.

148. Boxer M, Ellman L, Geller R, Wang C-A. Anemia in primaryhyperparathyroidism. Arch Intern Med 1977; 137:588–590.

149. Coen G, Calabria S, Bellinghieri G, Pecchini F, Conte F, Chip-pini MG, Ferrannini M, Lagona C, Mallamace A, Manni M,DiLuca M, Sardella D, Taggi F. Parathyroidectomy in chronicrenal failure: short and long-term results on parathyroid func-tion, blood pressure and anemia. Nephron 2001; 88:149–155.

150. McGonigle RJ, Wallin JD, Husserl F, Deftos LJ, Rice JC,O’Neill WJ, Fisher JW. Potential role of parathyroid hormoneas an inhibitor of erythropoiesis in the anemia of renal fail-ure. J Lab Clin Med 1984; 104:1016–1026.

151. Rao DS, Shih M-S, Mohini R. Effect of serum parathyroidhormone and bone marrow fibrosis on the response to ery-thropoietin in uremia. N Engl J Med 1993; 328:171–175.

152. Yasunaga C, Matsuo K, Yanagida T, Matsuo S, Nakamoto M,Goya T. Early effects of parathyroidectomy on erythropoietinproduction in secondary hyperparathyroidism. Am J Surg2002; 183:199–204.

153. DeissA,LeeGR,CartwrightGE.Hemolytic anemia inWilson’sdisease. Ann Intern Med 1970; 73:413–418.

154. Brady HR, Brenner BM, Lieberthal W. Acute renal failure.In: Brenner BM, ed. The Kidney. 5th ed. Philadelphia:W.B.Saunders, 1996:1206.

155. Ajuria JL, Kimmel PL. Choice of the hemodialysis mem-brane. In: Henrich W, ed. Principles and Practice of Dialysis.3rd ed. Philadelphia: Lippincott Williams and Wilkins, 2004:1–15.

156. Berkes SL, Kahn SI, Chazan JA, Garella S. Prolonged hemo-lysis from overheated dialysis. Ann Intern Med 1975;83:363–364.


157. Manzler AD, Schreiner AW. Copper-induced acute hemolyticanemia. Ann Intern Med 1970; 73:409–412.

158. Carlson DJ, Shapiro FL. Methemoglobin from well waternitrates: a complication of hemodialysis. Ann Intern Med1970; 73:757–759.

159. Eaton JW, Kolpin CF, Kjellstrand C-M, Jacob HS. Chlori-nated urban water: a cause of dialysis induced hemolyticanemia. Science 1973; 181:463–464.

160. Gallery ED, Blomfield J, Dixon SR. Acute zinc toxicity inhaemodialysis. BMJ 1972; 4:331–333.

161. Elliot H, Dryburg F, Fell G, Sabet S, MacDougall AI. Alumi-num toxicity during regular haemodialysis. BMJ 1978;1:1101–1103.

162. Joske RA, McAlister JM, Prankerd TAJ. Isotope investiga-tions of red cell production and destruction in chronic renaldisease. Clin Sci 1956; 15:511–522.

163. Yawata Y, Howe R, Jacob HS. Abnormal red cell metabolismcausing hemolysis in uremia: a defect potentiated by waterhemodialysis. Ann Intern Med 1973; 79:362–367.

164. Jacob HS, Eaton JW, Yawata Y. Shortened red blood cellsurvival in uremic patients: beneficial and deleterious effectsof dialysis. Kidney Int 1975; (suppl 2):139–143.

165. Yawata Y, Jacob HS. Abnormal red cell metabolism inpatients with chronic uremia: nature of the defect and its per-sistence despite adequate hemodialysis. Blood 1975; 45:231–239.

166. Lonergan ET, Semar M, Lange K. Transketolase activity inuremia. Arch Intern Med 1970; 126:851–854.

167. Giovannetti S, Cioni L, Balestri PL, Biagnini M. Evidencethat guanidines and some related compounds cause hemoly-sis in chronic uremia. Clin Sci 1968; 34:141–148.

168. Shainkin-Kestenbaum R, Giatt Y, Berlyne GM. The toxicityof guanidine compounds in the rat blood cell in uremia andthe effect of hemodialysis. Nephron 1982; 31:20–23.

169. Jacob HS, Amsden T. Acute hemolytic anemia with rigid cellsin hypophosphatemia. N Engl J Med 1971; 285:1446–1450.

712 Ajuria et al.

170. Kaiser U, Barth N. Haemolytic anemia in a patient withanorexia nervosa. Acta Haematol 2001; 106:133–135.

171. Melvin JD, Watts RG. Severe hypophosphatemia: a rarecause of intravascular hemolysis. Am J Hematol 2002;69:223–224.

172. Anagnostou A, Freid W, Kurtzman NA. Hematological conse-quences of renal failure. In: Brenner B, Rector F, eds. TheKidney. 2nd ed. Philadelphia: W.B.Saunders, 1981:2184–2212.

173. Bonomini M, Sirolli V, Gizzi F, Di Stante S, Grilli A, Felaco M.Enhanced adherence of human uremic erythrocytes to vascu-lar endothelium: role of phosphatidylserine exposure. KidneyInt 2002; 62:1358–1363.

174. Zwaall RFA, Schroit AJ. Pathophysiologic implications ofmembrane phospholipid asymmetry in blood cells. Blood1997; 89:1121–1132.

175. McEvoy L, Williamson P, Schlegel RA. Membrane phospholi-pid asymmetry as a determinant of erythrocyte recognitionby macrophages. Proc Natl Acad Sci USA 1986; 83:3311–3315.

176. Bonomini M, Sirolli V, Settefrati N, Dottori S, Di Liberato L,Arduini A. Increased erythrocyte phosphatidylserine expo-sure in chronic renal failure. Am J Kidney Dis 1999;10:1982–1990.

177. Bonomini M, Sirolli V, Reale M, Arduini A. Involvement ofphosphatidylserine exposure in the recognition and phagocyto-sis of uremic erythrocytes. Am J Kidney Dis 2001; 37:807–814.

178. Rosenmund A, Binswanger U, Straub PW. Oxidative injuryto erythrocytes, cell rigidity, and splenic hemolysis in hemo-dialyzed uremic patients. Ann Intern Med 1975; 82:460–465.

179. Forman S, Bischel M, Hochstein P. Erythrocyte deformity inuremic hemodialyzed patients. Ann Intern Med 1973; 79:841–843.

180. Hartley LCJ, Morgan TO, Innis MD, Clunie GJA. Splenect-omy for anaemia in patients on regular haemodialysis.Lancet 1971; 2:1343–1345.


181. Annuk M, Fellstrom B, Akerblom O, Zilmer K, Vihalemm T,Zilmer M. Oxidative stress markers in pre-uremic patients.Clin Nephol 2001; 56:308–314.

182. Canestrari F, Galli F, Giorgini A, Albertini MC, Galiotta P,Pascucci M, Bossu M. Erythrocyte redox state in uremic ane-mia: effects of hemodialysis and relevance of glutathionemetabolism. Acta Haematol 1994; 91:187–193.

183. Himmelfarb J, Stenvinkel P, Ikizler A, Hakim RM. Theelephant in uremia: oxidant stress as a unifying concept ofcardiovascular disease in uremia. Kidney Int 2002; 62:1524–1538.

184. Siems WG, Sommerburg O, Grune T. Erythrocyte free radicaland energy metabolism. Clin Nephrol 2000; 53(suppl 1):S9–S17.

185. Carrel R, Winterbourn C, Rachmilewitz E. Activated oxygenand hemolysis. Br J Haematol 1975; 30:259–264.

186. Giardini O, Taccone-Gallucci M, Lubrano R, Ricciardi-TenoreG, Bandino D, Silvi I, Ruberto U, Casciani CU. Evidence ofred blood cell membrane lipid peroxidation in haemodialysispatients. Nephron 1984; 36:235–237.

187. Delmas-Beauvieux MC, Combe C, Peuchant E,CarbonneauMA,DubourgL, dePrecigout V, AparicioM, ClercM. Evaluation of red blood cell lipoperoxidation in hemodia-lysed patients during erythropoietin therapy supplementedor not with iron. Nephron 1995; 69:404–410.

188. Hirayama A, Nagase S, Gotoh M, Takemura K, Tomida C,Ueda A, Aoyagi K, Terao J, Koyama A. Hemodialysis doesnot influence the peroxidative state already present inuremia. Nephron 2000; 86:436–440.

189. Haag-Weber M, Horl WH. Dysfunction of polymorphonuclearleukocytes in uremia. Semin Nephrol 1996; 16:192–201.

190. Rosenkranz AR, Templ E, Traindl O, Heinzl H, Zlabinger GJ.Reactive oxygen product formation by human neutrophils asan early marker for biocompatibility of dialysis membranes.Clin Exp Immunol 1994; 98:300–305.

191. Chiu D, Lubin B. Abnormal vitamin E and glutathioneperoxidase levels in sickle cell anemia. Evidence for increased

714 Ajuria et al.

susceptibility to lipid peroxidation in vivo. J Lab Clin Med1979; 94:542–548.

192. Usberti M, Gerardi G, Micheli A, Tira P, Bufano G, Gaggia P,Movilli E, Cancarini GC, DeMarinis S, D’Avolio G, Broccoli R,Manganoni A, Albertin A, Di Lorenzo D. Effects of a vitaminE-bonded membrane and of glutathione on anemia and ery-thropoietin requirements in hemodialysis patients. J Nephrol2002; 15:558–564.

193. UsbertiM,GerardiG, BufanoG, TiraP,Micheli A, Albertini A,Floridi A, Di Lorenzo D, Galli F. Effects of erythropoietin andvitamin E-modified membrane on plasma oxidative stressmarkers and anemia of hemodialyzed patients. Am J KidneyDis 2002; 40:590–599.

194. Shurtz-Swirski R, Kristal B, Shasha SM, Shapiro G, Geron R,Sela S. Interaction between erythropoietin and peripheralpolymorphonuclear leukocytes in continuous ambulatory dia-lysis patients. Nephron 2002; 91:759–761.

195. Zachee P, Ferrant A, Daelemans R, Coolen L, Goossens W,Lins RL, Couttenye M, De Broe ME, Boogaerts MA. Oxida-tive injury to erythrocytes, cell rigidity and splenic hemolysisin hemodialyzed patients before and during erythropoietintreatment. Nephron 1993; 65:288–293.

