CONTENTS American Journal of Kidney Diseases

VOL 47, NO 5, SUPPL 3, MAY 2006

CONTENTS American Journal of AJKD Kidney Diseases

KDOQI Clinical Practice Guidelines and Clinical Practice Recommendations for Anemia in Chronic Kidney Disease Tables...... S1 Figures ...... S2 Abbreviations and Acronyms...... S4 Work Group Membership ...... S7 KDOQI Advisory Board Members...... S8 Foreword...... S9 I. EXECUTIVE SUMMARY S11 II. CLINICAL PRACTICE GUIDELINES AND CLINICAL PRACTICE RECOMMEN- DATIONS FOR ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S16 1.1. Identifying Patients and Initiating Evaluation ...... S17 1.2. Evaluation of Anemia in CKD...... S28 2.1. Hb Range ...... S33 3.1. Using ESAs ...... S54 3.2. Using Iron Agents...... S58 3.3. Using Pharmacological and Nonpharmacological Adjuvants to ESA Treatment in HD-CKD...... S71 3.4. Transfusion Therapy ...... S79 3.5. Evaluating and Correcting Persistent Failure To Reach or Maintain Intended Hb...... S81 III. CLINICAL PRACTICE RECOMMENDATIONS FOR ANEMIA IN CHRONIC KID- NEY DISEASE IN CHILDREN S86 1.1: Identifying Patients and Initiating Evaluation ...... S87 1.2: Evaluation of Anemia in CKD...... S89 2.1: Hb Range ...... S90 3.1: Using ESAs ...... S93 3.2: Using Iron Agents...... S99 3.3: Using Pharmacological and NonPharmacological Adjuvants to ESA Treatment in HD-CKD...... S105 3.4: Transfusion Therapy ...... S106 3.5: Evaluating and Correcting Persistent Failure To Reach or Maintain Intended Hb Level ...... S107 IV. CLINICAL PRACTICE RECOMMENDATIONS FOR ANEMIA IN CHRONIC KID- NEY DISEASE IN TRANSPLANT RECIPIENTS S109 V. APPENDIX 1: METHODS OF EVIDENCE REVIEW AND SYNTHESIS ...... S117 Work Group Biographies...... S126 Acknowledgments ...... S131 References...... S132 KDOQI Disclaimer SECTION I: USE OF THE CLINICAL PRACTICE GUIDELINES AND CLINICAL PRACTICE RECOMMENDATIONS

These Clinical Practice Guidelines (CPGs) and Clinical Practice Recommendations (CPRs) are based upon the best information available at the time of publication. They are designed to provide information and assist decision-making. They are not intended to deﬁne a standard of care, and should not be construed as one. Neither should they be interpreted as prescribing an exclusive course of management. Variations in practice will inevitably and appropriately occur when clinicians take into account the needs of individual patients, available resources, and limitations unique to an institution or type of practice. Every health-care professional making use of these CPGs and CPRs is responsible for evaluating the appropriateness of applying them in the setting of any particular clinical situation. The recommendations for research contained within this document are general and do not imply a speciﬁc protocol. SECTION II: DEVELOPMENT PROCESS

The National Kidney Foundation makes every effort to avoid any actual or potential conflicts of interest that may arise as a result of an outside relationship or a personal, professional or business interest of a member of the Work Group. Specifically, all members of the Work Group are required to complete, submit and sign a Disclosure Questionnaire showing all such relationships that might be perceived as real or potential conflicts of interest. All affiliations are published in their entirety at the end of this publication in the Biographical Sketch section of the Work Group members.

In citing this document, the following format should be used: National Kidney Foundation. KDOQI Clinical Practice Guidelines and Clinical Practice Recommendations for Anemia in Chronic Kidney Disease. Am J Kidney Dis 47:S1-S146, 2006 (suppl 3).

Support for the development of the KDOQI Clinical Practice Guidelines and Clinical Practice Recommendations for Anemia in Chronic Kidney Disease was provided by Amgen, Inc.

The National Kidney Foundation gratefully acknowledges the support of Amgen, Inc. as the founding and principal sponsor of KDOQI. Tables

Table 1. Key Differences Between Current Guidelines (KDOQI Anemia 2006) and Previous Anemia Guidelines (KDOQI 2000 and EBPG 2004) ...... S15 Table 2. Relationship Between Hb Level and Kidney Function ...... S18 Table 3. Relationship Between Hct and Kidney Function...... S19 Table 4. Prevalence of Anemia by Level of Kidney Function...... S20

Table 5. Difference in Hb and Hct From Reference by Category of CCr...... S20 Table 6. Difference in Hb and Hct From Reference According to Category of BSA and Normalized eGFR ...... S21 Table 7. Hb for Males by Race/Ethnicity and Age: United States, 1988 to 1994...... S24 Table 8. Hb for Females by Race/Ethnicity and Age: United States, 1988 to 1994 ...... S25 Table 9. Normal Increase in Hb Levels Related to Long-Term Altitude Exposure...... S26 Table 10. Maximum Hb Cutoff Values for Anemia in Pregnancy...... S27 Table 11. Adjustment to Hb for Smoking by Number of Packets per Day...... S27 Table 12. RCTs Examining Effects of Distinct Hb Targets/Levels on Key Clinical Outcomes in the HD-CKD and PD-CKD Populations ...... S34 Table 13. RCTs Examining Effects of Distinct Hb Targets/Levels on QOL in the HD-CKD and PD-CKD Populations...... S36 Table 14. Non-CVD/Mortality AE Rates in RCTs Examining Distinct Hb Targets/Levels in HD-CKD and PD-CKD Populations: ESA versus ESA...... S38 Table 15. Non-CVD/Mortality AE Rates in RCTs Examining Distinct Hb Targets/Levels in the HD-CKD and PD-CKD Populations: ESA versus Placebo ...... S40 Table 16. Effects of Distinct Hb Targets/Levels on Key Clinical Outcomes in the ND-CKD Population ...... S41 Table 17. RCTs Examining Effects of Distinct Hb Targets/Levels on QOL in the ND-CKD Population ...... S42 Table 18. Non-CVD/Mortality AE Rates in RCTs Examining Distinct Hb Targets/Levels in the CKD Population ...... S43 Table 19. RCTs Examining Effects of Distinct Hb Targets/Levels on Exercise Capacity in the HD-CKD and PD-CKD Populations ...... S44 Table 20. Target Hb Levels in the HD-CKD and PD-CKD Populations...... S45 Table 21. Target Hb Levels in the ND-CKD Population...... S48 Table 22. Anemia Management Protocol Information for Initial Anemia Management Used in ESA RCTs ...... S56 Table 23. RCTs for Ferritin and TSAT Targets in the HD-CKD Population ...... S60 Table 24. Value of Baseline Ferritin for Assessing Likelihood of Response to IV Iron Therapy in the HD-CKD Population ...... S61 Table 25. Value of Ferritin in Assessing Iron Storage Excess or Deficiency...... S61 Table 26. Value of Baseline TSAT for Assessing Likelihood of Response to IV Iron Therapy in the HD-CKD Population ...... S62 Table 27. Value of Baseline CHr for Assessing Likelihood of Response to IV Iron Therapy...... S62 Table 28. Mean Monthly Dose of Iron in Adult Prevalent HD-CKD Patients on ESA Therapy.... S63 Table 29A. Comparative RCTs of IV versus PO Iron Administration and Efficacy of Anemia Management in the HD-CKD and PD-CKD Populations...... S65 Table 29B. Comparative RCTs of IV versus PO Iron Administration and Efficacy of Anemia Management in the ND-CKD Population ...... S65 Table 30. Comparative RCTs of PO Iron Administration versus Placebo/Control and Efficacy of Anemia Management in the HD-CKD and PD-CKD Populations ...... S66 Table 31. Serious AEs of Iron Agents in Patients Naïve to Tested Iron Agent (N Ͼ 100) ...... S66 Table 31A. IV versus PO Iron Agents in the HD-CKD and PD-CKD Populations ...... S67

American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S1-S3 S1 S2 FIGURES

Table 31B. IV versus PO Iron Agents in the ND-CKD Population ...... S67 Table 32. RCTs Evaluating Effects of Treatment with IV L-Carnitine on Hb Levels and ESA Doses in the HD-CKD Population...... S72 Table 33. Use of L-Carnitine as an Adjuvant to ESA Treatment in the HD-CKD Population...... S73 Table 34. RCTs Evaluating Effects of Treatment With IV Ascorbic Acid on Hb Levels and ESA Doses in the HD-CKD Population ...... S74 Table 35. Use of Ascorbic Acid as an Adjuvant to ESA Treatments in the HD-CKD Population...... S74 Table 36. RCTs Evaluating Effects of Treatment With IM Androgens on Hb Levels in the HD-CKD Population ...... S75 Table 37. Use of Androgens as an Adjuvant to ESA Treatment in the HD-CKD Population...... S76 Table 38. Published Experience in Patients With Anemia, CKD, and a Preexisting Hematologic Disorder ...... S84 Table 39. Hb Levels (g/dL) in Children Between 1 and 19 Years for Initiation of Anemia Workup...... S88 Table 40. Hb Levels (g/dL) in Children Between Birth and 24 Months for Initiation of Anemia Workup...... S88 Table 41. Literature Review of Anemia in Transplant CKD: Key Conclusions...... S110 Table 42. Prevalance of PTA by Duration of Posttransplantation Period ...... S112 Table 43. Example of a Summary Table...... S119 Table 44. Systematic Review Topics and Screening Criteria...... S120 Table 45. Literature Search Yield of Primary Articles for Systematic Review Topics...... S121 Table 45A. Assessment of Study Applicability...... S121 Table 46. Details of First-Look Topics, Ovid Literature Searches, and Yield by Topic ...... S122 Table 46A. Grades for Study Quality...... S122 Table 47. Evaluation of Studies of Prevalence...... S123 Table 48. Format for Guidelines...... S123 Table 49. Balance of Benefit and Harm ...... S124 Table 50. Definitions of Grades for Quality of Overall Evidence ...... S124 Table 51. Example of an Evidence Profile ...... S125 Figures

Figure 1. Kidney Failure in the United States ...... S12 Figure 2. Relationship Between Current and Previous Anemia Guidelines...... S14 Figure 3. Relationship Between Hb Level and eGFR ...... S21 Figure 4. Relationship Between eGFR and Prevalence of Anemia as Deﬁned by Differing Hb Levels for (A) Males and (B) Females...... S21 Figure 5. Prevalence of Anemia by CKD Stage ...... S22

Figure 6. Relationship Between Level of Renal Function, Reﬂected by SCr Level or GFR, and Prevalence of Anemia, Deﬁned at Different Hb Cutoff Levels, Among Patients under the Care of Physicians ...... S22 Figure 7. Hb Percentiles by GFR in the Canadian Multicentre Longitudinal Cohort Study ...... S22 Figure 8. Prevalence of Low Hb Level by Category of GFR in the Canadian Multicentre Longitudinal Cohort Study...... S23 Figure 9. Prevalence of Low Hb Level by eGFR in Patients with Diabetics Compared with the General Population ...... S23 Figure 10. Changes in Hb Levels and eGFR Over 2 years in Adult Patients with CKD Stages 3 and 4 Not Treated with ESA ...... S23 Figure 11. Distribution of Hb Levels in Adult Males and Females by Age Group...... S26 Figure 12. Distribution of Ferritin and TSAT Values by Age Group in Males and Females ...... S30 FIGURES S3

Figure 13. Distribution of Hb Levels by Ferritin Range in Patients with CKD ...... S31 Figure 14. Distribution of Hb Levels by TSAT Range in Patients with CKD...... S31 Figure 15. Target and Achieved Hb Levels in 18 RCTs Comparing Higher Against Lower (or Placebo/Control) Hb Targets for ESA Therapy ...... S50 Figure 16. Baseline Mean Pretreatment Hb Levels Compared with Achieved Mean Hb Levels in the Upper and Lower Target...... S51 Figure 17. Exposure to Eprex¨ and Case Counts of PRCA ...... S82 Figure 18. Mean Epoetin Dose per Patient per Administration by Percentile of Dose (1st, 5th to 95th, and 99th) ...... S82 Figure 19. Process of Triaging a Topic to a Systematic Review or a Narrative Review ...... S118 Acronyms and Abbreviations ⌬ Change 1 Increase 2 Decrease ABP Ambulatory blood pressure ABPM Ambulatory blood pressure monitoring ACE Angiotensin-converting enzyme ADE Adverse drug event AE Adverse event ARBs Angiotensin receptor blockers ARF Acute renal failure AST Aspartate transaminase BID Twice daily BIW Twice weekly BP Blood pressure BSA Body surface area CABG Coronary artery bypass graft CBC Complete blood count

CCr Creatinine clearance CESDS Center for Epidemiologic Studies Depression Scale CFU-E Erythroid colony-forming units CHF Congestive heart failure CHOIR Correction of Hemoglobin and Outcomes in Renal Insufficiency CHr Content of hemoglobin in reticulocytes CI Confidence interval CKD Chronic kidney disease CMV Cytomegalovirus CPG Clinical practice guideline CPR Clinical practice recommendation CREATE Cardiovascular Risk Reduction by Early Anemia Treatment With Epoetin Beta Trial CRI Corrected reticulocyte index CV Cardiovascular CVA Cerebrovascular accident CVD Cardiovascular disease D/C Discontinuation DBP Diastolic blood pressure DNA Deoxyribonucleic acid DOQI Dialysis Outcomes Quality Initiative EBPG European Best Practices Guideline eGFR Estimated glomerular filtration rate ERI Erythropoietin resistance index ERT Evidence Review Team ESA Erythropoiesis-stimulating agent ESRD End-stage renal disease FACIT Functional Assessment of Chronic Illness Therapy FDA Food and Drug Administration GFR Glomerular filtration rate GI Gastrointestinal Hb Hemoglobin Hct Hematocrit HD Hemodialysis

S4 American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S4-S6 ACRONYMS AND ABBREVIATIONS S5