196. Soldati L, Adamo D, Zerbi S, Caumo A, Spaventa R, Bianchi G,Vezzoli G. Erythrocyte voltage-dependent calcium influx isreduced in hemodialyzedpatients.Kidney Int 1999; 56:190–197.

197. Corry DB, Tuck ML, Brickman AS, Yanagawa N, Lee DBN.Sodium transport in red blood cells from dialyzed uremicpatients. Kidney Int 1986; 29:1197–1202.

198. Cheng JT, Kahn T, Kaji DM. Mechanism of alteration ofsodium potassium pump of erythrocytes from patientswith chronic renal failure. J Clin Invest 1984; 74:1811–1820.

199. Izumo H, Izumo S, DeLuise M, Flier JS. Erythrocyte Na, Kpump in uremia: acute correction of a transport defect byhemodialysis. J Clin Invest 1984; 74:581–588.

200. Kelly RA, Canessa ML, Steinman TI, Mitch WE. Hemodialy-sis and red cell cation transport in uremia: Role of membranefree fatty acids. Kidney Int 1989; 35:595–603.


201. Zidek W, Rustemeyer T, Schluter W, Karas M, Kisters K.Isolation of an ultrafilterable Ca(2þ)-ATPase inhibitor fromthe plasma of uraemic patients. Clin Sci 1992; 82:659–665.

202. Mason EE. Gastrointestinal lesions occurring in uremia. AnnIntern Med 1952; 37:96–105.

203. Chenney K, Bonnin JA. Haemorrhage, platelet dysfunction,and other coagulation defects in uremia. Br J Haematol1961; 8:215–227.

204. Rath CE, Maillard JA, Schreiner GE. Bleeding tendency inuremia. N Engl J Med 1957; 257:808–811.

205. Ando M, Iwamoto Y, Suda A, Tsuchiya K, Nihei H. Newinsights into the thrombopoietic status of patients on dialysisthrough the evaluation of megakaryocytopoiesis in bone mar-row and of endogenous thrombopoietin levels. Blood 2001;97:915–921.

206. Eknoyan G, Brown GH. Biochemical abnormalities of plate-lets in renal failure. Evidence for decreased platelet seroto-nin, adenosine diphosphate and Mg-dependent adenosinetriphosphate. Am J Nephrol 1981; 1:17–23.

207. Kyrle PA, Stockenhuber F, Brenner B, Gossinger H,Korninger C, Pabinger I, Sunder-Plassmann G, Balcke P,Lechner K. Evidence for an increased generation of prostacy-clin in the microvasculature and an impairment of the plate-let alpha-granule release in chronic renal failure. ThrombHaemost 1988; 60:205–208.

208. Smith MC, Dunn MJ. Impaired platelet thromboxane produc-tion in renal failure. Nephron 1981; 29:133–137.

209. Remuzzi G, Benigni A, Dodesini P, Schieppati A, Livio M, DeGaetano, Day JS, Smith WL, Pinca E, Patrignani P, PatronoC. Reduced platelet thromboxane formation in uremia: evi-dence for a functional cyclooxygenase defect. J Clin Invest1983; 71:762–768.

210. Remuzzi G, Cavenaghi AE, Mecca G, Donati MB, de GaetanoG. Prostacyclin-like activity and bleeding in renal failure.Lancet 1977; 2:1195–1197.

211. Zoja C, Vigano G, Bergamelli A, Benigni A, de Gaetano G,Remuzzi G. Prolonged bleeding time and increased vascular

716 Ajuria et al.

prostacyclin in rats with chronic renal failure: effects of con-jugated estrogens. J Lab Clin Med 1988; 112(Sep):380–386.

212. Noris M, Benigni A, Boccardo P, Aiello S, Gaspari F,Todeschini M, Figliuzzi M, Remuzzi G. Enhanced nitric oxidesynthesis in uremia: implications for platelet dysfunction anddialysis hypotension. Kidney Int 1993; 44:445–450.

213. Remuzzi G, Perico N, Zoja C, Corna D, Macconi D, Vigano G.Role of endothelium-derived nitric oxide in the bleedingtendency of uremia. J Clin Invest 1990; 86:1768–1771.

214. Castillo R, Lozano T, Escolar G, Revert L, Lopez J, Ordinas A.Defective platelet adhesion on vessel subendothelium in ure-mic patients. Blood 1986; 68:337–342.

215. Escolar G, Cases A, Bastida E, Garrido M, Lopez J, Revert L,Castillo R, Ordinas A. Uremic platelets have a functionaldefect affecting the interaction of von Willebrand factor withglycoprotein IIb–IIIa. Blood 1990; 76:1336–1340.

216. Benigni A, Boccardo P, Galbusera M, Monteagudo J, DeMarco L, Remuzzi G. Reversible activation defect of the plate-let glycoprotein IIb–IIIa complex in patients with uremia. AmJ Kidney Dis 1993; 22:668–676.

217. Mannucci PM, Remuzzi G, Pusineri F, Lombardi R,Valsecchi C, Mecca G, Zimmerman TS. Deamino-8-d-argininevasopressin shortens the bleeding time in uremia. N Engl JMed 1983; 308:8–12.

218. Kazatchkine M, Sultan Y, Caen JP, Bariety J. Bleeding inrenal failure: a possible cause. BMJ 1976; 2:612–615.

219. Rabiner SF. The effect of dialysis on platelet function ofpatients with renal failure. Ann N Y Acad Sci 1972; 201:234–242.

220. Rabiner SF, Drake RF. Platelet function as an indicator ofadequate dialysis. Kidney Int 1975; (suppl 2):144–146.

221. DiMinno G,Martinez J,McKeanML,De LaRosa J, Burke JF,Murphy S. Platelet dysfunction in uremia. Multifaceted defectpartially corrected by dialysis. Am J Med 1985; 79:552–559.

222. Mannucci PM, Canciani MT, Rota L, Donovan BS. Responseof factor VIII=von Willebrand factor to DDAVP in healthy


subjects and patients with haemophilia A and vonWillebrand’s disease. Br J Haematol 1981; 47:283–293.

223. Vigano G, Gaspari F, Locatelli M, Pusineri F, Bonati M,Remuzzi G. Dose-effect and pharmacokinetics of estrogensgiven to correct bleeding time in uremia. Kidney Int 1988;34:853–858.

224. Zoja C, Vigano G, Corna D, Salmona M, Remuzzi G,Garattini S. Oral zeranol shortens the prolonged bleeding timeof uremic rats. Kidney Int 1990; 38:96–100.

225. Janson PA, Jubelirer SJ, Weinstein MS, Deykin D. Treat-ment of bleeding tendency in uremia with cryoprecipitate.N Engl J Med 1980; 303:1318–1322.

226. Livio M, Gotti E, Marchesi D, Mecca G, Remuzzi G,De Gaetano G. Uremic bleeding: role of anemia and beneficialeffect of red cell transfusions. Lancet 1982; 2:1013–1015.

227. Vigano G, Benigni A, Mendogni D, Mingardi G, Mecca G,Remuzzi G. Recombinant human erythropoietin to correcturemic bleeding. Am J Kidney Dis 1991; 18:44–49.

228. Akizawa T, Kinugasa E, Kitaoka T, Koshikawa S. Effects ofrecombinant human erythropoietin and correction of anemiaon platelet function in hemodialysis patients. Nephron 1991;58:400–406.

229. El-Shahawy MA, Francis R, Akmai M, Massry SG. Recombi-nant human erythropoietin shortens the bleeding time andcorrects the abnormal platelet aggregation in hemodialysispatients. Clin Nephrol 1994; 41:308–313.

230. Zwaginga JJ, Ijsseldijk MJW, de Groot PG, Kooistra M, Vos J,van Es A, Koomans HA, Struyvenberg A, Sixma JJ. Treat-ment of uremic anemia with recombinant erythropoietin alsoreduces the defects in platelet adhesion and aggregationcaused by uremic plasma. Thromb Haemost 1991; 66:638–647.

231. Cotter D, Thamer M, Kimmel PL, Sadler JH. Secular trendsin recombinant erythropoietin therapy among the U.S. hemo-dialysis population: 1990–1996. Kidney Int 1998; 54:2129–2139.

718 Ajuria et al.

232. Obrador GT, Roberts T, St Peter WL, Frazier E, Pereira BJG,Collins AJ. Trends in anemia at initiation of dialysis in theUnited States. Kidney Int 2001; 60:1875–1884.

233. Obrador GT, Ruthazer R, Arora P, Kausz AT, Pereira BJG.Prevalence of and factors associated with suboptimal carebefore initiation of dialysis in the United States. J Am SocNephrol 1999; 10:1793–1800.

234. Valderrabano F, Horl WH, Macdougall IC, Rossert J,Rutkowski B, Wauters JP. Pre-dialysis survey on anaemiamanagement. Nephrol Dial Transplant 2003; 18:89–100.

235. Centers for Medicare and Medicaid Services. 2002 AnnualReport, End Stage Renal Disease Clinical Performance Mea-sures Project. Baltimore, MD: Department of Health andHuman Services, Centers for Medicare and Medicaid Ser-vices, Center for Beneficiary Choices, December 2002.

236. Flanigan MJ, Rocco MV, Frankenfield DL, Bailie G,Frederick PR, Prowant BF, Taylor L. 1996 Peritoneal dialy-sis-core indicators report. Am J Kidney Dis 1998; 32:1–9.

237. Kimmel PL, Greer J, Milam R, Thamer M. Trends in erythro-poietin therapy in the US dialysis population: 1995–1998.Semin Nephrol 2000; 20:335–344.

238. Hynes DM, Stroupe KT, Greer JW, Reda DJ, FrankenfieldDL, Kaufman JS, Henderson WG, Owen WF, Rocco MV,Wish JB, Kang J, Feussner JR. Potential cost savings of ery-thropoietin administration in end stage renal disease. Am JMed 2002; 112:169–175.

239. Kaufman JS, Reda DJ, Fye CL, Goldfarb DS, Henderson WG,Kleinman JG, Vaamonde CA. Subcutaneous compared withintravenous epoetin in patients receiving hemodialysis. NEngl J Med 1998; 339:578–583.

240. Kausz AT, Obrador GT, Pereira BJG. Anemia managementin patients with chronic renal insufficiency. Am J KidneyDis 2000; 36(suppl 3):S39–S51.

241. Eschbach JW, Kelly MR, Haley NR, Abels RI, Adamson JW.Treatment of the anemia of progressive renal failure withrecombinant human erythropoietin. N Eng J Med 1989;321:158–163.