HD-CKD Hemodialysis-dependent chronic kidney disease HDF Hemodiafiltration HPS Hemophagocytic syndrome HTN Hypertension HU Hydroxyurea HUI Health Utilities Index HUS Hemolytic uremic syndrome IHD Ischemic heart disease IM Intramuscular IU International Unit IV Intravenous KDIGO Kidney Disease: Improving Global Outcomes KDOQI Kidney Disease Outcomes Quality Initiative KDQ Kidney Disease Questionnaire KDQOL Kidney Disease Quality of Life KEEP Kidney Early Evaluation Program KLS Kidney Learning System KPS Karnofsky Performance Scale LV Left ventricular LVD Left ventricular dilation LVEDd Left ventricular end-diastolic diameter LVH Left ventricular hypertrophy LVMI Left ventricular mass index LVVI Left ventricular volume index MAP Mean arterial blood pressure MCH Mean corpuscular hemoglobin MCHC Mean corpuscular hemoglobin concentration MCV Mean corpuscular volume MDRD4 Modification of Diet in Renal Disease MI Myocardial infarction MR Mitral regurgitation N, n Number of subjects N/A Not applicable NAPRTCS North American Pediatric Renal Transplant Cooperative Study nd Not documented ND-CKD NonÐdialysis-dependent chronic kidney disease NHANES National Health and Nutrition and Examination Survey NKF National Kidney Foundation NOS Not otherwise specified NS Not significant OR Odds ratio PD Peritoneal dialysis PD-CKD Peritoneal dialysisÐdependent chronic kidney disease PHRBC Percent hypochromic red blood cells PHRC Percent hypochromic red cells PO Oral PRCA Pure red cell aplasia Pt Patient PTA Posttransplantation anemia QoAL Quality of American Life QOL Quality of life S6 ACRONYMS AND ABBREVIATIONS

QOW Every other week RCT Randomized controlled trial rHuEPO Recombinant human erythropoietin RQoLP Renal Quality of Life Proﬁle RR Relative risk SAE Severe adverse event SBP Systolic blood pressure SC Subcutaneous SCD Sudden cardiac death

SCr Serum creatinine SF-36 36-Item Medical Outcomes Study Short-Form Health Survey SIP Sickness Impact Proﬁle SQUID Superconducting quantum interference device TDS Total daily supplement TID Thrice daily TIW Thrice weekly TRESAM Transplant European Survey on Anemia Management TSAT Transferrin saturation TTO Time trade-off TTP Thrombotic thrombocytopenic purpura Tx-ND-CKD NonÐdialysis-dependent kidney transplant recipient population U Unit URR Urea reduction ratio USFDA United States Food and Drug Administration USRDS United States Renal Data System vs Versus WHO World Health Organization ZPP Zinc protoporphyrin Anemia in Chronic Kidney Disease Work Group Membership

Work Group Co-Chairs David B. Van Wyck, MD Kai-Uwe Eckardt, MD University of Arizona College of Medicine University of Erlangen-Nuremberg Tucson, AZ Erlangen, Germany

Work Group John W. Adamson, MD Patricia Bargo McCarley, RN, MSN, NP Blood Center of SE Wisconsin Diablo Nephrology Medical Group Blood Research Institute Walnut Creek, CA Milwaukee, WI Hans H. Messner, MD George R. Bailie, MSc, PharmD, PhD University Health Network Albany College of Pharmacy Princess Margaret Hospital Albany, NY Toronto, Canada Jeffrey S. Berns, MD Allen R. Nissenson, MD University of Pennsylvania School of Medicine UCLA Medical Center Philadelphia, PA Los Angeles, CA Steven Fishbane, MD Gregorio T. Obrador, MD Winthrop University Hospital Universidad Panamericana School of Medicine Mineola, NY Mexico City, Mexico Robert N. Foley, MD John C. Stivelman, MD Nephrology Analytical Services Northwest Kidney Center Minneapolis, MN Seattle, WA Sana Ghaddar, RD, PhD American University of Beirut Colin T. White, MD Faculty of Agriculture and Food Sciences British Columbia Children’s Hospital Beirut, Lebanon Vancouver, Canada John S. Gill, MD, MS Liaison Member University Of British Columbia Francesco Locatelli, MD St. Paul’s Hospital Azienda Ospedaliera DI Lecco Vancouver, Canada Lecco, Italy Kathy Jabs, MD Iain C. Macdougall, MD Vanderbilt University Medical Center King’s College Hospital Nashville, TN London, England

Evidence Review Team National Kidney Foundation Center for Guideline Development and Implementation at Tufts-New England Medical Center, Boston, MA Katrin Uhlig, MD, MS, Project Director and Program Director, Nephrology Christina Kwack Yuhan, MD, Assistant Project Director

Amy Earley, BS Ashish Mahajan, MD, MPH Rebecca Persson, BA Priscilla Chew, MPH Gowri Raman, MD

In addition, support and supervision was provided by:

Joseph Lau, MD, Program Director, Evidence Based Medicine Ethan Balk, MD, MPH, Evidence Review Team Co-Director Andrew S. Levey, MD, Center Director KDOQI Advisory Board Members

Adeera Levin, MD, FACP KDOQI Chair Michael Rocco, MD, MSCE KDOQI Vice-Chair

Garabed Eknoyan, MD Nathan Levin, MD, FACP KDOQI Co-Chair Emeritus KDOQI Co-Chair Emeritus

George Bailie, PharmD, PhD William Mitch, MD Bryan Becker, MD Joseph V. Nally, MD Peter G. Blake, MD, FRCPC, MBB.Ch Gregorio Obrador, MD, MPH Allan Collins, MD, FACP Rulan S. Parekh, MD, MS Peter W. Crooks, MD Thakor G. Patel, MD, MACP William E. Haley, MD Brian J.G. Pereira, MD, DM Alan R. Hull, MD Neil R. Powe, MD Lawrence Hunsicker, MD Claudio Ronco, MD Bertrand L. Jaber, MD Anton C. Schoolwerth, MD Cynda Ann Johnson, MD, MBA Raymond Vanholder, MD, PhD George A. Kaysen, MD, PhD Nanette Kass Wenger, MD Karren King, MSW, ACSW, LCSW David Wheeler, MD, MRCP Michael Klag, MD, MPH Winfred W. Williams, Jr., MD Craig B. Langman, MD Shuin-Lin Yang, MD Derrick Latos, MD Donna Mapes, DNSc, MS Linda McCann, RD, LD, CSR Ex-Ofﬁcio Ravindra L. Mehta, MD, FACP Josephine Briggs, MD Maureen Michael, BSN, MBA David Warnock, MD NKF-KDOQI Guideline Development Staff Donna Fingerhut Anthony Gucciardo Margaret Fiorarancio Kerry Willis, PhD Richard Milburn VOL 47, NO 5, SUPPL 3, MAY 2006

American Journal of AJKD Kidney Diseases Foreword

his publication of the Clinical Practice these ‘non anemia’ purposes, but both the T Guidelines and Clinical Practice Recom- basic and clinical science of ESA therapy are mendations for Anemia represents the second advancing at a rapid pace. update of these guidelines since the first guide- This updated set of guidelines on anemia is line on this topic was published in 1997. The first unique and different from the previous anemia set of guidelines established the importance of guidelines. Firstly, the guidelines have been sepa- hemoglobin in dialysis patients, and established rated into evidence based guidelines and clinical guidelines and targets for the treatment of ane- practice recommendations, based on the strength mia in dialysis patients. The first update, pub- of evidence. The term guideline is reserved for lished in 2000, described anemia in a wider that which is robust enough to be used, if appro- spectrum of the chronic kidney patients, and priate, as a clinical performance measure. Clini- included those not on dialysis. cal practice recommendations are those recom- A number of important randomized clinical mendations based on expert opinion of the trials and large observational studies have been working group, but lacking sufficient hard data; commenced or completed in CKD populations CPRs are also susceptible to testing in a clinical both on and off dialysis, over the past 5 years. trial. Secondly, these guidelines incorporated in- The key questions that have been addressed in dividuals from countries outside the US (Europe, these recent studies have been the optimal UK, Middle East, Mexico and Canada) and at- hemoglobin target for CKD patients. To date tempted to build on most recently published there has been little support normalizing hemo- European Best Practice guidelines (2004), as globin, though a number of important studies well as the previously published KDOQI ver- have not been completed or reported (CHOIR sions. Lastly, the working group was clearly and TREAT), and the community eagerly awaits focused on ensuring that this document was them. Nonetheless, the key aspects of the pub- robust and clear for the reader. There was an lished data from both controlled trials and overt recognition that research recommendations observational studies, support the ongoing im- should be organized and described in sufficient portance of hemoglobin levels in risk stratifica- detail to ensure that prior to the next update of tion of patients with CKD, whether with native any anemia guideline, there would be new data or transplant kidneys, or on dialysis therapies. addressing gaps in our current knowledge. Thus, Much has been learned about the causes of there are no concrete research recommendations erythropoietin resistance, and some small stud- in this version; they will be published in a ies have reported methods of treating ESA separate document in the near future. The in- resistance. Furthermore, there are multiple new tended effect of this change in how the research insights into the optimizing treatment of ane- recommendations are presented is to provide a mia with iron supplementation and ESA. Novel actions of ESAs on cognitive function, as well as preservation of kidney and cardiac function © 2006 by the National Kidney Foundation, Inc. have been described. To date there are no 0272-6386/06/4705-0101$32.00/0 clinical trials to support the use of ESA for doi:10.1053/j.ajkd.2006.04.023

American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S9-S10 S9 S10 FOREWORD guidepost for funding agencies and investigators Learning System (KLS) component of the Na- to target research efforts in those areas that will tional Kidney Foundation is developing imple- provide important information that will benefit mentation tools that will be essential to the patient outcomes. success of these guidelines. This final version of the Clinical Practice In a voluntary and multidisciplinary undertak- Guidelines and Recommendations for Anemia ing of this magnitude, many individuals make has undergone revision in response to comments contributions to the final product now in your during the public review, an important and inte- hands. It is impossible to acknowledge them gral part of the KDOQI guideline process. None- individually here, but to each and every one of theless, as with all guideline documents, there is them we extend our sincerest appreciation. This a need for revision in the light of new evidence, limitation notwithstanding a special debt of grati- and more importantly, a concerted effort to trans- tude is due to the members of the Work Group, late the guidelines into practice. Considerable and their co-chairs, David Van Wyck and Kai- effort has gone into their preparation over the Uwe Eckardt. It is their commitment and dedica- past 2 years, and every attention has been paid to tion to the KDOQI process that has made this their detail and scientific rigor, no set of guide- document possible. lines and clinical practice recommendations, no matter how well developed, achieves its purpose Adeera Levin, MD unless it is implemented and translated into clini- KDOQI Chair cal practice. Implementation is an integral component of the KDOQI process, and accounts for Michael Rocco, MD the success of its past guidelines. The Kidney KDOQI Vice-Chair EXECUTIVE SUMMARY

INTRODUCTION that contribute to anemia may be helpful. Ane- Anemia commonly contributes to poor qual- mia is the clinical manifestation of a decrease in ity of life (QOL) in patients with chronic circulating red blood cell mass and usually is kidney disease (CKD). Fortunately, among the detected by low blood hemoglobin (Hb) concen- disorders that may afflict patients with CKD, tration. The cause, treatment, and prognostic anemia is perhaps the most responsive to treat- significance of anemic disorders vary widely. ment. Anemia was the subject of one of the Causes are distinguished clinically by markers of first efforts of the National Kidney Foundation the magnitude and appropriateness of a marrow (NKF) to improve patient outcomes through response to anemia. Under usual conditions, bone the development, dissemination, and implemen- marrow generates approximately 200 billion new tation of Dialysis Outcomes Quality Initiative cells per day to match the number of senescent (DOQI) Clinical Practice Guidelines.1 The first cells removed from circulation. The expected update of these guidelines appeared in 2001 compensatory response to anemia is a height- under the newly organized NKF-Kidney Dis- ened rate of erythropoiesis. Failure to demon- ease Outcomes Quality Initiative (KDOQI).2 strate a compensatory response signifies slowed In 2004, the NKF-KDOQI Steering Committee or defective erythropoiesis. Thus, hyperprolifera- appointed a Work Group and Evidence Review tive disorders reflect increased destruction of red Team (ERT) to undertake the first comprehen- blood cells with normal marrow response, sive revision of the KDOQI Clinical Practice whereas hypoproliferative and maturation disor- Guidelines for the Management of Anemia in ders reflect impaired red blood cell production. CKD. This Executive Summary provides a The cellular and molecular biology of erythro- brief background description of CKD and ane- poiesis has important implications for understand- mia, outlines the scope of the guidelines and ing, evaluating, and treating anemia in patients the methods of evidence review and synthesis, with CKD. Effective circulating red blood cell and provides the complete text of the guideline mass is controlled by specialized interstitial cells statements. in the kidney cortex that are exquisitely sensitive to small changes in tissue oxygenation.11 If tis- BACKGROUND sue oxygenation decreases because of anemia or About CKD other causes, these cells sense hypoxia and produce erythropoietin.12 Within erythroid islands, CKD is a worldwide public health issue.3 In the autonomous unit of erythropoiesis in mar- the United States, the incidence and prevalence row, receptors on the surface of the earliest red of kidney failure are increasing (Fig 1), out- blood cell progenitors, erythroid colony-forming comes are poor, and the cost is high.4 The preva- units (CFU-Es), bind erythropoietin. Binding of lence of earlier stages of CKD is approximately erythropoietin to erythropoietin receptors sal- 100 times greater than the prevalence of kidney vages CFU-Es and the subsequent earliest eryth- failure, affecting almost 11% of adults in the roblast generations from preprogrammed cell United States.4,5 There is growing evidence that death (apoptosis), thereby permitting cell sur- some of the adverse outcomes of CKD can be vival and division and the eventual expansion of prevented or delayed by preventive measures, erythropoiesis. If successful, these erythropoietin- early detection, and treatment. Strategies to im- stimulated events increase the production of re- prove outcomes include Clinical Practice Guide- 4 ticulocytes, restore normal circulating red blood lines (CPGs) for CKD and for the management 13 3 6 7 cell mass, and correct tissue hypoxia. of hypertension, dyslipidemia, bone disease, 8 9 In anemic patients, 1 or more steps in this nutrition, and cardiovascular disease (CVD) in patients with CKD. autoregulatory sequence may fail. In the pres-