242. Lim VS, Kirchner PT, Fangman J, Richmond J, DeGowin RL.The safety and the efficacy of maintenance therapy of recom-binant human erythropoietin in patients with renal insuffi-ciency. Am J Kidney Dis 1989; 14:496–506.

243. Frenken LAM, Wetzels JFM, Sluiter HE, Koene RAP. Evi-dence for renal vasodilation in pre-dialysis patients duringcorrection of anemia by erythropoietin. Kidney Int 1992;41:384–387.

244. Frankenfield D, Johnson CA, Wish JB, Rocco MV, Madore F,Owen WF, for the ESRD Core Indicators Workgroup. Anemiamanagement of adult hemodialysis patients in the U.S.:results from the 1997 ESRD Core Indicators Project. KidneyInt 2000; 57:578–589.

245. Silverberg DS, Blum M, Agbaria Z, Schwartz D, Zubkov A,Yachnin T, Iaina A. Intravenous iron for the treatment ofpredialysis anemia. Kidney Int 1999; 55(suppl 69):S79–S85.

246. Baile GR, Frankenfield DL, Prowant BF, McClellan W, RoccoMV. Erythropoietin and iron use in peritoneal dialysispatients: report from the 1997 HCFA End-Stage Renal Dis-ease Core Indicators Project. Am J Kidney Dis 1999;33:1187–1189.

247. Winearls CG, Oliver DO, Pippard MJ, Reid C, Downing MR,Coles PM. Effect of human erythropoietin derived fromrecombinant DNA on the anaemia of patients maintainedby chronic haemodialysis. Lancet 1986; 2:1175–1178.

248. Fishbane S, Frei GL, Maesaka J. Reduction in recombinanthuman erythropoietin doses by the use of chronic intravenousiron supplementation. Am J Kidney Dis 1995; 26:41–46.

249. Owen WF, Szczech L, Johnson C, Frankenfield DL. Nationalperspective on iron therapy as a clinical performance mea-sure for maintenance hemodialysis patients. Am J KidneyDis 1999; 34(suppl 2):S5–S11.

250. Eknoyan G, Levin N. NKF-K=DOQI clinical practice guide-lines for anemia of chronic kidney disease: update 2000. AmJ Kidney Dis 2001; 37(suppl 1):S182–S238.

251. Adamson JW, Eschbach JW. Erythropoietin for end-stagerenal disease. N Engl J Med 1998; 339:625–627.

720 Ajuria et al.

252. Besarab A, Bolton WK, Browne JK, Egrie JC, Nissenson AR,Okamoto DM, Schwab SJ, Goodkin DA. The effects of normalas compared with low hematocrit values in patients withcardiac disease who are receiving hemodialysis and epoetin.N Eng J Med 1998; 339:584–590.

253. Veys N, Vanholder R, De Guyper K, Ringoir S. Influence oferythropoietin on dialyzer re-use, heparin needs, and ureakinetics in maintenance hemodialysis patients. Am J KidneyDis 1994; 23:52–59.

254. Spinowitz BS, Arslanian J, Charytan C, Golden RA, Rascoff J,Galler M. Impact of epoetin beta on dialyzer clearance andheparin requirements. Am J Kidney Dis 1991; 18:668–673.

255. Acchiardo SR, Quinn BP, Moore LW, Burk LB, Miles DE.Evaluation of hemodialysis patients treated with erythro-poietin. Am J Kidney Dis 1991; 17:290–294.

256. Eschbach JW,AbdulhadiMH,BrowneJK,DelanoBG,DowningMR, Egrie JC, Evans RW, Friedman EA, Graber SE, Haley NR,Korbet S, Krantz SB, Lundin PA, Nissenson AR, Ogden DA,Poganini EP, Rader B, Rutsky EA, Stivelman J, Stone WJ,Teschan P, van Stone JC, Van Wyck DB, Zuckerman K, Adam-son JW. Recombinant human erythropoietin in anemic patientswith end-stage renal disease. Results of a phase III multicenterclinical trial. Ann Intern Med 1989; 111:992–1000.

257. Raine AEG, Roger SD. Effects of erythropoietin on blood pres-sure. Am J Kidney Dis 1991; 18(suppl 1):76–83.

258. Abraham PA, Macres MG. Blood pressure in hemodialysispatients during amelioration of anemia with erythropoietin.J Am Soc Nephrol 1991; 2:927–936.

259. Kaupke CJ, Kim S, Vaziri ND. Effect of erythrocyte mass onarterial blood pressure in dialysis patients receiving mainte-nance erythropoietin therapy. J Am Soc Nephrol 1994;4:1874–1878.

260. Neff MS, Kim KE, Persoff M, Onesti G, Swartz C. Hemody-namics of uremic anemia. Circulation 1971; 43:876–883.

261. Raine AE. Hypertension, blood viscosity, and cardiovascularmorbidity in renal failure: implication of erythropoietintherapy. Lancet 1988; 1:97–100.


262. Churchill DN, Muirhead N, Goldstein M, Posen G, Fay W,Beecroft ML, Gorman J, Taylor DW. Probability of thrombo-sis of vascular access among hemodialysis patients treatedwith recombinant human erythropoietin. J Am Soc Nephrol1994; 4:1809–1813.

263. Lacombe C. Resistance to erythropoietin. N Engl J Med 1996;334:660–662.

264. Stivelman JC. Resistance to recombinant human erythro-poietin therapy: a real clinical entity? Semin Nephrol 1989;9(suppl 2):8–11.

265. Bamonti-Catena F, Buccianti G, Porcella A, Valenti G,Como G, Finazzi S, Maiolo AT. Folate measurements inpatients on regular hemodialysis treatment. Am J KidneyDis 1999; 33:492–497.

266. Vlahakos DV, Canzanello VJ, Madaio MP, Madias NE. Ena-lapril-associated anemia in renal transplant recipients trea-ted for hypertension. Am J Kidney Dis 1991; 17:199–205.

267. Yoshida A, Morozumi K, Suganuma T, Aoki J, Sugito K,Nakamura S, Ikeda M, Oikawa T, Fujinami T, KawaharaH. Angiotensin-converting enzyme inhibitor and anemia ina patient undergoing hemodialysis. Nephron 1991; 59:334–335.

268. Cruz DN, Perazella MA, Abu-Alfa AK, Mahnensmith RL.Angiotensin-converting enzyme inhibitor therapy in chronichemodialysis patients: any evidence of erythropoietinresistance? Am J Kidney Dis 1996; 28:535–540.

269. Casedevall N, Nataf J, Viron B, Kolta A, Kiladjian JJ, Martin-Dupont P, Michaud P, Papo T, Ugo V, Teyssandier I, Varet B,Mayeux P. Pure red-cell aplasia and antierythropoietin antibo-dies in patients treated with recombinant erythropoietin. NEngl J Med 2002; 346:469–475.

270. Foley RN, Parfrey PS, Harnett JD, Kent GM, Murray DC,Barre PE. The impact of anemia on cardiomyopathy, morbid-ity, and mortality in end-stage renal disease. Am J KidneyDis 1996; 28:53–61.

271. Foley RN, Parfrey PS, Morgan J, Barre PE, Campbell P,Cartier P, Coyle D, Fine A, Handa P, Kingma I, Lau CY,

722 Ajuria et al.

Levin A, Mendelssohn D, Muirhead N, Murphy B, Plante RK,Posen G, Wells GA. Effect of hemoglobin levels in hemodialy-sis patients with asymptomatic cardiomyopathy. Kidney Int2000; 58:1325–1335.

272. Levin A, Thompson CR, Either J, Carlisle EJF, Tobe S,Mendelssohn D, Burgess E, Jindal K, Barrett B, Singer J,Djurdjev O. Left ventricular mass index increase in earlyrenal disease: impact of decline in hemoglobin. Am J KidneyDis 1999; 34:125–134.

273. Levin A, Singer J, Thompson CR, Ross H, Lewis M. Prevalentleft ventricular hypertrophy in the predialysis population:identifying opportunities for intervention. Am J Kidney Dis1996; 27:347–354.

274. Harnett JD, Foley RN, Kent GM, Barre PE, Murray D,Parfrey PS. Congestive heart failure in dialysis patients: pre-valence, incidence, prognosis and risk factors. Kidney Int1995; 47:884–890.

275. Wizemann V, Kaufmann J, Kramer W. Effect of erythropoie-tin on ischemia tolerance in anemic hemodialysis patientswith confirmed coronary artery disease. Nephron 1992;62:161–165.

276. Fowler NO, Holmes JC. Blood viscosity and cardiac output inacute experimental anemia. J Appl Physiol 1975; 39:453–456.

277. Canadian Erythropoietin Study Group. Association betweenrecombinant human erythropoietin and quality of life andexercise capacity of patients receiving haemodialysis. BMJ1990; 300:573–578.

278. Xia H, Ebben J, Ma JZ, Collins AJ. Hematocrit levels andhospitalization risks in hemodialysis patients. J Am SocNephrol 1999; 10:1309–1316.

279. Ma JZ, Ebben J, Xia H, Collins AJ. Hematocrit level andassociated mortality in hemodialysis patients. J Am SocNephrol 1999; 10:610–619.

280. Mann J, Gerstein HC, Pogue J, Bosch J, Yusef S. Renal insuf-ficiency as a predictor of cardiovascular outcomes and theimpact of ramipril: the HOPE randomized trial. Ann InternMed 2001; 134:629–636.


281. Parfrey PS, Harnett JD, Griffiths SM, Gault MH, Barre PE.Congestive heart failure in dialysis patients. Arch InternMed 1988; 148:1519–1525.

282. Silberberg JS, Barre PE, Prichard SS, Sniderman AD. Impactof left ventricular hypertrophy on survival in end stage renaldisease. Kidney Int 1989; 36:286–290.

283. Foley RN, Parfrey PS, Sarnak MJ. Clinical epidemiology ofcardiovascular disease in chronic renal disease. Am J KidneyDis 1998; 32(suppl 3):S112–S119.

284. Foley RN, Parfrey PS, Harnett JD, Kent GM, Martin CJ,Murray DC, Barre PE. Clinical and echocardiographic dis-ease in patients starting end-stage renal disease therapy.Kidney Int 1995; 47:186–192.

285. Brannon ES, Merrill AJ, Warren JV, Stead EA Jr. Thecardiac output in patients with chronic anemia as measuredby the technique of right atrial catheterization. J Clin Invest1945; 24:332–336.