About Anemia © 2006 by the National Kidney Foundation, Inc. Before considering a patient with both anemia 0272-6386/06/4705-0102$32.00/0 and CKD, a brief introduction to the processes doi:10.1053/j.ajkd.2006.03.010

American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S11-S15 S11 S12 EXECUTIVE SUMMARY

prompting a decrease in transferrin saturation (TSAT) and thereby promoting iron-deficiency erythropoiesis.15 Thus, the anemia of inflammation characteristically is hypoproliferative and not infrequently includes features suggesting iron- deficiency erythropoiesis. This synopsis provides an introduction to the subject and purpose of the following CPGs for patients with both anemia and CKD. Guide- line 1.1 describes identification of the patient Fig 1. Kidney failure in the United States. Preva- with anemia and CKD. Guideline 1.2 describes lence estimates of kidney failure treated by dialysis the recommended initial evaluation; Guideline and transplantation (end-stage renal disease [ESRD]) 2.1, the goal of treatment; and Guidelines 3.1, in the United States. Prevalence refers to the number of patients alive on December 31st of the year. 3.2, and 3.3, the use of therapeutic agents. Upper and lower estimates reflect the effect of potential changes in ESRD incidence, diabetes preva- SCOPE OF THE GUIDELINES lence, ESRD death rate, and underlying age and race structure of the US population after the year 2000. New findings, new agents, and the need for Reprinted with permission.10 an expanded scope prompt the need for a comprehensive revision of existing NKF- KDOQI CPGs for the Treatment of Anemia in ence of kidney disease, erythropoietin produc- CKD. In preparing the current guidelines, the tion may be impaired, leading to erythropoi- Anemia Work Group members broadened our etin deficiency and the apoptotic collapse of inquiry to include all stages of CKD, identify early erythropoiesis. If folate and vitamin B12 areas of concern to current practitioners, adopt are lacking, deoxyribonucleic acid (DNA) syn- a structured intensive evidence review process thesis is impaired, and erythroblasts, normally not previously used, apply that process to both undergoing rapid division during this period of newly available literature and literature exam- erythropoiesis, succumb to apoptosis.13 In on- ined in the development of previous guideline going folate or vitamin B12 deficiency, disor- versions, formulate conclusions that distin- dered DNA synthesis, maturation arrest, and guish evidence-based guidelines from expert- ineffective early erythropoiesis lead to a mac- opinion–based clinical practice recommenda- rocytic anemia. Conversely, if iron is lacking, tions (CPRs), and present both guidelines and the Hb-building steps that follow rapid cell recommendations in a new format to more division are affected: synthesis of both heme clearly describe what is not known. To ensure and globin slow, and erythropoiesis is im- that the next update profits from evidence we paired. Reticulocytes that emerge from mar- currently lack, we identified limitations of cur- row are few, poorly hemoglobinized, and small. rently available evidence and, in a subsequent A hypochromic microcytic anemia eventually report, will identify priorities for needed re- results. Inflammation, a disorder common search. among patients with CKD, appears to impair both the early erythropoietin-dependent period Intended Reader of erythropoiesis and the later iron-dependent Our intended reader is the practitioner who period. Inflammatory cytokines inhibit erythro- manages patients with CKD, including neph- poietin production, directly impair growth of rologists, primary-care providers, cardiolo- early erythroblasts, and—especially in the ab- gists, nurse practitioners, nurses, and dieti- sence of erythropoietin—promote death by li- tians. Clinical pharmacists, quality outcomes gand-mediated destruction of immature erythro- directors, and clinical investigators will find blasts.14 In addition, inflammatory cytokines the document useful. We write primarily for stimulate hepatic release of hepcidin, which si- practitioners in North America. Nevertheless, multaneously blocks iron absorption in the gut from the onset, we have coordinated our ef- and iron release from resident macrophages, forts with guideline development processes EXECUTIVE SUMMARY S13 elsewhere, ensured that Work Group member- The overall strength of each guideline state- ship includes experts from Latin America and ment was rated by assignment as either strong Europe, and planned so that this document or moderately strong. A strong rating indicates may serve as the last of the KDOQI Anemia “it is strongly recommended that clinicians Guidelines and a foundation for the first truly routinely follow the guideline for eligible pa- global guideline under the auspices of the tients. There is high-quality evidence that the Kidney Disease: Improving Global Outcomes practice results in net medical benefit to the (KDIGO) initiative (www.kdigo.org). patient.” The moderately strong rating indicates “it is recommended that clinicians rou- Scope tinely follow the guideline for eligible patients. We address the target population of patients There is at least moderately high-quality evi- with CKD stages 1 to 5 not on dialysis therapy, dence that the practice results in net medical on hemodialysis (HD) or peritoneal dialysis benefit to the patient.” Overall, the strength of (PD) therapy, or with a kidney transplant in the the guidelines and recommendations was based full range of practice settings in which they are on the extent to which the Work Group could encountered. However, the evidence continues be confident that adherence will do more good to derive disproportionately from findings in than harm. Strong guidelines require support facility-based HD patients. We have not been by evidence of high quality. Moderately strong unmindful of cost implications. However, net guidelines require support by evidence of at medical benefit to the patient is the foundation least moderately high quality. Incorporation of of each of our guidelines and recommenda- additional considerations modified the linkage tions. Our commitment, in short, is to assist between quality of evidence and strength of practitioners in the care of patients by describ- guidelines, usually resulting in a lower strength ing best practice and the evidence for it. In this of the recommendation than would be support- way, individual practitioners who face local able based on the quality of evidence alone. reimbursement constraints may make informed Our objective is to describe the evidence decisions armed with appropriate facts. Simi- base for key elements in the identification, larly armed, those responsible for guiding reim- evaluation, and management of patients with bursement policy may make informed deci- CKD-associated anemia. For topics on which sions consistent with the evidence. we undertook a systematic literature review, we present detailed information, usually in the Evidence, Opinion, Guidelines, and form of summary evidence tables. These topics Recommendations include Hb thresholds for initiating therapy, When the quality of evidence is high or Hb level therapeutic goals, iron status goals, or moderately high, we present a CPG based on efficacy of adjuvants in achieving Hb goals. evidence, rate the quality of that evidence, and Results that bear directly on patient lives (mor- distinguish the evidence-based guideline by tality, morbidity, QOL, and adverse events enclosing it in a text box. When the quality of [AEs]) receive particular attention. Topics for evidence is low, very low, or missing, but the which the evidence base is limited deserve topic is important to practitioners, we offer a brief mention and receive it. Information that CPR and cite limitations to the current litera- involves implementation, application, or proto- ture. In a subsequent document, we will pro- col development, we leave to the NKF Kidney pose and prioritize needed research. Through- Learning System (KLS) resources. out this document, a text box surrounds each When the quality of evidence is low, we will evidence-based guideline, and the strength of follow acknowledgement of these limitations in the guideline is rated as “strong” or “moder- a separate publication by encouraging further ately strong.” The phrase, “In the opinion of investigation. Work Group members” precedes each CPR. Distinguishing evidence-based guidelines from Relationship to Previous Guidelines CPRs is in keeping with recent advances in The lineage of the current document derives KDOQI policy and practice. from both the most recent (2001) version of the S14 EXECUTIVE SUMMARY

Fig 2. Relationship between current and previous anemia guidelines.

KDOQI Guidelines for Anemia2 and the most lines into sections on identiﬁcation, deﬁnition, recent (2004) version of the European Best Prac- and evaluation of anemia (Guideline 1.1 and tices Guidelines (EBPGs) for anemia (Fig 2).16 1.2); Hb treatment target (Guideline 2.1); and Our intention is to simplify the current guide- management (Guidelines 3.1-4.1), with parallel EXECUTIVE SUMMARY S15

Table 1. Key Differences between Current Guidelines (KDOQI Anemia 2006) and Previous Anemia Guidelines (KDOQI 2000 and EBPG 2004)

KDOQI 2000 Anemia EBPG 2004 Anaemia KDOQI 2006 Anemia Topic Guideline Guideline Guideline Reason KDOQI 2006 Differs from Prior Guidelines Definition of Anemia by Hb <12.0 g/dL in males and <12.0 g/dL males <13.5 g/dL males KDOQI 2006 uses more recent NHANES data set, defines anemia postmenopausal females <11.0 g/dL females <12.0 g/dL females as any Hb below the 5th percentile for the adult, gender-specific <11.0 g/dL in premenopausal population. Among males, no adjustment is made for age >70 females and prepubertal years, to exclude the possibility that pathological conditions patients contribute to lower Hb values. Among females, the 5th percentile determination is made only among individuals without evidence of iron deficiency, as defined by TSAT <16% or ferritin <25 ng/mL. bH tegraT bH 21-11 Ld/g 0.11> /g Ld tegrat ≥11 g/dL, caution when Current guideline reflects QOL benefits at Hb maintained ≥11.0 >12.0 in CVD not intentionally maintaining g/dL, risks when intentionally maintaining Hb >13.0, and recognition recommended Hb >13 g/dL that Hb will often exceed 13 g/dL unintentionally, without evidence Hb >14.0 g/dL not desirable of increased risk, in patients with Hb intent to treat ≥11.0 g/dL. Target Iron Status TSAT (%) TSAT (%) TSAT (%) TSAT: lower limit: 20 lower limit: 20 lower limit ≥20 Current guideline reflects unchanged lower bound for iron therapy; upper limit: 50 target: 30-50 upper limit of TSAT not specified.

Ferritin (ng/mL) Ferritin (ng/mL) Ferritin (ng/mL) Ferritin: lower limit: 100 lower limit: 100 lower limit: 200 HD-CKD Current guideline distinguishes HD- from non–HD-CKD on basis of target: 200-500 100 non-HD-CKD available evidence. Lower limit sets objective of iron therapy. There > 500 not routinely is insufficient evidence to assess harm and benefit in maintaining recommended ferritin > 500 ng/mL . In HD-CKD, 200 ng/mL reflects evidence for substantial efficacy of IV iron at ferritin <200 ng/mL. Adjuvants L-Carnitine Not recommended Not recommended for Not routinely Current guideline based on low-quality evidence which shows lack general use recommended of efficacy Asco etabr Not r uo it len y Current guideline reflects combination of safety concerns and low recommended quality evidence of efficacy dnA dnA or sneg eleS c evit su e Not ocer mme dedn C u tnerr lediug ni e tcelfer s suoires as ef ty noc cer .sn ecnedivE rof efficacy is low quality,. organization for pediatrics. Key differences be- Indicators of Efficacy tween current and past guidelines are set out in Efficacy of anemia management: Efficacy of Table 1. anemia management is indicated in a single patient by an Hb level greater than the target threshold, and OPERATING DEFINITIONS in a group of patients, by the percentage of patients maintained at greater than target threshold. Erythropoiesis-stimulating agent (ESA): The ESA efficacy: ESA efficacy is indicated by the term ESA applies to all agents that augment dose of ESA needed to achieve or maintain Hb erythropoiesis through direct or indirect action levels at greater than the target threshold. on the erythropoietin receptor. Currently avail- Adjuvant efficacy: Adjuvant efficacy is indi- able ESAs include epoetin alfa, epoetin beta, and cated by the percentage of patients who success- darbepoetin alfa. fully exceed the target Hb threshold or achieve an Therapeutic goal: A therapeutic goal is the increase in Hb level greater than 1 g/dL without intended goal of current therapy, not the cur- initiating ESA therapy or increasing ESA doses to rent result of previous therapy. This definition greater than baseline. Among patients with steady- applies equally to therapeutic goals for both state Hb levels greater than target threshold, adju- Hb and iron. vant efficacy is indicated by ESA dose reduction. II. CLINICAL PRACTICE GUIDELINES AND CLINICAL PRACTICE RECOMMENDATIONS FOR ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS CPR 1.1. IDENTIFYING PATIENTS AND INITIATING EVALUATION

Identifying anemia is the first step in evaluating dialysis therapy at very low levels of kidney the prognostic, diagnostic, and therapeutic signifi- function.23,24 cance of anemia in the patient with CKD. The reported prevalence of anemia by CKD stage depends in large part on the size of the 1.1.1 Stage and cause of CKD: In the opinion of the Work Group, Hb study; whether study participants are selected testing should be carried out in all from the general population, are at high risk for patients with CKD, regardless of stage CKD, or are patients already under a physician’s or cause. care; what level of Hb is defined as constituting 1.1.2 Frequency of testing for anemia: anemia; and whether patients do or do not have In the opinion of the Work Group, Hb diabetes. levels should be measured at least Studies reviewed for the purposes of guideline annually. statement 1.1.1 include those of patients with 1.1.3 Diagnosis of anemia: CKD before dialysis therapy, those with kidney In the opinion of the Work Group, transplants, and those on dialysis therapy. The diagnosis of anemia should be made reviewed literature spans close to 40 years of and further evaluation should be un- investigation up to the year 2000 and describes dertaken at the following Hb concen- clinical findings of researchers as they explore trations: the relationships between Hb level or hematocrit ● <13.5 g/dL in adult males. (Hct) and kidney function (Table 2 and 3). The ● <12.0 g/dL in adult females. majority of available data were derived from studies of small sample size, most of which are BACKGROUND cross-sectional studies, or baseline data from Anemia develops early in the course of CKD clinical trials of variable size and robustness. and is nearly universal in patients with CKD In 12 of the 21 studies reviewed, there was an stage 5.17 The purpose of specifying Hb level association between level of Hb or Hct and the thresholds to define anemia is to identify patients selected measure of kidney function. They also who are most likely to show pathological pro- demonstrate variability in levels of Hb or Hct at cesses contributing to a low Hb level and who each level of kidney function, whether assessed therefore are most likely to benefit from further by using serum creatinine (SCr) concentration, anemia evaluation. Thus, the current guideline creatinine clearance (CCr), or estimated GFR (Guideline 1.1) identifies Hb level thresholds (eGFR). However, the consistency of the informa- that should trigger diagnostic evaluation (Guide- tion they provide indicates a trend toward lower line 1.2). Conversely, Hb levels that should trig- Hb levels at lower levels of GFR and variability ger ESA therapy are defined in Guideline 2.1. in Hb levels across GFR levels. More recent studies used large databases to RATIONALE examine the relationship between Hb level or Hct and kidney function (Table 4). A cross- sectional analysis examined 12,055 ambulatory Stage and Cause of CKD adult patients (Ն18 years of age) who had at

Hb testing should be carried out in all patients least 2 SCr measurements 2 years apart and at with CKD, regardless of stage or cause. Anemia least 1 Hct measurement and weight recorded is associated with CKD of any cause,18-22 includ- between 1990 and 1998.25 Results should pro- ing transplant-associated CKD (see Guideline gressively lower Hct at an estimated CCr less 4.1). The severity of anemia in patients with than 60 mL/min/1.73 m2 in men and 40 mL/min/ CKD is related to both the degree of loss of 1.73 m2 in women (Table 5). When GFR normal- glomerular ﬁltration rate (GFR) and the cause of ized to body surface area (BSA) was estimated kidney disease. The lowest Hb levels are found by using the Modiﬁcation of Diet in Renal Dis- in anephric patients and those who commence ease (MDRD) equation, both men and women

American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S17-S27 S17 S18 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

developed a statistically significant decrease in and 45 mL/min/1.73 m2 in women, the associ- Hct at the same BSA-normalized GFR, 50 mL/ ated Hb level is lower than that seen at eGFRs min/1.73 m2. This relationship persisted when above those thresholds (Fig 3). The relationship analyses were restricted to patients who were is observed best in the distribution of the lowest normochromic and normocytic, ie, presumably fifth percentile of Hb levels. In general, the lower nonÐiron deficient. However, the change in age- the eGFR at less than the respective threshold, and race-adjusted Hct with decreasing levels of the lower the associated Hb level, but confidence GFR was greater in men than women. It should intervals (CIs) are broad at the lowest eGFR be noted that only 10% of men and 6% of women levels because the number of observed individu- had an eGFR of 50 mL/min/1.73 m2 or less als with a low eGFR is small. For example, only (MDRD formula), and the indications for measur- 3.5% of women and 1.4% of men in the survey 2 ing SCr and Hct were unknown. showed an eGFR of 30 mL/min/1.73 m or less. Two studies examined the relationship be- The relationship between GFR and prevalence tween prevalent Hb concentration and GFR by of anemia is determined in large part by the Hb using results from the National Health and Nutri- concentration used to define anemia. In NHANES tion and Examination Survey (NHANES) III III, prevalence of a Hb level less than 13 g/dL database. NHANES III is a cross-sectional sur- increases at less than a threshold eGFR of 60 vey of nutritional and health status in 15,419 mL/min/1.73 m2 in males and 45 mL/min/1.73 individuals randomly selected from the general m2 in females (Fig 4). However, the prevalence US population from 1988 to 1994. Results from of a Hb level less than 11 g/dL is not greater than NHANES III26,27 are consistent with those ob- reference except at an eGFR less than 30 mL/min/ tained in ambulatory adult patients (Table 6).25 1.73 m2 in both males and females. Again, pre- At an eGFR less than 75 mL/min/1.73 m2 in men cise estimates of prevalence are limited by rela- ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S19

tively small numbers; in particular, less than a stage 5) were 12.8 Ϯ 1.5, 12.4 Ϯ 1.6, 12.0 Ϯ GFR of 30 mL/min/1.73 m2 (N ϭ 52). 1.6, and 10.9 Ϯ 1.6 g/dL, respectively.29 The The relationship between prevalence of ane- prevalence of untreated anemia (defined as a mia and GFR may be changing as a result of Hb level Ͻ 12, 10 to 12, and Ͻ 10 g/dL) for earlier identification and treatment of anemia. different CKD stages (Fig 6) is much greater Comparing results of NHANES III with than that reported in NHANES surveys.26,27 A NHANES IV shows a lower prevalence of greater disease burden among patients com- anemia for each CKD stage in the more recent pared with randomly selected individuals pro- survey (Fig 5).28 In this analysis, anemia was vides a likely explanation for these findings. defined by World Health Organization (WHO) Because data points at GFR levels less than criteria (Hb level Ͻ 12 g/dL in women and Ͻ 30 mL/min/1.73 m2 were scarce in all except 13 g/dL in men). the previous cross-sectional study,29 the Cana- The prevalence of anemia among patients is dian Multicentre Study30 was used to demon- much greater than that observed among indi- strate trends in a large cohort of patients before viduals randomly surveyed in the general popu- dialysis therapy (Fig 7). The prevalence of lation. In a cross-sectional study of 5,222 adult anemia was greatest at the lowest levels of patients with CKD who were selected from GFR, but approached 20% among patients 237 US physician practices (including family with a GFR of 30 to 44 mL/min/1.73 m2 (Fig practice, internal medicine, nephrology, and 8). Similarly, in another study of 131,848 pa- endocrinology), mean Hb levels for an MDRD tients who began dialysis therapy in the United eGFR of 60 or greater (CKD stages 1 and 2), States between 1995 and 1997, proportions of 30 to 59 (CKD stage 3), 15 to 29 (CKD stage patients with Hcts less than 30% and less than 4), and less than 15 mL/min/1.73 m2 (CKD 36% immediately before the initiation of dialy- S20 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

Table 4. Prevalence of Anemia by Level of Kidney Function

Definition of Level of Kidney Author, Year N Applicability Prevalence of Anemia Results Quality Anemia Hb (g/dL) Function Diabetic Nondiabetic Men & Women >50 GFR ≥89 8.7% 6.9% El-Achkar, 2005 37 5,380 yr, Hb <12; Women GFR 60-89 7.5% 7% ≤50 yr, Hb <11 GFR 30-59 22.2% 7.9% GFR <30 52.4% 50% (Likelihood [%] of Hb level <12 g/dL) a Men Women 26 Men Hb 10-12 Hsu, 2002 15,971 CrCl >80 3% 7% Women Hb10-12 CrCl 41-50 4% 10% CrCl 21-30 12% 25% GFR ≥ 09 1 %8. Men Hb <12 b RFG 98-06 %3.1 Astor, 2002 27 15,419 Women Hb <11 GFR 3 0 95- %2.5 GFR 1 5-29 44 . %1 c Men Women GFR >41-50 9% (7/77) 5% (18/333) Hsu, 2001 25 8,220 Hb <11 GFR 31-40 10% (3/29) 10% (12/116) d GFR 20-30 0% (0/17) 33% (9/27) GFR ≤20 60% (12/20) 52% (11/21) GFR ≥60 %7.62 60> GFR ≥30 41.6% McClellan, 2004 29 5,222 Hb <12 30> GFR ≥15 53.6% 1< RFG 1< 5 .57 5% CrCl >90 e <1% CrCl 98-06 <2% 85 Fehr, 2004 4,760 Hb <11 CrCl 95-04 %3~ CrCl 02 93- 51~ % CrCl 2< 0 %83~ Footnotes: a. Data given for age 61-70 and Hb <12 g/dL. b. To focus on clinically significant and potentially treatable anemia, a more stringent definition (Hb <12 for men, Hb <11 for women) was used for most analyses. However, in analysis using the WHO definition (Hb <13 for men and Hb <12 for women) the overall prevalence for men was 3.5% vs. women 10.7%. c. All prevalence rates adjusted for age = 60. This corresponds to approximately 160,000 (SE 44,000) noninstitutionalized civilians in the US in 1991. d. At any given level of CrCl, men had steeper slope of decrease in Hb, as did blacks> Mexican >whites, older vs. younger men, and younger vs. older women. e. No statistical pooling between categories of creatinine clearance stages.

Coding of Outcome: statistically significant increase in anemia with decreasing kidney function.

sis therapy were 67.5% and 93%, respec- extension of the same audit, patients with type 2 31 2 tively. diabetes with a CCr of 60 to 90 mL/min/1.73 m Patients with diabetes are more prone to both were twice as likely to have anemia than those 2 develop anemia and develop anemia at earlier with a CCr greater than 90 mL/min/1.73 m . stages of CKD than their nondiabetic counter- Patients with diabetes with a CCr less than 60 parts (Fig 9).32-37 In a cross-sectional clinical mL/min/1.73 m2 were twice as likely to have audit of 820 Australian patients with diabetes, anemia as those with a CCr of 60 to 90 mL/min/ anemia was 2 to 3 times more prevalent in 1.73 m2.36 In the Kidney Early Evaluation Pro- patients with diabetes compared with the general gram (KEEP 2.0), a cross-sectional community- population at all levels of GFR (Fig 9). In an based screening program aimed at detecting CKD ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S21

among high-risk patients (those with diabetes, 2) is relatively low in individuals randomly hypertension, or a family history of kidney dis- selected from the general population, but is not ease), the prevalence of anemia at each level of uncommon in patients with CKD under the eGFR from patients with CKD stages 2 through care of a physician or identified by virtue of 5 was greater than in patients without diabetes: being at high risk for CKD; (2) among individu- 8.7% versus 6.9% in stage 2 (P ϭ not significant als from a general population, mean Hb levels [NS]), 7.5% versus 5.0% in stage 3 (P ϭ 0.015), decrease and anemia develops consistently only 22.2% versus 7.9% in stage 4 (P Ͻ 0.001), and when GFR is less than 60 mL/min/1.73 m2 52.4% versus 50% in stage 5 (P ϭ 0.88).37 (stage 3 CKD); (3) the prevalence of anemia The following conclusions can be drawn increases at later stages of CKD (CKD stages 4 from these studies: (1) the prevalence of ane- and 5); (4) there is significant variability in Hb mia at higher levels of GFR (CKD stages 1 and levels at any given level of kidney function;

Fig 3. Relationship between Hb level and eGFR. Data represent a cross- sectional survey of individuals randomly selected from the general US population (NHANES III). Median and 95% conﬁdence limits around the 95th and 5th percentiles are shown for males and females at each eGFR interval. Adapted with permission.27

Fig 4. Relationship between eGFR and prevalence of anemia as deﬁned by differing Hb levels for (A) males and (B) females. Data from NHANES III. Reprinted with permission.27 S22 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

Fig 5. Prevalence of anemia by CKD stage. Anemia, Fig 7. Hb percentiles by GFR in the Canadian Multi- defined as an Hb level less than 13.0 g/dL in males and centre Longitudinal Cohort Study. Patients were re- less than 12.0 g/dL in females, was plotted by stage of ferred to nephrologists between 1994 and 1997. Re- CKD in results from NHANES III compared with sults shown exclude patients receiving ESA therapy NHANES IV. USRDS 2004.28 or with an arteriovenous fistula. Reprinted with permission.4,30 and (5) anemia among patients with diabetes compared with patients without diabetes is subsequently decreased. During 2 years, a signifi- more prevalent, more severe, and occurs ear- cant proportion of patients eventually required lier in the course of CKD. ESA therapy. However, among those who did not These observations underscore the need to require ESA therapy, mean Hb values remained measure Hb levels in every patient with CKD, relatively stable (Fig 10). This evidence suggests regardless of stage or cause. the need for regular surveillance of Hb levels in patients with CKD without ESA therapy. The Frequency of Hb Testing in Patients With CKD statement Hb levels should be measured at least Hb levels should be measured at least annu- annually emphasizes that more frequent Hb sur- ally. Because little is known about the natural veillance likely will be needed for selected pa- history of anemia in patients with CKD, precise tients, including those with greater disease bur- information is unavailable to determine the opti- den, unstable clinical course, or evidence of mum frequency of Hb testing in patients with previous Hb level decline. Frequency of Hb CKD. The recommendation that patients be evalu- testing in patients already undergoing ESA ated at least annually rests on observations from therapy is addressed in Guideline 3.1.1. clinical trials that (in the absence of ESA therapy) the natural history of anemia in patients with Diagnosis of Anemia CKD is a gradual decline in Hb levels over Diagnosis of anemia should be made and time.38,39 In 1 trial,39 patients assigned to the further evaluation should be undertaken at Hb lower Hb target (9.5 to 10.5 g/dL) entered the concentrations less than 13.5 g/dL in adult males trial with Hb values greater than target (mean, and less than 12.0 g/dL in adult females. The 11.8 g/dL) and received ESA only if Hb levels recommended thresholds for defining anemia

Fig 6. Relationship between level of renal function, reﬂected by

SCr level or GFR, and prevalence of anemia, deﬁned at different Hb cutoff levels, among patients under the care of physicians. Re- printed with permission.29 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S23

rent evidence of pathological conditions that may cause or contribute to anemia.32 Because we cannot exclude potential pathological states, we cannot assume that lower Hb levels in older males are normal. Therefore, we make no adjustment for age among males. The recommended threshold of less than 12.0 g/dL in females reflects exclusion from the healthy population of individuals whose iron status test results suggest that iron deficiency is contributing to a low Hb level. Fig 8. Prevalence of low Hb level by category of The recommended definition of anemia dif- GFR in the Canadian Multicentre Longitudinal Cohort fers from previous CPGs. The WHO defines 4,30 Study. Reprinted with permission. anemia as a Hb level less than 13.0 g/dL in adult men and less than 12 g/dL in adult represent the mean Hb of the lowest fifth percen- women.42 Differences between the current rec- tile of the sex-specific general adult population ommendation and the WHO definition arise 40,41 (Table 7 and 8; Fig 11). The conclusion that from differences in the data source for the anemia therefore is defined as a Hb level less general population: the WHO definition is based than 13.5 g/dL in adult males and less than 12 on sparse data obtained prior to 1968, whereas g/dL in adult females (Fig 3) assumes a lack of the definition proposed in the current guide- adjustment downward for age in males and an lines is based on more recent NHANES III adjustment upward for iron deficiency in fe- data. The recently published Revised EBPG males. Although mean Hb level representing the for the Management of Anaemia in Patients fifth percentile decreases among males older with Chronic Renal Failure also defines ane- than 60 years, it is clear that a substantial fraction mia as a Hb level less than 13 g/dL in men, but of older males with low Hb levels show concur- adjust this to less than 12 g/dL for men older

Fig 9. Prevalence of low Hb level by eGFR in patients with diabetes compared with the general population. Open circles (Œ) indicate the diabetic population, while closed circles () represent general population data from NHANES III. A: Hb <11 g/dL; B: WHO criteria (men, Hb <13 g/dL; women, Hb <12 g/dL). Reprinted with permission.35