286. Powe NR, Griffiths RI, Watson AJ, Anderson GF, de LissovoyG, Greer JW, Herbert RJ, Milam RA, Whelton PK. Effect ofrecombinant erythropoietin on hospital admissions, readmis-sions, length of stay, and costs of dialysis patients. J Am SocNephrol 1994; 4:1455–1465.

287. Ofsthun N, LaBrecque J, Lacson E, Keen M, Lazarus JM.The effects of higher hemoglobin levels on mortality andhospitalization in hemodialysis patients. Kidney Int 2003;63:1908–1914.

288. Collins AJ, Ma JA, Ebben J. Impact of hematocrit on morbid-ity and mortality. Semin Nephrol 2000; 20:345–349.

289. Fellner SK, Lang RM, Neumann A, Korcarz C, Borow KM.Cardiovascular consequences of the correction of the anemiaof renal failure with erythropoietin. Kidney Int 1993;44:1309–1315.

290. Metra M, Cannella G, La Canna G, Guaini T, Sandrini M,Gaggiotti M, Movilli E, Dei Cas I. Improvement in exercisecapacity after correction of anemia in patients with end-stagerenal failure. Am J Cardiol 1991; 68:1060–1066.

724 Ajuria et al.

291. Macdougall IC, Lewis NP, Saunders MJ, Cochlin DL,Davies ME, Hutton RD, Fox KAA, Coles GA, Williams JD.Long-term cardiorespiratory effects of amelioration of renalanemia by erythropoietin. Lancet 1990; 335:489–493.

292. Davenport A. The effect of treatment with recombinanthuman erythropoietin on human skeletal muscle functionin patients with end-stage renal failure treated with regularhemodialysis. Am J Kidney Dis 1993; 22:685–690.

293. Revicki DA, Brown RE, Feeny DH, Henry D, Teehan BP,Rudnick MR, Benz RL. Health-related quality of life asso-ciated with recombinant human erythropoietin therapy forpredialysis chronic renal disease patients. Am J Kidney Dis1995; 25:548–554.

294. Evans RW, Rader B, Manninen DL, for the Cooperative Mul-ticenter EPO Clinical Trial Group. The quality of life of hemo-dialysis recipients treated with recombinant humanerythropoietin. JAMA 1990; 263:825–830.

295. Evans RW. Recombinant human erythropoietin and the qual-ity of life end-stage renal disease patients: a comparativeanalysis. Am J Kidney Dis 1991; 18(suppl 1):S62–S70.

296. Marsh JT, Brown WS, Wolcott D, Carr CR, Harper R,Schweitzer SV, Nissenson AR. rHuEPO treatment improvesbrain cognitive function in anemic dialysis patients. KidneyInt 1991; 39:155–163.

297. Ramirez G, Bittle PA, Saunders H, Rabb HAA, Bercu BB.The effects of corticotrophin and growth hormone releasinghormones on their respective secretory axes in chronichemodialysis patients before and after correction of anemiawith recombinant human erythropoietin. J Clin EndocrinolMetab 1994; 78:63–69.

298. Schaefer RM, Kokot F, Heidland A. Impact of recombinanterythropoietin on sexual function in hemodialysis patients.Contrib Nephrol 1989; 76:273–282.

299. Sobh MA, Abd el Hamid IA, Atta MG, Refaie AF. Effect oferythropoietin on sexual potency in hemodialysis patients: apreliminary study. Scand J Urol Nephrol 1992; 26:181–185.


300. Diez JJ, Iglesias P, Selgas R. Influence of recombinanthuman erythropoietin on growth hormone response to growthhormone releasing hormone in uremic patients. Clin Nephrol1994; 41:119.

301. Veys N, Vanholder R, Ringoir S. Correction of deficient pha-gocytosis during erythropoietin treatment in maintenancehemodialysis patients. Am J Kidney Dis 1992; 19:358–363.

302. Sennesael JJ, Van der Niepen P, Verbeelen DL. Treatmentwith recombinant human erythropoietin increases antibodytiters after hepatitis B vaccination in dialysis patients. Kid-ney Int 1991; 40:121–128.

303. Keane WF, Brenner BM, De Zeeuw D, Grunfeld JP, McGill J,Mitch WE, Ribeiro AB, Shahinfar S, Simpson RL,Snapinn SM, Toto R, for the RENAAL Study Investigators.The risk of developing end-stage renal disease in patientswith type 2 diabetes and nephropathy: the RENAAL study.Kidney Int 2003; 63:1499–1507.

726 Ajuria et al.

26

Anemia of Chronic Disease inInflammatory Bowel Diseases

CHRISTOPH GASCHE

Department of Medicine 4,Division of Gastroenterology and Hepatology,

Medical University of Vienna andGeneral Hospital Vienna, Vienna, Austria

IMPORTANCE OF ACD IN IBD

Inflammatory bowel disease represents a cluster of chronicinflammatory diseases of the intestine. Individual cases areclassified into either ulcerative colitis or Crohn’s disease byestablished clinical and pathological criteria (1,2). Thosepatients who cannot be allocated are temporarily grouped asindeterminate colitis until one or the other clinical phenotypebecomes apparent. Most patients get diagnosed in their thirddecade of life. The growing frequency of Crohn’s disease inchildren in recent years is not well understood. It is more

727

likely that this is due to earlier and stronger exposure toenvironmental factors rather than to genetic anticipation (3).Symptoms typically include recurrent episodes of diarrhea,diffuse or crampy abdominal pain, arthralgia, and fever. Bloodtests may show an increase in C-reactive protein, leukocytes,and platelets and a decrease in erythrocytes and hemoglobin.

Anemia is indeed a common problem in IBD. Cohort stu-dies report a prevalence of 30%within the general IBD popula-tion but no population based data exist (4). In adolescent andadult cases, significant anemia (hemoglobin below 10.0 g=dL)is already present at 10–20% at the time of diagnosis. Thisanemia may improve upon therapy and during clinical remis-sion, but may recur at flares of disease. Anemia is not only acomplication of inflammatory bowel diseases but rather a typi-cal symptom of it. The hemoglobin levels or hematocrit is actu-ally part of most disease activity scores (5).

In IBD, the hemoglobin level typically drops slowly overperiods of months or years. Acute intestinal bleeding may befound in cases with severe ulcerative colitis, but this hasbecome an infrequent event. Anemia may be the first sign ofdisease. Anemia associated symptoms in IBD include delayedpuberty, amenorrhoea in premenopausal women, or erectiledysfunction in young men. Also other anemia related symp-toms can be primarily observed in IBD including fatigue, head-ache, dizziness, shortness of breath, or tachycardia. Thesesymptoms are caused by a decrease in blood oxygen concentra-tion and peripheral hypoxia. Compensatory blood shifting fromthe mesenteric arteries can decrease the perfusion of intestinalmucosa, which is specifically unwanted in IBD (6). As a result,motility disorder (with irritable bowel disease-like symptomssuch as diarrhea, pain, or constipation), nausea, or loss of appe-tite may contribute to bowel discomfort and weight loss.

SPECIFIC PATHOGENESIS OF ANEMIAIN IBD

Iron Deficiency Anemia

Anemia in IBD may develop by several mechanisms, alone orin combination (Fig. 1). The most prevalent factor is iron

728 Gasche

deficiency (4). Blood loss that occurs in the diseased intestinecannot be matched by duodenal iron absorption. Such a nega-tive iron balance often exists in IBD. Though specific investi-gations are missing, the intestinal ulcerations are consideredas the site of blood loss (7). In particular patients, dietaryrestrictions may also add to iron deficiency. Iron malabsorp-tion was entertained as cause of iron deficiency in Crohn’sdisease, however, convincing studies are missing (8,9). Theo-retically, malabsorption may be relevant to the 5% withduodenal Crohn’s disease. Still, iron deficiency is the mostprominent factor in IBD-associated anemia.

Anemia of Chronic Disease

Chronic bowel inflammation may also affect hematopoiesis.This is not only true for erythropoiesis but also for megakar-yopoiesis. Thrombocytosis is one of the most remarkable find-ings in active IBD and was related to high levels of acutephase proteins and IL-6 (10,11). The exact mechanism andspecifically the relative impact of erythropoietin or thrombo-poietin on increased platelet production are not well

Figure 1 Etiology of IBD-associated anemia.

ACD in Inflammatory Bowel Diseases 729

understood (12,13). Direct experimental data on the presenceof anemia of chronic disease in IBD are also missing. Circulat-ing levels of erythropoietin were found inadequately low tothe degree of anemia (4,14). In some patients, anemia alsopersisted despite intensive iron repletion and responded wellto pharmacological doses of erythropoietin (15,16). Anemia inIBD was related to high IL-1 beta secretion from circulatingmonocyte (14) but no correlations of serum erythropoietinand IL-6 concentrations were found (4). It is believed that ery-thropoietin production is inhibited by cytokines derived fromthe inflamed gut. Indeed, high concentrations of TNF-alpha,IL-1 or IL-6 mRNA or protein were shown in the inflamedgut mucosa, in peri-intestinal adipose tissue of the mesentery,or in the circulation (17–22). Such cytokines can interferewith renal erythropoietin production (23). Although cytokinerelated inhibition of erythropoiesis is suggestive in IBD,direct experimental studies are missing.

Drug-Associated Anemia

Several drugs that are commonly used for treatment of IBDcan interfere with erythropoiesis. Sulfasalazine for long-termuse in ulcerative colitis may impair folate absorption in somepatients leading to mild anemia and macrocytosis (24). Thisfolate related side effect of sulfasalazine is not reported formesalazine, but both drugs share some risk of drug-associatedpure-red-cell aplasia or aplastic anemia (25–27). More impor-tantly, both 6-mercaptopurine and azathioprine have directmyelosuppressive effects. They represent the most effectivetherapy for Crohn’s disease and are therefore frequently used(28). The risk of developing leukopenia or aplasia is specifi-cally high in individuals with low thiopurine methyltrans-ferase activity (29). The enzyme activity of thiopurine methyl-transferase is genetically determined. A pharmacogeneticapproach with thiopurine methyltransferase genotyping orenzymatic testing of thiopurine methyltransferase activitybefore initiating therapy can identify patients at risk(30,31). Different from life threatening aplasia, mild anemiaand macrocytosis commonly develops after several months

730 Gasche

of therapy and is associated with good compliance and generalresponse to treatment (32).

Vitamin Deficiencies

Cobalamin or folate deficiencies may also occur independent ofdrug therapy either through malabsorption or dietary restric-tion. Macrocytic anemia was early recognized as a clinical fea-ture of Crohn’s disease (33). Nowadays, both vitamins arewidely supplementedandclinical evidence of vitamindeficiencyis uncommon.