Fig 10. Changes in Hb levels and eGFR over 2 years in adult patients with CKD stages 3 and 4 not treated with ESA. Oral iron supplements and occasional IV iron treatment were administered to maintain TSAT at greater than 20% and ferritin levels greater than 100 ng/mL. Data given as mean SD. Although mean Hb levels showed ؎ little or no change, the number of study patients remaining without ESA therapy de- clined. Results courtesy of the authors, from control group patients previously described. Reprinted with permission.39 S24 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

than 70 years. Because the EBPGs do not ciency), and less than 12 g/dL for postmeno- adjust for the potential contribution of iron pausal women (no change). deficiency in women, the EBPG threshold for The importance of identifying patients with women is less than 11.5 g/dL.16 Finally, the anemia in the presence of CKD is 2-fold. First, a previous KDOQI Anemia Guidelines recom- Hb value in the lowest fifth percentile of the mended less than 12 g/dL for men (reflecting general population may signify the presence of use of a different data set), less than 11 g/dL significant nutritional deficits, systemic illness, for women of reproductive age (different data or other disorders that warrant attention. Second, set, no attempt to exclude effect of iron defi- anemia in patients with CKD is a known risk ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S25

factor for a number of significant adverse patient Hb values, in addition to sex, and therefore must outcomes, including hospitalizations, CVD, cog- be considered in patients with CKD. The current nitive impairment, and mortality.16,43-50 Regard- definition for anemia reflects results from adult less of whether concurrent disorders, if any, are patients older than 18 years, of all races and treatable, awareness of their presence is likely to ethnic groups, and living at relatively low alti- be helpful to the clinician. Similarly, the benefits tude (Ͻ1,000 m or 3,000 ft). of anemia treatment (see Guideline 2.1) cannot Altitude has a direct impact on red blood cell extend to patients whose anemia remains uniden- number, mass, and volume.51 Generally, Hb con- tified. centration can be expected to increase by about Altitude, age, race, and smoking each contrib- 0.6 g/dL in women and 0.9 g/dL in men for each ute to the interpretation of the normal range of 1,000 m of altitude above sea level.52 This in- S26 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

Fig 11. Distribution of Hb levels in adult males and females by age group. The 5th, 50th, and 95th percentiles are shown for each age group. In females, results reflect only individuals with serum ferritin greater than 25 ng/mL or TSAT greater than 16% because lower values may contribute to low Hb levels. Dash-dot-dash lines indicate recommended cutoff value to define anemia. (Source: US Renal Data System [USRDS] special data request, NHANES 1999-2002 data set). crease in Hb levels seems to be caused at least in with current opinion that anemia is not a normal part by increased erythropoietin production.53 consequence of aging.56 The threshold Hb level defining anemia in pa- Among women of reproductive age, menstrual tients living at high altitude should be adjusted losses, the potential for iron deficiency, and the upward, in keeping with the degree of elevation effect of pregnancy each deserve consideration. (Table 9). Although Hb values in women are lower than In children, mean Hb level and the variability those in men (Table 7 and 8), Hb concentrations around the mean vary highly with age. These in iron-replete women remain stable between 20 findings are discussed elsewhere in more detail and 80 years of age.54 Regardless of the presence (in CPR for Children 1.1). Among adult males, or absence of CKD, the prevalence of anemia is as previously noted, the lower fifth percentile of greater in women compared with men.55 Menses Hb values declines as age advances (Fig 11). One and pregnancy may each contribute to this find- study reported a mean decrease of 1 to 1.5 g/dL ing. Menstrual losses of iron average 0.3 to 0.5 in Hb concentration in males as they increased in mg/d and may result in mostly iron-deficiency age from 50 to 75 years; however, this trend was anemia.42 Among pregnant women, Hb concen- not seen in women, in whom Hb concentration tration decreases during the first and second remained stable between 20 and 80 years of trimesters largely because of the dilutional effect age.54 Although it previously was believed that of expanding blood volume. In the third trimes- decreases in Hb levels might be a consequence of ter, Hb concentration remains low in pregnant normal aging, evidence has accumulated that women who do not take iron supplements, but anemia reflects poor health and increased vulner- gradually increases toward prepregnancy levels ability to adverse outcomes in older persons.32 in those who take iron supplements.57-59 Table 10 Consequently, the current recommendation makes shows maximum Hb cutoff values for anemia no adjustment for aging in men. An association during pregnancy. between an increasing prevalence of anemia and Hb values also vary significantly between older age has been reported in most,27,55 but not races, with African-American individuals consis- all,29 studies of patients with CKD, in keeping tently showing Hb concentrations 0.5- to 0.9-

Table 9. Normal Increase in Hb Levels Related to Long-Term Altitude Exposure Altitude (meters) Increase in Hb (g/dL) 000,1< 0 000,1 2.0+ 005,1 5.0+ 000,2 8.0+ 005,2 3.1+ 000,3 9.1+ 005,3 7.2+ 000,4 5.3+ 005,4 5.4+ Reprinted with permission.42 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S27

Table 10. Maximum Hb Cutoff Values for Anemia in Pregnancy Weeks’ Gestation Hb Concentration (g/dL) 21 0.11 61 6.01 02 5.01 42 5.01 82 7.01 23 0.11 63 4.11 04 9.11 retsemirT tsriF 0.11 dnoceS 5.01 drihT 0.11

Maximum cutoff values for anemia are based on the 5th percentile from the third National Health and Nutrition Examination Survey (NHANES III), which excluded persons who had a high likelihood of iron deficiency. Maximum values for anemia during pregnancy are based on values from pregnant women who had adequate supplementation. Reprinted with permission.41 g/dL lower than those of Caucasoid or Oriental LIMITATIONS 60-62 populations. Because the reason for this dispar- Many studies that examined the relationship ity in Hb level distributions by race has not been between Hb level and kidney function: determined and could reflect increased disease burden, this guideline does not provide race-specific 1. Have been cross-sectional and not longitu- cutoff values for anemia.41,63,64 Anemia may de- dinal in design. velop at an earlier stage in the course of CKD 2. Described patients entered into clinical trials among non-Hispanic African Americans compared or seen by nephrologists, which are not a truly with individuals of other races.27 African-Ameri- representative sample of patients with CKD. can patients have a greater prevalence of anemia 3. Included small numbers of patients with at every stage of CKD compared with whites.55 lower levels of kidney function. Last, smoking is associated with elevated 4. Used a great variety of methods to assess carboxyhemoglobin levels, which are associ- level of kidney function. It therefore is ated with a compensatory increase in total Hb difficult to determine whether the variabil- concentration.65 If the same definition of ane- ity in Hb at levels of kidney function is mia is applied to both smokers and nonsmokers, caused by variability in measurements of the risk for developing anemia is lower in current kidney function or variability associated or past smokers than nonsmokers.55 Because with CKD itself. approximately 20% to 40% of patients with 5. Used the MDRD4 formula to estimate CKD are current or past smokers,55 consider- GFR, the precision of which decreases at ation should be given to the significance of a higher levels of kidney function. smoking history in interpreting results of Hb 6. Did not describe the cause of the anemia in measurement (Table 11). patients with CKD.

Table 11. Adjustment to Hb for Smoking by Number of Packets per Day puorG bH )Ld/g( srekomsnoN 0 )lla( srekomS )lla( 3.0+ 0.5-1 packet/day +0.3 1-2 packets/day +0.5 >2 packets/day +0.7 Reprinted with permission.55 CPR 1.2.: EVALUATION OF ANEMIA IN CKD

Anemia in patients with CKD is not always RATIONALE caused by erythropoietin deficiency alone. Initial laboratory evaluation therefore is aimed at iden- Initial Assessment of Anemia tifying other factors that may cause or contribute The CBC provides information about the sever- to anemia or lead to ESA hyporesponsiveness. ity of anemia, adequacy of nutrients (including folate, vitamin B12, and iron), and adequacy of 1.2.1 In the opinion of the Work Group, bone marrow function. initial assessment of anemia should Severity of anemia is assessed best by measur- include the following tests: ing Hb concentration rather than Hct because Hb 1.2.1.1 A complete blood count (CBC) is a stable analyte that is measured directly. The including—in addition to the Hb assay is standardized and is not influenced by Hb concentration—red blood differences in instrumentation. Conversely, Hct cell indices (mean corpuscu- measurement is relatively unstable and lacks lar hemoglobin [MCH], mean standardization. The Hct result is derived indi- corpuscular volume [MCV], rectly by automated analyzers and is instrumen- mean corpuscular hemoglo- tation dependent.16 Analytical advantages of Hb bin concentration [MCHC]), level over Hct are described next.2 white blood cell count, and Although blood-sample storage conditions differential and platelet count. have no effect on Hb measurement, Hct in- 1.2.1.2 Absolute reticulocyte count. creases with storage temperature and duration 1.2.1.3 Serum ferritin to assess iron because stored red blood cells swell. The MCV stores. from which the Hct is calculated (MCV ϫ eryth- 1.2.1.4 Serum TSAT or content of Hb rocyte count) increases after 8 hours of storage at in reticulocytes (CHr) to as- room temperature and after 24 hours when refrig- 86 sess adequacy of iron for erated. The degree of MCV increase after erythropoiesis. sample storage may erroneously elevate the resulting Hct value by as much as 2 to 4 Hct 87 BACKGROUND percentage points. Long sample shipping dis- tances and sample transit times, as commonly Although erythropoietin deficiency is com- required for US dialysis facilities, increase the mon among patients with anemia and CKD, risk for introducing error into Hct results. other potential causes and potentially contrib- Hyperglycemia is associated with an increase uting disorders should be identified or ex- in MCV (but not Hb level) and therefore spuri- cluded. The recommended laboratory evalua- ously elevates the Hct result.88,89 tion provides information regarding the degree The coefficient of variation for same-sample and cause of anemia, activity of the erythroid Hct is greater than that for Hb, largely because of and nonerythroid marrow, and assessment of variability in the number and size of erythrocytes iron stores and iron availability for erythropoi- counted to calculate the Hct.90 Within-run and esis. In particular, clinicians should consider between-run coefficients of variation in auto- causes of anemia other than erythropoietin mated analyzer measurements of Hb are one half deficiency when: (1) severity of the anemia is and one third those for Hct, respectively.91 disproportionate to the deficit in renal func- For these reasons, Hct is an unacceptable test tion; (2) there is evidence of iron deficiency, or to evaluate anemia, and Hb should be the stan- (3) there is evidence of leukopenia or thrombo- dard measure for assessing anemia. cytopenia. An evaluation of the cause of ane- Hb should be measured on standardized auto- mia should precede initiation of ESA therapy. mated blood count analyzers in an accredited This guideline provides a framework for under- laboratory. In patients with nonÐdialysis-depen- standing the tests used in the initial evaluation dent (ND) CKD (ND-CKD) and patients with of anemia and sheds light on the utility of test PD-dependent CKD (PD-CKD), the timing of results in evaluating iron status. the blood sample draw is not critical because

S28 American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S28-S32 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S29 plasma volume in these patients remains rela- When these findings are present, further diagnos- tively constant. However, in patients with HD- tic evaluation may be indicated. Hypochromia dependent CKD (HD-CKD), interdialytic weight (low MCH) likely reflects longstanding iron- gain contributes to a dilutional decrease in Hb deficiency erythropoiesis. level, whereas intradialytic ultrafiltration pro- In general, the anemia of CKD is normochro- motes a contractional increase in Hb level. Thus, mic and normocytic; that is, morphologically a sample obtained immediately before dialysis indistinguishable from the anemia of chronic and during volume expansion underestimates the disease. It characteristically is hypoproliferative: euvolemic Hb level, whereas a sample obtained erythropoietic activity is low, consistent with immediately after dialysis overestimates the euv- insufficient erythropoietin stimulation. Prolifera- olemic Hb. In a study of 68 stable HD patients, tive activity is assessed by determination of the mean predialysis versus postdialysis Hb levels absolute reticulocyte count, the reticulocyte in- Ϯ Ϯ were 10.5 1.3 and 11.5 1.3 g/dL, respec- dex, and the reticulocyte production index. The 92 tively. There was a strong linear inverse corre- normal absolute reticulocyte count ranges from lation between percentage of change in Hb and 40,000 to 50,000 cells/␮L of whole blood. The Hct values and percentage of change in body reticulocyte index is calculated from the ratio of weight, indicating that increments in Hb and Hct observed to normal reticulocyte count. Thus: values after dialysis were a direct function of the intradialytic variation in body weight and degree of ultrafiltration. Moreover, the predialysis Hb Reticulocyte index level measured after a long interdialytic period ϭ ͓observed absolute reticulocyte count͔ (3 days) was 0.5 to 0.6 g/dL less than the predi- Ϭ ͓normal absolute reticulocyte count͔ alysis Hb level measured after a short interdialytic period (2 days). Among all pre-HD and Conversely, the reticulocyte production index post-HD Hb values, levels measured at the end corrects for the effects of erythropoietin-stimu- of short intervals were closest to the mean Hb lated early release of reticulocytes from the bone value of the week, derived from calculation of marrow. Early release shortens the fraction of the area under the curve.93 time reticulocytes mature in marrow and propor- Given the relationship between Hb level and tionally prolongs their maturation time in circula- interdialytic weight gain in patients with HD- tion. The result of early release is an increase in CKD, midweek predialysis sampling theoreti- total reticulocyte count. However, that increase cally is optimal. However, information that pa- reflects the premature shift to the circulation, not tient outcomes are affected by sampling day in increased erythroid production.94 Normal matu- patients with HD-CKD is lacking. Moreover, ration time in circulation is 1 day. The expected logistical considerations preclude sampling all maturation time, presuming a sufficient erythro- patients with HD-CKD at midweek. Therefore, poietin response to anemia, increases to 1.5 days sampling for Hb determination should be per- at Hb values between 10 and 13 g/dL, 2 days at formed before dialysis without specific reference values between 7 and 10 g/dL, and 2.5 days at to dialysis day. In the interpretation of results, values between 3 and 7 g/dL. The reticulocyte consideration should be given to the potential production index is calculated by dividing the effect of the patient’s volume status. reticulocyte index by the expected maturation In addition to Hb, other reported results of the time. Thus: CBC may convey important clinical information.