Obscure Causes of IBD-Associated Anemia

There is basically no form of anemia that has not beendescribed in relation to IBD. Of particular interest, however,is the association of autoimmune hemolytic anemia and ulcera-tive colitis (34–55). In certain cases, red cell antibodies wereproduced in B-cells that had been isolated from the diseasedcolon, or colectomy has shown to cure the disease (56,57).Nevertheless, the vast number of publications on this entityare case reports. They do not allow a conclusion on the actualclinical frequency, which has been described in up to 1.7% ofreferred cases with ulcerative colitis (58).

Various hematological coincident diseases have also beenreported in IBD such as myelodysplastic syndrome, aplasticanemia, or glucose-6-phosphate dehydrogenase deficiency,but without relevance to general clinical practice (59–63).

CURRENT THERAPY

Monitoring of IBD patients with complete blood counts is aroutine measure. Hypochromia and microcytosis indicate irondeficiency with or without concomitant anemia of chronicdisease. Ferritins are usually in the lower normal range,which however does not rule out functional iron deficiency(64,65). Transferrin saturation is typically low as well.

Macrocytosis with mild or no anemia occurs duringazathioprine or 6-mercaptopurine therapy, but may also point


to a cobalamine or folate deficiency. C-reactive protein is use-ful to estimate the intensity of inflammation, specifically inCrohn’s disease. Measurements of transferrin, the solubletransferrin receptor, and erythropoietin plasma levels mayhelp to identify patients with predominant anemia of chronicdisease (66).

Iron Deficiency

Because the majority of patient with IBD-associated anemiaare iron deficient, the therapeutic priority is effective ironreplacement. Oral iron supplements that contain iron in formof ferrous salts (Fe2þ) are widely used in clinical practicebut are frequently associated with side effects. These sideeffects may be fundamentally different from non-IBD patients(67,68). Ferrous salts typically increase common IBD-relatedbowel complaints such as diarrhea or pain. In contrast, innon-IBD patients, the most prevalent side effect is constipa-tion. IBD-related side effects were thought to be due to theferrous form of iron and caused by local induction of oxidativestress (7,69). Indeed, a decrease in plasma oxygen scavengers(as consequence of increased intestinal oxidant activity) wasshown in Crohn’s patients upon treatment with ferrous fuma-rate (68). Ferric trimaltol (Fe3þ) was better tolerated thanferrous sulfate, but it is no readily available in most countries(70).

To overcome the obstacles of oral iron therapy, parent-eral iron preparations (mainly iron sucrose) have becomewidely used for anemia in IBD (67). The direct administrationof iron into circulation requires formulation that prevents thecellular toxicity of iron salts (71). More than 1000 infusionswith iron sucrose (Fe3þ) were given in controlled studieswithout major adverse events, in particular when used as adilute solution (16,66,72). Ten milliliters of iron sucrose (twoampoules Venofer�, Vifor, Switzerland; corresponding to200mg Fe3þ) were diluted in 250mL 0.9% sodium chlorideand given twice weekly during the first two weeks and onceweekly thereafter. Paravenous infusion of dilute solutionsoccurred but did not produce local skin necrosis. After

732 Gasche

1200mg iron sucrose (4weeks) about 65% and after 2000mg(8weeks) about 75% of patients responded (16,66,72). Withlittle variation, the manufacturer recommends to dilute5mL iron sucrose in 100mL 0.9% sodium chloride and to giveit intravenously at least over 15min. Acceptable safety hasbeen shown for single infusions with 7mg=kg body weight(or a maximum of 500mg per infusion) (73). No appropriatedose finding study has been performed and no data exist forrepeated infusions. In the predialysis setting, 300mg perinfusion appeared to be safe (74).

Anemia of Chronic Disease

Anti-TNF strategies have the potency to interact with themechanisms of anemia of chronic disease. A variety of thera-peutic approaches have been tested to specifically inhibit TNFin patients with IBD including infliximab, the CDP571 anti-body, etanercept (a p75 TNF receptor: Fc fusion protein), oronercept (the soluble p55 TNF receptor) (75). Infliximab hasbecome a cornerstone in the treatment of Crohn’s diseasebut is still experimental in ulcerative colitis. It selectivelybinds to soluble and surface-bound TNF and thereby neutra-lizes TNF activity. As a logic consequence of this principle,TNF-induced impairment of erythropoiesis is expected toimprove. Indeed, such effects have been observed in inflixi-mab treated patients with rheumatoid arthritis (76). Reportson similar effects in IBD-related anemia are currently miss-ing. Etanercept, which is not effective in the treatment ofCrohn’s disease, had also no positive effect on hemoglobinlevels (77). Besides infliximab, any drug that leads to healingof the intestinal ulcerations (such as known for azathioprine)may theoretically also improve anemia of chronic disease inthe long term (28,78). Specific studies, however, have not beenperformed.

Supraphysiologic doses of erythropoietin can overcomeACD-associated inhibition of erythropoiesis (79). In IBD, thefirst patients were reported after restoration of their ironreservoir and exclusion of other causes of anemia (15,80).Double-blind controlled trials followed, using either oral (14)


or intravenous (72) concomitant iron supplementation. Alsostudies in pediatric patients demonstrated a significant effectof erythropoietin (81). Since iron sucrose turned out to behighly cost effective on its own, erythropoietin was furtherused only in patients who failed to respond to iron sucrose(82). At this point, virtually all IBD patients with IBD-associated anemia can be successfully treated with the combi-nation of iron sucrose and erythropoietin (Fig. 2).

UNANSWERED ISSUES

As shown in various clinical trials including also patientswith IBD-associated anemia, successful therapy of anemia

Figure 2 Adjusting therapy to hemoglobin levels. Low doses offerrous sulfate or if available ferric trimaltol should be given topatients with iron deficiency but without manifestation of anemia.The suggested dose and frequency of iron sucrose relate to thedegree of anemia (200–600mg per week). Erythropoietin (10.000IU two to three times a week) should be used if there is a baselinetransferrin concentration below 2.9 g=L or no response to ironsucrose within 4weeks. (Reproduced with permission from GUT,Ref. 82.)

734 Gasche

increases the quality of life (72). Two central questions,however, remain: at what hemoglobin level should we initiatetreatment and what is the target level of hemoglobin thatshould be reached? In the oncology setting, the relationshipbetween hemoglobin increase and quality of life gain rangedfrom 8 to 14 g=dL (with the largest improvement between 11and 13g=dL) indicated that our current treatment triggerand target levels may be too low (83). It is tempting to specu-late that IBD patients who are mostly in the middle oftheir working life will profit even more from any gain inhemoglobin.

In IBD, it is our main goal to prevent or treat iron defi-ciency early before significant anemia develops. The adaptiveincrease of iron absorption capacity during iron deficiency isrelatively small. Most parts of the ingested iron are not takenup but passed on to the lower ileum and colon, the sites ofinflammation in Crohn’s disease and ulcerative colitis. Whenthis iron comes in contact with the ulcerated intestinal wall, itmay enhance the local production of reactive oxygen speciesand increase tissue injury. This has been demonstrated inanimal models of IBD (69). The undesired effects of oxidativestress limit the use of oral iron in IBD (84). Iron and free oxy-gen radicals may cause DNA damage and may potentially fuelcolorectal carcinogenesis in IBD. The procarcinogenic effectsof iron on the development of colorectal cancer have just beendemonstrated in patients carrying hemochromatosis (HFE)gene mutations (85).

The time required for uncomplexed ferrous iron (Fe2þ) toundergo oxidation to the ferric state (Fe3þ) is dependent onmany factors, the dominant being pH, temperature, and oxy-gen concentration. Ferric iron has less pro-oxidant potential,but is sparingly insoluble and generally bio-unavailable. Itis currently unknown, whether the good tolerability of ironsucrose is because of the ferric status. Iron sucrose is adminis-tered intravenously and thus does not pass the inflamedbowel segments locally as oral products do. Comparativestudies are needed that compare the pro-oxidant effects offerric and ferrous compounds as well as of the oral and intra-venous administration route.


REFERENCES

1. Lennard-Jones JE. Classification of inflammatory boweldisease. Scand J Gastroenterol Suppl 1989; 170:2–6.

2. Gasche C, Scholmerich J, Brynskov J, D’Haens G, Hanauer SB,Irvine EJ, et al. A simple classification of Crohn’s disease: reportof the Working Party for the World Congresses of Gastroenter-ology, Vienna 1998. Inflamm Bowel Dis 2000; 6(1):8–15.

3. Picco MF, Goodman S, Reed J, Bayless TM. Methodologicpitfalls in the determination of genetic anticipation: the caseof Crohn disease. Ann Intern Med 2001; 134(12):1124–1129.

4. Gasche C, Reinisch W, Lochs H, Parsaei B, Bakos S, Wyatt J,et al. Anemia in Crohn’s disease. Importance of inadequateerythropoietin production and iron deficiency. Dig Dis Sci1994; 39(9):1930–1934.

5. Best WR, Becktel JM, Singleton JW. Rederived values of theeight coefficients of the Crohn’s Disease Activity Index (CDAI).Gastroenterology 1979; 77(4 Pt 2):843–846.

6. Vatner SF. Effects of hemorrhage on regional blood flow distri-bution in dogs and primates. J Clin Invest 1974; 54(2):225–235.

7. Oldenburg B, Koningsberger JC, Brge Henegouwen GP, VanAsbeck BS, Marx JJ. Iron and inflammatory bowel disease.Aliment Pharmacol Ther 2001; 15(4):429–438.

8. de Vizia B, Poggi V, Conenna R, Fiorillo A, Scippa L. Ironabsorption and iron deficiency in infants and children withgastrointestinal diseases. J Pediatr Gastroenterol Nutr 1992;14(1):21–26.

9. Bartels U, Pedersen NS, Jarnum S. Iron absorption and serumferritin in chronic inflammatory bowel disease. Scand JGastroenterol 1978; 13(6):649–656.

10. Gasche C, Reinisch W, Lochs H, Gangl A. Interleukin-6production in Crohn’s disease. Gastroenterology 1992; 103(3):1120–1121.

11. Shah A, Morgan G, Rose JD, Fifield R, Rhodes J. Platelet num-ber and size in relation to serum orosomucoid concentration inCrohn’s disease. Med Lab Sci 1989; 46(1):79–80.