Deficiency of folate or vitamin B12 may lead to macrocytosis, whereas iron deficiency or inher- Reticulocyte production index ited disorders of Hb formation (␣-or␤-thalasse- ϭ ͓reticulocyte index͔ mia) may produce microcytosis. Macrocytosis Ϭ ͓expected maturation time͔ with leukopenia or thrombocytopenia suggests a generalized disorder of hematopoiesis caused by In an anemic patient, a reticulocyte production toxins (eg, alcohol), nutritional deficit (vitamin index greater than 3 is evidence of a normal

B12 or folate deﬁciency), or myelodysplasia. proliferative response to anemia, whereas a pro- S30 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

Fig 12. Distribution of ferritin and TSAT values by age group in males and females. Data are given as 5th, 50th, and 95th percentiles. Centers for Disease Control and Prevention; National Center for Health Sta- tistics, 2005.40 duction index of 2 or less is regarded as hypopro- ment of the adequacy of iron supply for erythro- liferative, ie, little or no effective response. Al- poiesis generally requires results of either TSAT though there is significant between-patient vari- or, when available, CHr. We further recommend ability in absolute reticulocyte count, the test is testing Hb, ferritin, and TSAT or CHr together sufficiently useful to serve its intended purpose because the combination provides important in- as a semiquantitative marker of erythropoietic sight into external iron balance and internal iron activity. The specific utility of the reticulocyte distribution. The distribution of ferritin and TSAT production index for the diagnosis and manage- by age and sex in the general population is ment of anemia in patients with CKD has not shown in Fig 12. been evaluated. Testing iron status before treating anemia in By contrast, erythropoietin levels are not rou- patients with CKD serves 2 purposes: to assess tinely useful in distinguishing erythropoietin de- the potential contribution of iron deficiency to ficiency from other causes of anemia in clinical 18,19,85,95,96 anemia and determine whether further evalua- settings. tion for sources of gastrointestinal (GI) bleeding Evaluating Iron Status in Anemic Patients is needed. Iron status testing is poorly suited to With CKD serve a third purpose, the prediction of respon- Iron status test results reflect either the level of siveness to iron therapy, because no single iron iron in tissue stores or the adequacy of iron for test or combination of tests discriminates iron- erythropoiesis. Serum ferritin level is the only responsive from iron-unresponsive patients and available blood marker of storage iron. Tests that because patients respond to intravenous (IV) iron reflect adequacy of iron for erythropoiesis in- therapy even when iron status results are substan- clude TSAT, MCV, and MCH and the related tially greater than the range associated with evi- indices, percentage of hypochromic red blood dence of iron deficiency (Guideline 3.2). cells (PHRC) and content of Hb in reticulocytes The potential contribution of iron deficiency (CHr). Because long sample transport and stor- to anemia is demonstrated best by examining the age times spuriously elevate PHRC results, the relationship between iron status test results and PHRC test is not suitable for routine use in many Hb difference from a reference value. Not surpris- patient-care settings. MCV and MCH decrease to ingly, the nature of the relationship depends on less than normal range only after long-standing whether the test measures iron stores or iron iron deficiency. Thus, timely and reliable assess- adequacy for erythropoiesis. ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S31

Fig 13. Distribution of Hb levels by ferritin range in patients with CKD. Results shown as -mean 95% ؎ CI for parameter estimates. Re printed with permission.26

The relationship between measures of iron difference from reference in patients with ND- stores (serum ferritin) and Hb difference from CKD is shown in Fig 14. The relationship be- reference is shown in Fig 13. In patients with tween TSAT and Hb values is continuous and ND-CKD, Hb values do not decrease to less than relatively linear throughout the encountered range reference level except at the lowest ferritin levels of TSAT values (Fig 14). There is no obvious (Ͻ25 ng/mL in males and Ͻ11 ng/mL in fe- threshold or cutoff value of TSAT below which males). These ferritin results closely approxi- the prevalence of anemia is sharply higher or the mate the fifth percentile of serum ferritin values level of Hb is sharply lower. Hb values are in a randomly selected population (males Ն 20 significantly lower than reference (ie, Hb results years, 33 ng/mL; females Ն 60 years, 15 ng/mL; seen in patients with a TSAT of 20% to 29%) US Renal Data System [USRDS] special data among both males and females only when TSAT request, NHANES 1999 to 2002 data set) and is less than 16%. By comparison, analysis of likely represent sex-specific thresholds at which TSAT values in a randomly selected survey popu- iron stores are exhausted in patients with ND- lation showed that the fifth percentile of TSAT CKD. results in males 20 years and older is 12.5%, and In patients with HD-CKD, the relationship that in females 60 years and older is 10.3% between serum ferritin and Hb values is less (USRDS special data request, NHANES 1999 to clear, probably because the relationship between 2002 data set). ferritin level and iron stores may be disturbed. In In short, in patients with ND-CKD undergoing relatively healthy patients with HD-CKD before evaluation for anemia, ferritin levels less than 25 widespread use of IV iron therapy, the finding of ng/mL in males and less than 12 ng/mL in a ferritin level less than 50 ng/mL was not females suggest that storage-iron depletion is uncommon97 and was associated with absent contributing to anemia. Serum ferritin level is bone marrow iron in approximately 80% of pa- less reliable in the evaluation of iron stores in tients.98 However, in patients with a greater patients with HD-CKD than in those with ND- disease burden, absent iron stores may still be CKD. Iron-deficiency erythropoiesis is most found at ferritin levels approaching or even ex- likely to contribute to anemia when TSAT results ceeding 200 ng/mL.99 are less than 16%. However, the clinical utility of The relationship between adequacy of iron TSAT is impaired by the absence of a diagnostic supply for erythropoiesis (from TSAT) and Hb threshold.

Fig 14. Distribution of Hb levels by TSAT range in patients with CKD. Results shown as -mean ؎ 95% CI for parameter estimates. Re printed with permission.26 S32 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

Although iron status tests provide reasonable and specificity of occult blood testing in the markers to detect iron deficits, there is little target population is lacking. Moreover, there is information to guide what action should be taken no information comparing the clinical utility of in response, particularly in regard to further occult blood testing with that of endoscopy in the diagnostic GI workup. In the absence of men- relevant patient population with iron deficiency, strual bleeding, iron depletion and iron defi- anemia, and CKD or in any target population ciency result from blood loss from the GI tract. with iron deficiency. Finally, there is no evidence GI pathological states—including ulcer disease, that either endoscopy or other additional GI colonic polyps, colorectal carcinoma, Helicobac- imaging improves patient outcomes when rou- ter pylori positivity, and sprue—are common tinely conducted for the evaluation of iron defi- among iron-deficient patients referred for esopha- ciency. Thus, although colonoscopy can be part gogastroduodenoscopy and colonoscopy.100-103 of an age-appropriate cancer screening and Testing for occult blood in stool is not recom- esophagogastroduodenoscopy should be strongly mended as part of the evaluation of iron-deficient considered if iron deficiency is confirmed in an patients.104,105 As a screening test for colorectal anemic patient with CKD, there is insufficient carcinoma or precancerous colonic polyps, stool evidence to recommend their routine use for occult blood testing commonly is performed, but anemia evaluation. a high false-negative result rate renders it rela- In conclusion, although epidemiological evi- tively unreliable.106 However, in a patient with dence yields a consistent definition of iron deple- iron deficiency, the test provides no useful infor- tion (ferritin Ͻ 25 ng/dL) and iron deficiency mation because a positive stool test result only erythropoiesis (TSAT Ͻ 16%), in the absence of confirms a diagnosis that the blood test results sufficient evidence that demonstrates therapeutic signifying iron deficiency have already estab- consequences based in these threshold values lished (GI blood loss), while a negative stool test and in the presence of information that therapeu- result can only be falsely reassuring. tic intervention should be guided by other factors In patients with CKD, occult blood in stool (see Guideline 3.2), these definitions lack spe- was found in 6% to 7% of patients with HD- cific recommendations for action and thereby fall CKD and PD-CKD and nearly 20% of patients short of constituting a guideline statement or with ND-CKD.107 Among patients with heme- CPR. positive stools, follow-up endoscopy identifies GI pathological states (predominantly gastric and Further Evaluation duodenal) in approximately 60% of patients. These tests are recommended only for the the However, because the likelihood of finding GI initial evaluation of anemia. Should initial evalu- lesions in patients with CKD with and without ation yield evidence for disorders other than heme-negative stools has not been examined, erythropoietin deficiency or iron deficiency, fur- critical information to establish the sensitivity ther evaluation is warranted. CPG AND CPR 2.1. HB RANGE

Treatment thresholds in anemia management de- evidence that encompassed both the RCTs previ- scribe the intended goal of current treatment for the ously cited and RCTs published since the previ- individual patient. The Hb treatment range repre- ous analysis. Because each RCT included iron sents the intended goal of ESA and iron therapy. agents as an adjunct to ESA, the guideline 2.1.1 Lower limit of Hb: includes references to both. In patients with CKD, Hb should be RATIONALE 11.0 g/dL or greater. (MODERATELY STRONG RECOMMENDATION) Hb Level Should Be 11.0 g/dL or Greater 2.1.2 Upper limit of Hb: Evidence to support a recommended target Hb In the opinion of the Work Group, there level demands the highest order of methodologic is insufficient evidence to recommend rigor. To determine the recommended target Hb routinely maintaining Hb levels at 13.0 level, we confined consideration to results gained g/dL or greater in ESA-treated patients. from RCTs that enrolled patients who are representative of the key CKD populations (anemia in BACKGROUND patients with ND-CKD, PD-CKD, and HD-CD), Previous (2001) KDOQI Guidelines for the assigned patients to at least 2 distinct target Hb Treatment of the Anemia of CKD2 incorporated level groups (treatment versus placebo/control or the following lines of evidence to support a higher versus lower Hb level), assessed out- recommended Hb target level of 11.0 to 12.0 comes that are important to patients, and re- g/dL in patients with CKD: ported results of between-group comparisons. We based our conclusions on evidence that, ● Observational evidence showed that mor- when taken together, reflected high or at least tality rates were lower in HD patients with moderate overall quality. Such evidence has few Hb values close to this range in compari- limitations, shows consistency among trials, bears son to HD patients with lower Hb levels. direct relevance to the anemic CKD population, ● Observational evidence in patients with and lacks sparseness or substantial bias. We gave HD-CKD and those with ND-CKD showed weight to between-group differences (effect sizes) an association between anemia and left that are not only statistically significant, but also ventricular hypertrophy (LVH). clinically meaningful, particularly when consid- ● A reasonable body of evidence from ran- ering QOL. When weighing evidence of im- domized controlled trials (RCTs) compar- proved outcomes against the potential for in- ing distinct Hb level targets suggested that creased risk, we sought to determine the Hb partial correction of anemia, to levels of level(s) at which a patient can expect an overall approximately 11 to 12 g/dL, led to im- net benefit. Finally, we required that the recom- proved QOL. mended target Hb statement be clear and unam- ● A limited body of evidence from distinct- biguous and the target Hb level be achievable in target RCTs suggested that partial correction practice. of anemia, to levels of approximately 11 to Evidence supporting the statement that Hb 12 g/dL, led to partial regression of LVH. level should be 11.0 g/dL or greater includes ● A single large RCT examining patients results from 22 RCTs and is presented both in with HD-CKD with patients with overt detail for each trial Table 12 through 19) and in CVD suggested that Hb level targets of 14 summary for each outcome ( Table 20 and Table g/dL do not improve the primary study 21). The evidence is confined to results of be- outcome of death or myocardial infarction tween-group comparisons generated by trials ran- (MI) and appear to jeopardize dialysis domizing patients to distinct intent-to-treat Hb vascular access compared with Hb levels targets, using “ESA versus placebo” or “ESA of 10 g/dL. lower Hb target versus ESA higher Hb target” To develop the current guideline, we undertook a designs (Fig 15). We excluded evidence from comprehensive reexamination of the available observational studies. Cohort-based observa-

American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S33-S53 S33 S34

Table 12. RCTs Examining Effects of Distinct Hb Targets/Levels on Key Clinical Outcomes in the HD-CKD and PD-CKD Populations

CKD Stage (Achieved) (%) zation Adequacy (%) Applicability Arm 3 (%)

ESA vs. ESA 14.0 *Nonfatal *29.6 ESA High ∆Kt/V: HD- (12.7-13.3) MI vs. 21 vs. 31 See QOL Besarab, 1998 108 1,233 10.2 14 — NS –0.03 vs. +0.06 CKD 3.1 vs. 2.3 24.4 P = 0.001 Table ESA Low 10.0 (10.0) P <0.001 NS c NS c 13.5-14.5 CVA: ESA High (13.3) 4 vs. 1 ∆URR: HD- See QOL Parfrey, 2005 109 596 11.0 24 P = 0.045 *NS NS — 0 vs. +2 % — CKD 9.5-11.5 Table ESA Low Other CVD: P <0.05 (10.9) NS ESA High 13-14 (13) HD- Kt/V: See QOL Foley, 2000 110 146 10.4 11 9.5-10.5 NS *NS NS — — CKD ESA Low NSh Table (10.5) 11.5-13 *High vs. ESA High 3 vs. 3 vs. (11.7) Low: ADULTS IN DISEASE KIDNEY CHRONIC IN ANEMIA 72 CanEPO, 1990- HD- NS 118 7.0 6 9.5-11 — — NS — — ESA vs. 1991 49, 117 CKD ESA Low (10.2) Placebo: ESA vs. P <0.05 f Placebo: Placebo (7.4) P = 0.024 4-5 13.5-16.0 ESA High *See PD- (3.6) Furuland, 2003 118 416 10.9 12 — — NS NS — — QOL CKD 9-12 ESA Low Table HD-CKD d (11.3-11.7) ESA High e <11 (8.7) 8 vs. 5 vs. 23 HD- ESA Low (8.2) See QOL Suzuki, 1989 119 179 6.3 2 — — — — — ESA vs. CKD Table Placebo (6.1) Placebo: P <0.05 Furuland, 2005 120 13.5-16.0 ∆Kt/V: HD- ESA High Substudy of Furuland, 24 11.1 5.5 (14.3) — — — — –0.1 vs. 0 — — CKD 2003 ESA Low 9-12 (10.9) nd McMahon, HD- ESA High e 14 (14) 14 8.5 1.5 --- — — — — — P <0.02 1999,2000 121, 122 CKD ESA Low 10 (10) NMAI HOI INYDSAEI ADULTS IN DISEASE KIDNEY CHRONIC IN ANEMIA