736 Gasche

12. Kapsoritakis AN, Potamianos SP, Sfiridaki AI, KoukourakisMI, Koutroubakis IE, Roussomoustakaki MI, et al. Elevatedthrombopoietin serum levels in patients with inflammatorybowel disease. Am J Gastroenterol 2000; 95(12):3478–3481.

13. Gasche C, Dejaco C. Iron deficiency and thrombocytosis inIBD. Gastroenterology 2002; 122(4):A611.

14. Schreiber S, Howaldt S, Schnoor M, Nikolaus S, Bauditz J,Gasche C, et al. Recombinant erythropoietin for the treatmentof anemia in inflammatory bowel disease. N Engl J Med 1996;334(10):619–623.

15. Horina JH, Petritsch W, Schmid CR, Reicht G, Wenzl H, SillyH, et al. Treatment of anemia in inflammatory bowel diseasewith recombinant human erythropoietin: results in threepatients. Gastroenterology 1993; 104(6):1828–1831.

16. Gasche C, Dejaco C, Reinisch W, Tillinger W, Waldhoer T,Fueger GF, et al. Sequential treatment of anemia in ulcerativecolitis with intravenous iron and erythropoietin. Digestion1999; 60(3):262–267.

17. DesreumauxP,ErnstO,GeboesK,GambiezL,BerrebiD,Muller-Alouf H, et al. Inflammatory alterations in mesenteric adiposetissue in Crohn’s disease. Gastroenterology 1999; 117(1):73–81.

18. MacDonald TT, Hutchings P, Choy MY, Murch S, Cooke A.Tumour necrosis factor-alpha and interferon-gamma produc-tion measured at the single cell level in normal and inflamedhuman intestine. Clin Exp Immunol 1990; 81(2):301–305.

19. Isaacs KL, Sartor RB, Haskill S. Cytokine messenger RNAprofiles in inflammatory bowel disease mucosa detected bypolymerase chain reaction amplification. Gastroenterology1992; 103(5):1587–1595.

20. Reinecker HC, Steffen M, Witthoeft T, Pflueger I, Schreiber S,MacDermott RP, et al. Enhanced secretion of tumour necrosisfactor-alpha, IL-6, and IL-1 beta by isolated lamina propriamononuclear cells from patients with ulcerative colitis andCrohn’s disease. Clin Exp Immunol 1993; 94(1):174–181.

21. Mahida YR, Kurlac L, Gallagher A, Hawkey CJ. High circulat-ing concentrations of interleukin-6 in active Crohn’s diseasebut not ulcerative colitis. Gut 1991; 32(12):1531–1534.


22. Fiocchi C. Inflammatory bowel disease: etiology and pathogen-esis. Gastroenterology 1998; 115(1):182–205.

23. Faquin WC, Schneider TJ, Goldberg MA. Effect of inflamma-tory cytokines on hypoxia-induced erythropoietin production.Blood 1992; 79(8):1987–1994.

24. Swinson CM, Perry J, Lumb M, Levi AJ. Role of sulphasala-zine in the aetiology of folate deficiency in ulcerative colitis.Gut 1981; 22(6):456–461.

25. Laidlaw ST, Reilly JT. Antilymphocyte globulin for mesalazine-associated aplastic anaemia. Lancet 1994; 343(8903):981–982.

26. Abboudi ZH, Marsh JC, Smith-Laing G, Gordon-Smith EC.Fatal aplastic anaemia after mesalazine. Lancet 1994;343(8896):542.

27. Anttila PM, Valimaki M, Pentikainen PJ. Pure-red-cell aplasiaassociated with sulphasalazine but not 5-aminosalicylic acid.Lancet 1985; 2(8462):1006.

28. D’Haens G, Geboes K, Rutgeerts P. Endoscopic and histologichealing of Crohn’s (ileo-) colitis with azathioprine. Gastroint-est Endosc 1999; 50(5):667–671.

29. Black AJ, McLeod HL, Capell HA, Powrie RH, Matowe LK,Pritchard SC, et al. Thiopurine methyltransferase genotypepredicts therapy-limiting severe toxicity from azathioprine.Ann Intern Med 1998; 129(9):716–718.

30. Colombel JF, Ferrari N, Debuysere H, Marteau P, Gendre JP,Bonaz B, et al. Genotypic analysis of thiopurine S-methyl-transferase in patients with Crohn’s disease and severe myelo-suppression during azathioprine therapy. Gastroenterology2000; 118(6):1025–1030.

31. Dubinsky MC, Lamothe S, Yang HY, Targan SR, Sinnett D,Theoret Y, et al. Pharmacogenomics and metabolite measure-ment for 6-mercaptopurine therapy in inflammatory boweldisease. Gastroenterology 2000; 118(4):705–713.

32. Jobson B, Garza A, Sninsky CA. Red cell mean corpuscularvolume (MCV) correlates with 6 thioguanine nucleotide(6TG) levels during azathioprine or 6-MP therapy for Crohn’sdisease. Gastroenterology 2001; 120(5):A4.

738 Gasche

33. Plum P, Warburg E. Hematological changes, especially mega-locytic anemia, in regional ileitis. Acta Med Scand 1939;102:449–475.

34. Lehtinen M, Lehtinen E, Nikkila EA. Microangiopathic hemo-lytic anemia in ulcerative colitis: report of a case. Scand JGastroenterol 1968; 3(4):417–424.

35. Black AJ, Eisinger AJ, Loehry CA, Johnson GD. Ulcerativecolitis with autoimmune haemolytic anaemia. Br Med J1969; 2(648):31–32.

36. Bardana EJ Jr, Pirofsky B. Autoimmune hemolytic anemiaand ulcerative colitis: a multisystem immunodeficiency dis-ease? Int Arch Allergy Appl Immunol 1970; 37(3):325–336

37. Goldstone AH. Letter: autoimmune haemolytic anaemia inulcerative colitis. Br Med J 1974; 2(918):556.

38. Gorst DW, Leyland MJ, Delamore IW. Autoimmune haemoly-tic anaemia and ulcerative colitis. Postgrad Med J 1975;51(596):409–411.

39. Shashaty GG, Rath CE, Britt EJ. Autoimmune hemolyticanemia associated with ulcerative colitis. Am J Hematol1977; 3:199–208.

40. Goodacre RL, Ali MA, Vanderlinden B, Hamilton JD, CastelliM, Seaton T. Hemolytic anemia in patients receiving sulfasa-lazine. Digestion 1978; 17(6):503–508.

41. Altman AR, Maltz C, Janowitz HD. Autoimmune hemolyticanemia in ulcerative colitis: report of three cases, review ofthe literature, and evaluation of modes of therapy. Dig DisSci 1979; 24(4):282–285.

42. Shashaty GG. Hemolytic anemia and ulcerative colitis. Dig DisSci 1980; 25(2):154–155.

43. Wodzinski MA, Lawrence AC. Severe Coombs positive autoim-mune haemolytic anaemia associated with ulcerative colitis.Postgrad Med J 1985; 61(713):261–262.

44. Mechanick JI. Coombs’ positive hemolytic anemia followingsulfasalazine therapy in ulcerative colitis: case reports,review, and discussion of pathogenesis. Mt Sinai J Med1985; 52(8):667–670.


45. Basista MH, Roe DC. A case presentation of hemolytic anemiain ulcerative colitis and review of the literature. Am J Gastro-enterol 1986; 81(10):990–992.

46. Clements D, Carter C, Allan RN. Autoimmune haemolyticanaemia complicating ulcerative colitis. Br J Hosp Med 1988;40(1):72.

47. Gumaste V, Greenstein AJ, Meyers R, Sachar DB. Coombs-positive autoimmune hemolytic anemia in ulcerative colitis.Dig Dis Sci 1989; 34(9):1457–1461.

48. Martinez P, Lambert M, Vanheuverzwyn R. Coombs-positiveautoimmune hemolytic anemia complicating ulcerative colitis.A case report. Acta Gastroenterol Belg 1991; 54(1):34–37.

49. Veloso FT, Fraga J, Carvalho J, Dias LM, Salgado MC. Autoim-mune hemolytic anemia in ulcerative colitis. A case report withreviewof the literature. JClinGastroenterol1991; 13(4):445–447.

50. Iwama T, Imajo M, Mishima Y. Coombs-positive autoimmunehemolytic anemia in ulcerative colitis—a report of two cases.Jpn J Surg 1991; 21(3):341–343.

51. Breese E, Braegger CP, Corrigan CJ, Walker-Smith JA,MacDonald TT. Interleukin-2- and interferon-gamma-secret-ing T cells in normal and diseased human intestinal mucosa.Immunology 1993; 78(1):127–131.

52. Ramakrishna R, Manoharan A. Auto-immune haemolytic anae-mia in ulcerative colitis. Acta Haematol 1994; 91(2):99–102.

53. Hernandez F, Linares M, Ferrer L, Cuquerella J, Sanchez H,Tome A, et al. Auto-immune haemolytic anaemia in ulcerativecolitis: report of three cases. Acta Haematol 1994; 91(4):213–214.

54. Collazzo JA, Siddiqui MA, Deren JJ. Concomitant onset ofulcerative colitis and autoimmune hemolytic anemia in an83-yr-old woman. Am J Gastroenterol 1996; 91(10):2236–2237.

55. Murphy PT, Cunney R, Nolan A, O’Donnell JR. Autoimmunehaemolytic anaemia associated with ulcerative colitis. Ir MedJ 1996; 89(5):172–173.

56. Arner O, Brohult J, Engstedt L, Karlson R, Sallstrom T. Auto-immune hemolytic anemia in ulcerative colitis, cured by colect-omy. Report of a case. Acta Med Scand 1971; 189(4):275–277.

740 Gasche

57. Yates P, Macht LM, Williams NA, Elson CJ. Red cell autoanti-body production by colonic mononuclear cells from a patientwith ulcerative colitis and autoimmune haemolytic anaemia.Br J Haematol 1992; 82(4):753–756.

58. Giannadaki E, Potamianos S, Roussomoustakaki M, KyriakouD, Fragkiadakis N, Manousos ON. Autoimmune hemolyticanemia and positive Coombs test associated with ulcerativecolitis. Am J Gastroenterol 1997; 92(10):1872–1874.

59. Eng C, Farraye FA, Shulman LN, Peppercorn MA, KraussCM, Connors JM, et al. The association between the myelodys-plastic syndromes and Crohn disease. Ann Intern Med 1992;117(8):661–662.

60. Harewood GC, Loftus EVJ, Tefferi A, Tremaine WJ, SandbornWJ. Concurrent inflammatory bowel disease and myelodys-plastic syndromes. Inflamm Bowel Dis 1999; 5(2):98–103.

61. Dunn AM, Kerr GD. Pure red cell aplasia associated withsulphasalazine. Lancet 1981; 2(8258):1288.

62. Sheehan RG, Necheles TF, Lindeman RJ, Meyer HJ, Patter-son JF. Regional enteritis and granulomatous colitis asso-ciated with erythrocyte glucose-6-phosphate dehydrogenasedeficiency. N Engl J Med 1967; 277(21):1124–1126.

63. Schreiber S, Wedel S. Diagnosis and treatment of anemia ininflammatory bowel disease. Inflamm Bowel Dis 1997;3(3):204–216.

64. Thomson AB, Brust R, Ali MA, Mant MJ, Valberg LS. Irondeficiency in inflammatory bowel disease. Diagnostic efficacyof serum ferritin. Am J Dig Dis 1978; 23(8):705–709.

65. Hansen TM, Hansen NE, Birgens HS, Holund B, Lorenzen I.Serum ferritin and the assessment of iron deficiency in rheu-matoid arthritis. Scand J Rheumatol 1983; 12(4):353–359.

66. Gasche C, Waldhoer T, Feichtenschlager T, Male C, Mayer A,Mittermaier C, et al. Prediction of response to iron sucrose ininflammatory bowel disease-associated anemia. Am J Gastro-enterol 2001; 96(8):2382–2387.

67. Gasche C. Anemia in IBD: the overlooked villain. InflammBowel Dis 2000; 6(2):142–150.


68. Erichsen K, Hausken T, Ulvik RJ, Svardal A, Berstad A, BergeRK. Ferrous fumarate deteriorated plasma antioxidant statusin patients with Crohn disease. Scand J Gastroenterol 2003;38(5):543–548.

69. Oldenburg B, Berge Henegouwen GP, Rennick D, Van AsbeckBS, Koningsberger JC. Iron supplementation affects theproduction of pro-inflammatory cytokines in IL-10 deficientmice. Eur J Clin Invest 2000; 30(6):505–510.

70. Harvey RS, Reffitt DM, Doig LA, Meenan J, Ellis RD, Thomp-son RP, et al. Ferric trimaltol corrects iron deficiency anemiain patients intolerant of iron. Aliment Pharmacol Ther 1998;12(9):845–848.

71. Geisser P, Baer M, Schaub E. Structure=histotoxicity relation-ship of parenteral iron preparations. Arzneimittelforschung1992; 42(12):1439–1452.

72. Gasche C, Dejaco C, Waldhoer T, Tillinger W, Reinisch W,Fueger GF, et al. Intravenous iron and erythropoietin for ane-mia associated with Crohn disease. A randomized, controlledtrial. Ann Intern Med 1997; 126(10):782–787.

73. Schroder O, Schrott M, Blumenstein I, Jahnel J, Dignass AU,Stein J. A study for the evaluation of safety and tolerability ofintravenous high-dose iron sucrose in patients with iron defi-ciency anemia due to gastrointestinal bleeding. Z Gastroen-terol 2004; 42(8):663–667.

74. Chandler G, Harchowal J, Macdougall IC. Intravenous ironsucrose: establishing a safe dose. Am J Kidney Dis 2001;38(5):988–991.

75. Sandborn WJ, Targan SR. Biologic therapy of inflammatorybowel disease. Gastroenterology 2002; 122(6):1592–1608.

76. Papadaki HA, Kritikos HD, Valatas V, Boumpas DT, Eliopou-los GD. Anemia of chronic disease in rheumatoid arthritis isassociated with increased apoptosis of bone marrow erythroidcells: improvement following anti-tumor necrosis factor-alphaantibody therapy. Blood 2002; 100(2):474–482.

77. Sandborn WJ, Hanauer SB, Katz S, Safdi M, Wolf DG, BaergRD, et al. Etanercept for active Crohn’s disease: a randomized,

742 Gasche

double-blind, placebo-controlled trial. Gastroenterology 2001;121(5):1088–1094.

78. D’Haens G, van Deventer S, Van Hogezand R, Chalmers D,Kothe C, Baert F, et al. Endoscopic and histological healingwith infliximab anti-tumor necrosis factor antibodies inCrohn’s disease: a European multicenter trial. Gastroenterol-ogy 1999; 116(5):1029–1034.

79. Means RTJ, Krantz SB. Inhibition of human erythroidcolony-forming units by gamma interferon can be correctedby recombinant human erythropoietin. Blood 1991; 78(10):2564–2567.

80. Demirturk L, Hulagu S, Yaylaci M, Altin M, Ozel M. Serumerythropoietin levels in patients with severe anemia second-ary to inflammatory bowel disease and the use of recombi-nant human erythropoietin in patients with anemiarefractory to treatment [letter]. Dis Colon Rectum 1995;38(8):896–897.

81. Dohil R, Hassall E, Wadsworth LD, Israel DM. Recombinanthuman erythropoietin for treatment of anemia of chronicdisease in children with Crohn’s disease. J Pediatr 1998;132(1):155–159.

82. Gasche C, Lomer MC, Cavill I, Weiss G. Iron, anemia, andinflammatory bowel diseases. Gut 2004; 53(8):1190–1197.

83. Crawford J, Cella D, Cleeland CS, Cremieux PY, Demetri GD,Sarokhan BJ, et al. Relationship between changes in hemoglo-bin level and quality of life during chemotherapy in anemiccancer patients receiving epoetin alfa therapy. Cancer 2002;95(4):888–895.

84. Lih-Brody L, Powell SR, Collier KP, Reddy GM, Cerchia R,Kahn E, et al. Increased oxidative stress and decreased antiox-idant defenses in mucosa of inflammatory bowel disease. DigDis Sci 1996; 41(10):2078–2086.

85. Shaheen NJ, Silverman LM, Keku T, Lawrence LB, RohlfsEM, Martin CF, et al. Association between hemochromatosis(HFE) gene mutation carrier status and the risk of coloncancer. J Natl Cancer Inst 2003; 95(2):154–159.


Index

a-1-Antitrypsin, 567Absorption, defective, 384ACD See Anemia of chronic

disease.Achronic lymphatic

leukemia, 507ACI See Anemia of chronic

inflammation.Acidosis, metabolic, 678ACRD See Anemia of chronic

renal disease.Acute hemolytic transfusion

reaction, 406Acute lung injury, transfusion-

related, 405Acute renal failure, 678Adjuvant chemotherapy,

conventional, 529Adult respiratory distress

syndrome, 405AI See Anemia of inflammation.Allogeneic blood transfusion, 399

Allogeneic RBCs, buffy-coat-reduced, 421

Alveolar hemorrhage, 385Amenorrhea, 494Anaphylactic reactions, 391Androgen therapy, 165, 675Anemia, 73, 157, 229, 373, 437,

607, 621, 671, 728degree, 458drug-associated, 730of cancer, chronic, 490of chronic disease, 6, 134, 145,

206, 233, 335, 368, 381,443, 490, 593, 729

of chronic inflammation, 335of chronic renal disease, 674of inflammation, 615pernicious, 350sickle cell, 204

Angiogenesis, 560Angiotensin converting enzyme

inhibitors, 694

745

Ankylosing spondylitis, 636Antimicrobial activity, 231Aplasia, 730Apoptosis, erythroid cell, 368Apoptosis, 93, 206Arrhythmia, cardiac , 403Arthalgias, 392Arthritis, rheumatoid, 128,

372, 450Arthropathy, 255Autoantibodies, 638Autoimmune diseases, 565Autologous stem cell

transplantation, 154Azidothymidine antiviral

therapy, 157

BaCon study, 411Basolateral membrane, 106Bilirubin, 211Biliverdin, 211Bloodcell transfusions, 399, 569cells, white, 402loss, 383pressure, 693

Blunted erythropoietin, 660B-lymphocytes, 595Bone marrow, 283Bovine spongiform

encephalopathy, 415Bowel disease, inflammatory,

384, 727Buffy-coat-reduced allogeneic

RBCs, 421

Carcinoma, cervical, 441Carcinomas, 534Cardiac arrhythmia, 403Cardiomyopathy, 255Celiac disease, 384Cellapoptosis, erythroid, 368arteritis, giant, 641

[Cell]differentiation, erythroid, 33transfusions, blood, 381,

399, 569transfusions, red blood, 499, 660

Cerebral hypoxia, 495Cervical carcinoma, 441Chagas’, disease, 409Chemotherapy, cytotoxic , 490Chemotherapy, noncisplatin, 162Chemotherapy, 146, 172

cisplatin, 169Chronic

anemia of cancer, 490arthritis, juvenile, 160hemodialysis, 439infection, 158inflammatory

conditions, 566inflammatory rheumatic

diseases, 633lymphatic leukemia, 507myeloid disorders, 164renal failure, 148renal insufficiency, 671

Colony-forming units, 131erythroid , 661

Colorectal adenocarcinoma, 608Congestive heart disease, 438, 493Conventional adjuvant

chemotherapy, 529Creutzfeldt–Jakob disease, 407Crypt cells, intestinal, 107Cyclophosphamide, 168Cystic fibrosis, 7Cytokines, 111, 130, 146, 282

signaling, 33inflammatory, 660

Cytomegalovirus, 399Cytotoxic chemotherapy, 490

Darbopoietin alfa, 465, 505, 599Dcytb, 243Defective absorption, 384Deferiprone, 643

746 Index

Deficiency, Iron, 31, 69, 383,458, 732

Degree of anemia, 458Dermatomyositis, 634Desferroxamine, 561Dextran, iron (INFeD), 391Divalent metal transporter 1, 9Divalent-cation-sensitive

fluorophore, 109Drug-associated anemia, 730Dysfunction, neutrophil, 563Dyspnea, 493

Echinocytes, 685Edema, pulmonary, 493Effete erythrocytes, 110End stage renal disease, 440, 671Epithelial cell layer, 105Epoetin alpha, 599Epoietin alfa therapy, 463EPO

plasma levels, 149serum levels, 147production, 661receptor, 71

EpoR status, 575Erythrocytes, effete, 110

hypochromic, 639ferritin, 639

Erythrocytosis, 72, 201, 336Erythroid

burst forming unit, 661cell apoptosis, 368cell differentiation, 33colony-forming unit, 661progenitor cells, 336, 584progenitors, 134