Arm 1 Clinical Outcomes (Arm 1 vs. Arm 2 vs. Arm 3) a Mean Hb (g/dL) Author, Year N Arm 2 CVD Target Mortality Hospitali- Dialysis Transfusion (mo) b Events LVH QOL Quality Hb (g/dL) Follow-up Baseline (%) CKD Stage (Achieved) (%) zation Adequacy Applicability Arm 3 (%) ESA vs. Placebo 10.7-12.7 ∆U/pt/4 wk ESA PD- (11.2) –0.21 vs. Nissenson, 1995 123 152 8 6-9 — — NS — — — CKD +0.42 Placebo (8.0) P <0.05 10-11.7 HD- ESA 9 vs. 60 Bahlmann, 1991 124 129 7.7 6 (10.6-10.9) NS — NS — — — CKD P <0.05 Placebo (7.8) 10-11.7 *LVEDd ESA e (11.3) (mm) HD- Sikole, 1993 125 38 6.7 12 — 48 vs. — — — — — CKD Control (8.3) g 53 P = 0.002 ESA vs. Placebo in Pediatric Patients PD- 10.5-12 ESA *See CKD (11.2) Morris, 1993 126 11 7.3 6 — — — — — — QOL HD- Placebo (7) Table CKD Footnotes: * Primary Outcome, if clearly indicated. a. All baseline data given for arm 1, unless otherwise specified. b. Global Scores, if documented, are provided here. Refer to Tables 13 and 19 for details of QOL measurements. c. The primary outcome was a composite of nonfatal MI or death. RR 1.3, 95% CI 0.9-1.8, P>0.05. d 294 HD-CKD, 51 PD-CKD, 72 ND-CKD patients. e. Iron cointervention not documented. f. Data at 8 weeks. g. Median. h. LVD subgroup: Kt/V 1.41 vs.1.50, P=0.025

Coding of Outcomes: Mortality: All-cause mortality. CVD Event: Includes CHF exacerbation, MI, arrhythmias, angina, interventional procedure such as CABG or angioplasty, sudden death, CVA. LVH: As identified by echocardiography with minimum of 6 month follow-up. S35 S36

Table 13. RCTs Examining Effects of Distinct Hb Targets/Levels on QOL in the HD-CKD and PD-CKD Populations

mrA 1 Mean Hb LOQ mrA( 1 .sv mrA 2 .sv mrA 3 ) CKD Stage Follow- Applic- Arm 2 (g/dL) Primary Outcome Author, Year N Global Quality Baseline up (mo) ability Target of Study Scale/Test Vitality and Fatigue Other Measures of QOL Arm 3 QOL Hb (g/dL) (achieved) KDQOL 5 ( )DH ASE H hgi 13.5-14.5 (13.1) subscales: SF-36 Vitality and SF-36 Vitality: + Social Interaction: NS Parfrey, 324 24 a LV volume index Quality of Social 2005 109 Interaction 11.0 ESA Low 9.5-11.5 (10.8) FACIT Fatigue SN Scale Physical: + ESA High 11.5-13 (11.7) Relationships: + KDQ Fatigue: + 5 (HD) + CanEPO, Depression: QOL and functional 1990- 118 6 b Frustration: NS ESA Low 9.5-11 (10.2) capacity c 1991 49, 117 Physical: + SIP + 7.0 Psychosocial: NS Placebo (7.4) OTT SN 14.0 Besarab, 5 ( DH ) ASE H gi h Mortality and 1,233 14 (12.7-13.3) SF-36 Physical: + d 1998 108 nonfatal MI 10.2 ESA Low 10.0 (10.0) Physical Symptoms: NS ADULTS IN DISEASE KIDNEY CHRONIC IN ANEMIA Depression: + 5 (HD) KDQ f Fatigue: + ESA High 13-14 (13) Relationships: + ∆LV mass index in Frustration: NS individuals with Physical Function: NS Foley, 94 concentric LVH; ∆ 12 e General Health: NS 2000 110 cavity volume index Bodily Pain: NS SF-36 Vitality: NS in individuals with Social Functioning: nd 10.1 ESA Low 9.5-10.5 (10.5) LVD Emotional Role: nd Mental Health: nd Mean HUI = 0.81 HUI (CI: 0.78, 0.85) 4-5 Physical: + ESA High 13.5-16.0 (13.6) Furuland, (PD,HD) Relationship: NS 253 12 QOL and safety KDQ Fatigue: NS g 2003 118 9-12 Depression: + 10.9 ESA Low (11.4) Frustration: NS g Several primary McMahon, )DH( 5 )DH( ASE High 41 1( 4) outcomes, including Physical: NS 1999, 14 1.5 SIP + QOL and exercise 2000 121, 122 8.3 ESA Low 10 (10) performance Psychosocial: + NMAI HOI INYDSAEI ADULTS IN DISEASE KIDNEY CHRONIC IN ANEMIA

Arm 1 QOL (Arm 1 vs. Arm 2 vs. Arm 3 ) CKD Stage Mean Hb Follow- Applic- Arm 2 (g/dL) Primary Outcome Author, Year N Global Quality Baseline up (mo) ability Target of Study Scale/Test Vitality and Fatigue Other Measures of QOL Arm 3 QOL Hb (g/dL) (achieved) Pediatric Patients 5 Physical Performance/ Sleep: NS ESA 10.5-12 (11.2) QOL, diet, exercise Morris, 1993 (PD,HD) 25-part Parental General Health Diet: NS 10 8 tolerance, and PD NS 126 Questionnaire h (includes school School Performance: NS 7.0 Placebo (6.5) efficiency attendance): + Psychosocial: NS Footnotes: a. No CVD or LVD. b. All individuals had LVD or LVH at baseline, no CVD. c. Data shown for ESA arms vs. Placebo. All statistical comparisons for ESA High vs. ESA Low were not significant. d. Increased by 0.6 point for each percentage point increase in Hct. e. Free of marked comorbidity. f. Results given are from repeated measures analysis of variance. g. P = 0.05 h. 25-Part Parental Questionnaire, modified from a previously used questionnaire126. Questions covered various aspects of the child’s well-being and behavior, including mood and psychological behavior, social interaction, somatic complaints and general health, sleep, diet, school functioning, and physical performance.

Coding of Outcomes: Coding of comparison of study arm 1 versus study arm 2: (+) better, (–) worse with reference to benefit for patient.

Key to QOL Measurement Scales/Tests: 36-item Medical Outcomes Study Short-Form Health Survey (SF-36): Evaluates eight health-related aspects: physical function, social function, physical role, emotional role, mental health, energy, pain, and general health perceptions. Each portion of the test is scored on a scale that ranges from 0 (severe limitation) to 100 (no limitation). Center for Epidemiologic Studies Depression Scale (CESDS): Has been used extensively in epidemiological studies of the general community and chronic disease populations. It is scored from 0 to 60, with higher scores indicating a greater number of depression symptoms. Functional Assessment of Chronic Illness Therapy (FACIT): Collection of QOL questionnaires targeted to the management of chronic illness Health Utilities Index (HUI): Provides an overall index of health, derived from scores in 7 aspects: sensation, mobility, emotion, cognition, self-care, pain, and fertility. This is an interval scale that can vary in theory between 0 (death) and 1 (perfect health). Karnofsky Performance Scale (KPS): Uses 10 steps with scores ranging from 100 (normal, without any limitation) to 10 (moribund) as an overall indicator of functional ability and self-sufficiency. It is considered an objective quality-of-life indicator. Kidney Disease Quality of Life (KDQOL): Validated in dialysis patients. The Short Form (SF) version used for assessment of vitality. Kidney Diseases Questionnaire (KDQ): Is validated in dialysis patients. Contains 26 questions divided into 5 sections: patient-specific physical symptoms, fatigue, depression, relationships, and frustration. All questions are scored on a 7-point Likert scale (7 = no problem, 1 = severe problem). Quality of American Life (QoAL): Asks the patient to rate his or her satisfaction with life on a scale from 1 to 7, with higher scores indicating greater satisfaction. Renal Quality of Life Profile (RQoLP): Instrument based on constructs representing renal patient’s own QOL determinants. Sickness Impact Profile (SIP): Is a behavior-related questionnaire that evaluates non–disease-specific, sickness-related behavioral dysfunction. It is widely used for end-stage renal disease patients and in studies evaluating quality-of- life improvement with ESA treatment for end-stage renal disease-related anemia. The SIP includes 136 items grouped into 12 categories of activity in physical and psychological dimensions. Scores range from 0 (no behavioral dysfunction) to 100 (100% dysfunction in a category or group). Unlike the KPS and the KDQ, lower scores for the SIP indicate better QOL. Time Trade-off (TTO): Is a unidimensional measure of QOL that gives a value to a patient’s QOL ranging from 1.0 (full health) to 0 (patient is indifferent between life and death). This is a utility measure using the time trade-off hypothesis. S37 S38

Table 14. Non-CVD/Mortality AE Rates in RCTs Examining Distinct Hb Targets/Levels in HD-CKD and PD-CKD Populations: ESA versus ESA

Arm 1 AEs (Arm 1 vs. Arm 2 vs. Arm 3)

Author, Description of a N Arm 2 BP Change or Hypertension Access Thrombosis (%) Seizures Other Reported AE (mo) Year Intervention Drug (g/dL) Target Dialysis Dialysis Modality Mean Hb Description and Follow-up (achieved) (achieved)

Arm 3 Definition Outcome Definition Outcome of D/C Total Results IV or SC ESA 14.0 618 1.5X pretrial dose; ESA High 0 Besarab, (12.7-13.3) Mean SBP and DBP Both AV graft and 39% vs. 29% 108 HD b adjusted after 2 wk 14 NS NS — 1998 during the study c AV fistulae (P = 0.001) 10.0 615 IV or SC ESA adjusted ESA Low 0 (10.0)

13.5-14.5 284 ASE H igh AV fistulae, Overall treatment IV or SC ESA for 24 wk (13.3) Hypertension not permanent 23% vs. 19% emergent AE in ≥10% of Parfrey, 2005 109 HD to reach target then 24 NS — nd specified catheter, non-site (NS) patients: maintained for 72 wk 9.5-11.5 281 ESA Low specific embolism 96% vs. 94% d (10.9)

13.5-16.0 Individuals with at least 1 216 SC ESA TIW ESA High Complication in SAE NOS: 51% vs. 34 (13.4-14.3) 90 vs. 83 5% vs. 2% in HD Furuland, HD ∆ mean DBP from AV graft, AV 38.5% (NS) 118 e 12 mm Hg patients only —

2003 PD baseline fistulae or catheter Thromboembolic: ADULTS IN DISEASE KIDNEY CHRONIC IN ANEMIA SC ESA TIW or no 9-12 (P = 0.02) (NS) 200 ESA Low during study Event: 56 vs. 47 per arm 15 treatment (11.3-11.7) (NS) f

59 ESA giH h 11< )7.8( No. of AE NOS Suzuki, Increased dose of 5 vs. 4 vs. 1 58 HD IV ESA 3,000 IU TIW 2 ESA woL )2.8( — — — 6.7% vs. 8.3% nd h 1989 119 anti-HTN meds individuals vs. 1.7% per arm g 57 Placebo (6.1) SC ESA For LVH: 13-14 Mean SBP, DBP, 73 ESA high arm had a 24 ASE hgiH significant (13) during between 8% vs. 14% Foley, 2000 110 HD wk “ramping” phase. 11 SBP and AV access — — nd groups, and use of (NS) i 24 wk maintenance was 9.5-10.5 anti-HTN 73 ESA Low anti-HTN meds similar in both arms (10.5) For LVD: NS ESA 39 (11.6) 200 U/kg % of individuals with IV ESA TIW 56% vs. 52% Abraham, ESA increases in DBP ≥10 40 HD after HD session 2.5-4.5 (11.0) vs. 45% — — — — nd 1991 127 100 U/kg mm Hg and/or 25, 100 or 200 IU/kg (NS) ESA anti-HTN meds 42 (8.8) 25 U/kg NMAI HOI INYDSAEI ADULTS IN DISEASE KIDNEY CHRONIC IN ANEMIA

Arm 1 AEs (Arm 1 vs. Arm 2 vs. Arm 3)

Arm 3 Definition Outcome Definition Outcome Total D/C of Results 11.5-13 38 ASE hgiH 2 IV ESA (11.7) 5% vs. 5% # of Events: Nonspecific AEs: CanEPO Severe HTN AV graft or 0% vs. 5% vs. HD 100 IU/kg initial dose; 6 9.5-11 vs. 0% 7 vs. 4 vs. 1 63 vs. 61 vs. 65 per 1990 49 40 ASE woL required withdrawal j AV fistulae clotting 2.5% (NS) 2 titrated to achieve targets (10.2) (P = 0.01) (P = 0.01) arm k 40 Placebo (7.4) 0 HTN: Slightly more 41 ASE hgiH 41 )0.41( SBP >140 mm Hg, prevalent in Berns, 1999 128 HD ESA to maintain target 12 — — — — nd DBP >90 mm Hg; or Low vs. High 14 ESA Low 10 (10.1) ∆ anti-HTN meds. (NS) SC ESA 2x/wk 14 McMahon, 8 ASE iH hg Mean ABP for peak 0 if total dose <20,000 IU/wk; (14.0) 1999, 2000 day and nocturnal HD IV ESA TIW 1.5 NS — — — — (Crossover) 121, readings taken before if total dose >20,000 10 122 6 ESA Low and after HD 0 IU/wk (10) Footnotes: a. For “Other Reported AE” column, outcomes may be recorded in no. of events per arm, no. of events per patient, or % given heterogeneity in reporting. b. All individuals had evidence of CHF and IHD. c. Prestudy ABP had to be below 160/100 for 4 weeks prior to study. Subgroup analysis129: 31 patients; mean day & nocturnal BP readings for 24 hr were NS at baseline or at follow-up. d. Of these P >0.05 for all comparisons except headache was greater in the ESA High arm and skeletal pain and surgery were greater in ESA Low arm. e. Includes some predialysis patients, Stages 4-5 CKD. f. Thromboembolic events were defined by WHO classification. g. 7 of 10 continued treatment. h. Not documented per arm. 3 individuals receiving ESA discontinued treatment. i. Patients with ongoing access problems were specifically excluded. The event rates were small and study did not have enough statistical power to detect a moderate impact on access thrombosis; the proportion using natural fistulae in the Besarab study was 23% compared to 76% in this study. j. DBP was increased in patients on ESA compared to placebo. P = 0.001; no statistical difference between High ESA and Low ESA. k. Nonspecific events include: clotting of tubing in dialysis machine, flu-like symptoms, headache, red eye, epistaxis or hemorrhage, pain in chest, abnormal sense of taste, aches in bone and muscle.