Erythron, 155Erythrophagocytosis, 114, 202,

566, 686Erythropoiesis, 13, 73, 130, 146,

356, 369, 729level, 233effect, 461response, 445

Erythropoietin, 146, 236, 381, 398,504, 599, 672

blunted, 660endogenous, 368recombinant , 568therapy, 451, 573expression, 238

Fas-ligand, 420Febrile, nonhemolytic transfusion

reactions, 404Ferric hydroxide complex

(VENOFER), 391Ferritin, 22, 202, 258, 283, 455,731serum, 386

Ferroportin 1, 283Ferroprotein, 107Fluorophore, divalent-cation-

sensitive, 109Formation of heme, 572

Gallop rhythms, 492Glial cells, 260Glomerular filtration rate, 674Glycosylation, 599Granulocyte, neutrophil, 563

Haptoglobin, 272Heartdisease, congestive, 493disease, ischemic, 439failure, congestive, 438

Hematocrit, 398, 462, 660, 728Hematological malignancies, 497Hematuria, 273heme catabolism, 209

Heme, 2,209formation, 572iron, 285proteins, 209

Hemochromatosis, 12, 202, 444hereditary, 7, 107, 295protein, 107secondary, 500

Index 747

Hemodialysis, chronic, 439Hemodilution, 146Hemoglobin, 202, 285, 340, 439,

457concentration, 565, 621levels, 373,608

Hemoglobinuria, 385Hemolysis, 598,623hemolytic, 236hypoproliferative, 338

Hemolyticautoimmune anemia, 637anemia, 236

Hemopoietin, 675Hemoptysis, 384Hemorrhage, alveolar, 385Hemosiderin, 202, 262Hemosiderosis, pulmonary, 385Hemostasis, 688Hepatectomy, 154Hepatitis A virus, 409Hepatitis B virus, 407Hepatitis C virus, 407Hepatitis G virus, 409Hepatocellular damage, 386Hepatocytes, 107Hepatoma, 271Hepatonecrosis, 273Hepcidin, 15, 116, 218, 230, 262,

284, 566, 625Hepcidinuria, 234Hephaestin mutation, 243Hereditary hemochromatosis,

7, 107, 295Herpesvirus, human, 8, 409Heteropolymers, 272HFE genotype, 239HIV infection, 397, 620HIV-seronegative, 157HIV-seropositive, 157Homeostasis, iron, 105,282Humanbone marrow, 681EPO, 62EPO, recombinant, 63, 660

[Human]erythropoietin (rHuEpo)

therapy, recombinant,147, 376, 494, 571, 671

herpesvirus, 8,409immunodeficiency

virus. See HIVleukocyte antigen, 405T-cell lymphotropic viruses I

and II, 407Hydroxide complex (VENOFER),

ferric, 391Hyperferremia, 274Hyperferritinemia, 111Hypernephroma, 574Hyperoxia, 567Hyperparathyroidism, 682Hyperphosphatemia, 682Hypersplenism, 206Hypertrophy, left ventricular,

443Hypocalcemia, 403Hypochromia, 382, 731Hypochromic erthyrocytes, 639Hypoferremia, 382, 594Hypoferremic, 258Hypophosphatemia, 685Hypoproliferative anemia, 338Hypothermia, 403Hypoxia, 148, 236, 369, 492,

560, 675cerebral, 495tissue, 61

IL-1, 91Immune cells, 563Immunological

molecules, 282system, 281

Inducible nitric oxidesynthase, 561

Infection, chronic, 158Infection, 88InFed, 460

748 Index

Inflammatorybowel disease, 384,727chronic conditions, 566chronic rheumatic diseases, 633cytokines, 660

Infliximab, 733Innate immunity, 282, 561Interferons, 132Interleukin-1, 132,662Intestinal crypt cells, 107IRE=IRP system, 27Iron, 368, 560

chelation, 561deficiency anemia,

153, 339, 368, 382, 728deficiency, functional, 507deficiency, true, 568deficiency, 31, 69, 383, 458, 732Dextran (INFeD), 391homeostasis, 105, 282loading anemia, 338metabolism, 15, 109, 230,

282, 367release, 294responsive protein, 215, 216sucrose preparations, 391therapy, 388turnover, plasma, 444

Iron-binding capacity, total, 597Iron-depleted cells, 108Irradiation, sublethal, 168Irregular menstrual cycles, 494Ischemic heart disease, 439Isoelectric point, 63(Hyper) fibrinolysis, 661

Kaposi’s sarcoma, 622Kidney, polycystic, 677

Lactoferrin, 117, 240, 257, 283Left ventricular hypertrophy, 443Leukemia and lymphoma, 165Leukemia cell, 267

Leukemia, 165Leukocytes, polymorphonuclear,

687Leukopenia, 730Leukopoiesis, 674Lipophosphoglycan, 268Loading anemia, iron, 338Low serum ferritin concentration,

386Lymphatic leukemiaachronic, 507chronic, 507

Lymphocytes, 405Lymphoma, non-Hodgkin’s, 594Lymphoma, 165

M. tuberculosis, 617Macrocytosis, 730Macrophages, 109, 561Malabsorption, 384, 494Malaria, 623Malignancy, 373, 382, 490, 578hematological, 497lymphoproliferative, 598nonmyeloid, 527nonhematologic, 594

Malignant cells, 575Marrow dysfunction, 387Mean corpuscular volume, 738Megakaryocytopoiesis, 674Megakaryopoiesis, 729Megaloblastic signs, 664Melanotransferrin, 267Membrane, basolateral, 106Menorrhagia, 494Menstrual cycles, irregular, 494Metabolic acidosis, 678Metabolism, iron, 15, 109, 230,

282, 367Microcytosis, 382,731Monocyte=macrophages, 283Monocytes, 207Mononuclear phagocyte

system, 201Multiple myeloma, 507

Index 749

Multiple organ dysfunctionsyndrome, 88

Multiple organ failure, 89Myalgias, 392Myelodysplastic syndromes,

164, 521Myelofibrosis, 682Myeloid disorders, chronic, 164Myeloid growth factors, 521Myeloma, multiple, 507Myeloproliferative disorders, 594Myocardial infarction, 462Neoplasia, 255Neopterin, 148, 312, 564Nephrectomy, 154Neuroferritinopathy, 23Neutrophil granulocyte, 563Neutrophils, polymorphonuclear,

257

Nile virus, West, 408Nocturnal hemoglobinuria,

paroxysmal, 594Noncisplatin chemotherapy, 162Nonhematologic malignancy, 594Nonhemolytic transfusion

reactions, febrile, 404Non-Hodgkin’s lymphoma, 594Nonhypoxemic adults, 661Nonmyeloid malignancies, 527Normocytic, 387, 674NRAMP, 11

Odds ratio of mortality, 440of ACD, 644of chronic

inflammation, 335of chronic disease, 6, 134, 145,

206, 233, 335, 368, 381,443, 490, 593, 729

of chronic renaldisease, 674

Organelles, intracellular, 115Osteitis fibrosa, 683Ovotransferrin, 257

Parathyroidectomy, 683Paroxysmal nocturnal

hemoglobinuria, 594Pathogenesis, 145Pathogens, 116Pernicious anemia, 350Phagocyte system,

mononuclear, 201Phagocytosis, 110Phagolysosome, 209Phlebotomy, blood, 447Phlebotomy, 660Pica, 385Plasma Epo levels, 149Plasma transferrin, 107Plasma viscosity, 167Point isoelectric, 63Polycythemia, 150, 677Polymyalyia rheumatica, 641Polymyositis, 634Progenitor cells, erythroid, 134,

336, 584Pulmonary edema, 493Pulmonary hemosiderosis, 385Pure red cell aplasia, 508Purpura, post-transfusion, 406

Quality of life, 494

Radiation therapy, 530Radioimmunoassays, 161Radiotherapy, 146, 490Ratio of mortality, odds, 440Reactive oxygen intermediates, 4Recombinant erythropoietin, 568Recombinant human

erythropoietin (rHuEpo)therapy, 147, 376, 494,571, 671

Recombinant humanEpo, 63, 660

Red blood cell, 61, 202, 351Red blood cell expansion, 445

750 Index

Red cell aplasia, pure, 508Red cell mass, 72, 375, 383Renal disease, end stage,

440, 671Renal failure, acute, 678Renal failure, chronic, 148Renal failure, 682Renal insufficiency, chronic, 671Renal tubular dysfunction, 173

Resistance associated macrophageprotein 1 (NRAMP-1),natural, 561

Respiratory distress syndrome,Adult, 405

responsive elements, 108Reticulocyte counts, 457Reticuloendolial system, 560Rheumatoid arthritis, 128,

372, 450

Scleroderma, 634Secondary hemochromatosis,

500Sepsis, 87Sertoli cells, 260Serum concentrations, 337Serum Epo levels, 147Serum ferritin concentration, low,

386serum ferritin, 22,455

Serum ferritin, 386Severe acute respiratory

syndrome, 407severe, 608

Sickle cell anemia, 204Siderophilins, 257Siderophores, 117, 260Slight proteinuria, 494Soluble transferrin receptor, 596Spleen, 203Spondylitis, ankylosing, 636

Spongiform encephalopathy,bovine, 415

Stem cell factor, 134Sublethal irradiation, 168Sulfasalazine, 730syndromes, 93

Plasma iron turnover, 444chemotherapy, 172

Systemic inflammatory responsesyndrome, 87

Systemic lupus erthematosus,634

Tachycardia, 492Tachyphylaxis, 689TfR-F Index, 355Th-1-mediated immune

pathways, 382Thrombocytopenia, 407, 688Thrombocytosis, 729Tissue hypoxia, 61TNF and IL-1, 91TNF, 91Total iron-binding capacity, 597toxicity, 3of inflammation, 615

Transferrin and transferriniron, 639

Transferrin iron, 639Transferrin, plasma, 107Transferrin receptor, 287Transferrin, 9, 215, 257, 386, 616Transfusion, 571Transfusion-related acute lung

injury, 405Transketolase, 685True, Iron deficiency, 568Tuberculosis, 616Tubular cysts, proximal, 677Tumor cells, 560Tumor interaction, 491Tumor necrosis factor, 131, 595Tumor necrosis factor-a, 336,

560, 662

Index 751

Units, burst-forming, 131Units, colony-forming, 131Uremia, 679

Vampirism, 660Vascular endothelial growth factor

(VEGF), 560

Vitamin deficiencies, 731

West Nile virus, 408White blood cells, 402

Zidovudine therapy, 157Zymosan particles, 109, 110

752 Index

Date post:	29-Nov-2014
Category:	Documents
Upload:	bang-juntak
View:	337 times
Download:	0 times

Anemia

Documents