Coding of Outcomes: (Variable per Column Description) Hypertension: Includes mean changes in SBP, DBP, MAP, increase in use of anti-HTN medications, difficult to control hypertension. Access Thrombosis: Synthetic grafts and fistulae. S39 S40 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S41 S42

Table 17. RCTs Examining Effects of Distinct Hb Targets/Levels on QOL in the ND-CKD Population

CKD rA m 1 Mean Hb (g/dL) Primary LOQ A( mr 1 sv . mrA )2 Stage Follow- Applic- Author, Year N Target Outcome of Quality Baseline up (mo) ability Other Measures of Arm 2 (achieved) Study Scale/Test Global QOL Vitality and Fatigue Hb (g/dL) QOL 3 4- SE A hgiH 21 - 31 )1.21( Physical Health: NS Change in LV SF-36 NS Roger, 2004 38 155 24 a Mental Health: NS 11.2 ESA Low 9-10 (10.8) mass RQOLP NS Home Management: NS Selected SIP Scales 4-5 ESA 1 1.7 1( 2.1 ) Alertness Behavior: NS Social Interaction: NS Revicki, Physical Function: + 1995 132 Health-related 35 12 Selected SF-36 Scales Energy: + Roth, 1994 81 QOLb Role Function: NS Health Distress: NS QoAL NS 8.9 Control (9.0) CESDS NS Sexual Dysfunction NS ADULTS IN DISEASE KIDNEY CHRONIC IN ANEMIA Interview Footnotes: a. Excluded patients with unstable or poorly controlled angina, severe CHF (grade III-IV), severe chronic respiratory disease, symptomatic peripheral vascular disease, or a created arteriovenous fistula. b. The interview incorporated dimensions of HRQL identified to cover the expected effects of ESA therapy and anemia in predialysis CKD patients based on review of medical literature and discussions with nephrologists and nurses. The intent was to comprehensively measure broad areas of functioning and well-being.

Key for QOL Scales: 36-item Medical Outcomes Study Short-Form Health Survey (SF-36): Evaluates 8 health-related aspects: physical function, social function, physical role, emotional role, mental health, energy, pain, and general health perceptions. Each portion of the test is scored on a scale that ranges from 0 (severe limitation) to 100 (no limitation). Renal Quality of Life Profile (RQOLP): Instrument based on constructs representing renal patient’s own QOL determinants. Sickness Impact Profile (SIP): Established generic health status measure of disability associated with chronic illness. Quality of American Life Survey (QoAL): Is a life satisfaction scale from Campbell et al. The Quality of American Life. New York, NY, Russell Sage Foundation, 1976. Center for Epidemiologic Studies Depression Scale (CESDS): Has been used extensively in epidemiological studies of the general community and chronic disease populations. It is scored from 0 to 60, with higher scores indicating a greater number of depression symptoms. ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S43 S44

Table 19. RCTs Examining Effects of Distinct Hb Targets/Levels on Exercise Capacity in the HD-CKD and PD-CKD Populations

CKD mrA 1 Mean Hb Q LO mrA( 1 .sv mrA 2 v s. rA m )3 Stage Follow- Applic- (g/dL) Primary Outcome of Author, Year N Arm 2 Quality Baseline up (mo) ability Target Study Scale/Test Description Results Arm 3 Hb (g/dL) (achieved) 13.5-14.5 5 )DH( ASE iH gh (13.1) Patients are asked to cover as much Parfrey, 2005 109 324 24 a LV volume index 6-min Walk Test distance in an enclosed corridor as they NS 11.0 ESA Low 9.5-11.5 (10.8) can in 6 minutes

Naughton Stress ( 5 ( H )D E AS hgiH .11 5 1- 3 ( 1 )7.1 Treadmill NS Test CanEPO, 1990- b QOL and functional 118 6 1991 49, 117 ESA Low 9.5-11 (10.2) capacity c Patients are asked to cover as much 7.0 6-min Walk Test distance in an enclosed corridor as they NS Placebo (7.4) can in 6 minutes

Peak heart rate NS ( 5 ( H )D SE A H igh 14 ( 1 )4 McMahon, Several primary outcomes, d 1999, 14 1.5 including QOL and Exercise Test Peak O2 consumption + 2000 121, 122 exercise performance 8.3 ESA Low 10 (10) ADULTS IN DISEASE KIDNEY CHRONIC IN ANEMIA Work done + Pediatric Patients 5 5.01 ASE 5.01 21- .11( )2 QOL, diet, exercise 2 nim- aw l nik g SN e (PD,HD) Exercise Tolerance Morris, 1993 126 10 8 tolerance, and PD Test efficiency 7.0 Placebo (6.5) Treadmill NS f Footnotes: a. No CVD or LVD. b. All individuals had LVD or LVH at baseline, no CVD. c. Data shown for ESA arms vs. Placebo. All statistical comparisons for ESA High vs. ESA Low were not significant. d. With cycle ergometer. e. Not a significant improvement but did improve over study time. f. Only 3 children completed the treadmill test.

Coding of Outcomes: Coding of comparison of study arm 1 versus study arm 2: (+) better, (–) worse (with reference to benefit for patient). ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S45 S46 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S47

tional trials and cross-sectional analyses of large medical databases, such as within-group analyses in RCTs, consistently show that higher achieved Hb values (including Ն12 g/dL) are associated with improved patient outcomes, including lower mortality, less frequent hospitalization, and less severe LVH. Conversely, distinct- target RCTs consistently show that patients assigned to a Hb value of 13 g /dL or greater show no discernable improvement in survival, hospitalization, or LVH compared with patients assigned to Hb targets less than 13 g/dL and may be prone to excess adverse cardiovascular events. The failure of observational associations to be conﬁrmed by interventional trials renders use of observational evidence unsuitable to support the development of an intervention guideline statement. The statement that Hb level should be 11.0 g/dL or greater incorporates evidence from Hb targets ranging from 6 to 16 g/dL (Table 12 through Table 21; Fig 15). RCTs conducted before 1998 are characterized by small study size, higher target Hb levels within the range of 10 to 13 g/dL, and a lower target Hb level that reﬂects assignment to placebo or no-treatment control. Trials conducted thereafter are characterized by larger study size, higher target Hb levels within the range of 12 to 16 g/dL, and lower target Hb levels between 9 and 12 g/dL. The lesser magnitude of between-group Hb level differences seen in more recent trials is associated with Hb baseline values that are greater than those of early trials and lower target Hb level ranges that are substantially greater than Hb levels previously seen in placebo or untreated controls. In addition, mean Hb levels achieved in more recent trials frequently are at or less than the lower limit of the higher target Hb level range. Patients in the 19 RCTs reviewed to support the statement that Hb level should be 11.0 g/d or greater clearly are representative of the anemic patients with CKD that the statement intends to address. Ten RCTs enrolled patients with HD- CKD, 1 enrolled patients with PD-CKD, 2 enrolled patients with both HD and PD-CKD (Tables 12 to 14), and 9 enrolled patients with ND-CKD (Tables 15 and 16). However, the preponderance of evidence (and thus the greatest strength of evidence) lies in the HD-CKD patient group. Although sample size and thus power, S48

Table 21. Target Hb Levels in the ND-CKD Population

No. of Directness of Summary of Findings Studies & Total N Methodological the Evidence, Quality of Study of Quality of Consistency including Other Evidence for Importance Outcome Design Patients Studies across Studies Applicability Considerations Outcome Qualitative Description of Effect Size of Outcome All-Cause Some No important Some 5 RCTs 336 Sparse data c Very low Unable to assess benefit or harm with any certainty High Mortality limitations a inconsistencies uncertainty b Nonfatal Some No important Some 3 RCTs 249 Sparse data c Very low Unable to assess benefit or harm with any certainty High CV Events limitations d inconsistencies b uncertainty e No statistically significant benefit. Difference between No important No major arms is not statistically significant. However, a trend LVH 2 RCTs 307 No limitations None High Moderate inconsistencies uncertainties toward LV growth in those maintained at the lower Hct. Likely some benefit. One study (achieved Hb 12.1 vs. 10.8 g/dL*) showed no improvement in QOL with early Some Some QOL 2 RCTs 238 N/A g Sparse data Low ESA intervention. A second study (achieved Hb 10.5 High limitations f uncertainties h vs. 8.6 g/dL*) showed statistical improvement in the energy subscale, no overall score was provided. Unable to assess benefit or harm with any Transfusion Major Some certainty.There were more transfusions in the control 1 RCT 83 N/Aj Sparse data Very low Moderate Requirement limitations i uncertainty k arm (23%) than in the ESA arm (9%) but this difference was not statistically significant. Unable to assess benefit or harm with any certainty. The larger studies failed to show worsening of the rate ADULTS IN DISEASE KIDNEY CHRONIC IN ANEMIA of kidney disease progression in the ESA-treated arms. However, 2 very small studies suggested that Kidney Some to major Important Some dialysis started earlier in the occasional patient treated Disease 9 RCTs 502 None Low High limitations l inconsistencies m uncertainty n with ESA, but no statistics were given. The 2 largest, 1 Progression study showed statistically significant prolonged renal survival in the ESA treated group which was larger in the subgroup without diabetes than in the group with diabetes. Major Some Likely no harm. 0 vs. 2 individuals* in all studies noting Seizures 3 RCTs 39 N/A j Sparse data Very low Moderate limitations o uncertainty p seizures. Potential harm. Studies either showed no significant Blood change in BP with ESA treatment or there was a Major Some Some Pressure 7 RCTs 469 None Low statistically insignificant trend towards patients who Moderate limitations q inconsistencies r uncertainty s Change received ESA needing more antihypertensive management. Balance of Benefit and Harm: Quality of Overall Evidence: Net Benefit ≥at11 Hbg/dL based on QOL; Uncertain Trade Off with Increasing Hb Levels, QOL Benefit May Be Low Offset by Potential for Harm ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S49

effect size, and strength of evidence generally are lower in patients with ND-CKD than patients with HD-CKD or PD-CKD, the direction of effect for key outcomes appears to be the same. Accordingly, the guideline statement is designed or quality, studies with to address all anemic patients with CKD regardless of the presence, absence, or mode of dialysis linical relevance, inconsistent

ven poor quality of reporting. therapy. Endpoints measured to support the statement that Target Hb levels should be maintained at rtained as the primary outcome. Refer to Quality 1 study, few studies of po greater than 11 g/dL reﬂect key clinical outcomes important to patients. Measured outcomes include all-cause mortality, nonfatal cardiovascu- ncies is difficult to assess gi to difficult ncies is

ed with variable and uncertain c ed lar events, LVH, hospitalization, QOL, transfu- ort duration to measure outcome. secondary outcomes of the studies.

cal significance of effect.. effect.. cal significance of sion requirement, access thrombosis, other throm- lled in some of the studies. lled in some of the studies.

and 4 Grade C studies. boembolic events, seizure, blood pressure change, B and 3 Grade C studies. dialysis adequacy in patients with HD-CKD, and kidney disease progression in patients with ND- CKD (Table 12 and 16). Accordingly, the Work Group rated the importance of these outcomes as reporting on BP as an outcome. few subjects or no studies with the outcome asce and Directness grades. and Directness grades. high or moderately high (Table 20 and Table 21). j. to 1 of the following: Unable to assess due only B Grade 2 endpoint. k. Secondary A, Grade l. 3 m. Inconsistency in the statisti n. BP and diet tightly contro o. 2 studies of low quality and sh p. Seizures were not primary or Grade 1 A, Grade q. 3 r. The importance of the inconsiste s. Different outcome measures us Most distinct-target RCTs clearly have inadequate power to compare AE rates between pa- d to tients in lower and higher Hb target groups for in any of

s) and 1 the key safety outcomes, mortality, MI, and cerebrovascular events (Table 12 and 16). There are 2 exceptions. In patients with HD-CKD with CVD, patients assigned to a Hb target level of 14 g/dL showed increased risk for death or MI compared with patients assigned to an Hb level of 10 not attributed to intervention g/dL,108 a ﬁnding that did not reach signiﬁcance (relative risk [RR], 1.3; 95% CI, 0.9 to 1.9). in none of the RCTs was morbidity attribute However, safety concerns prompted early study termination. In a second trial of patients with HD-CKD without symptomatic heart disease or ically relevant, separation of Hb levels between arm left ventricular dilation, patients assigned to a Hb target of 13.5 to 14.5 g/dL showed a greater rate a primary endpoint; mortalitywas endpoint; a primary of cerebrovascular events than those assigned to an Hb target range of 9.5 to 11.5 g/dL (P ϭ 0.045).109 The risk for vascular access thrombosis in

gnificant, but perhaps not clin patients with HD-CKD likely increases as the target Hb level increases, over a wide range of udies to evaluate the outcome. ralizable as mortality was not endpoints; ralizable since CV events were secondary potential Hb level targets. Patients assigned to ed to asses QOL parameters. ansfusions were not defined.

and 1 Grade studies. Hb targets of either 9.5 to 11 or 11.5 to 13 g/dL show increased access thrombosis compared with placebo-treated controls,49 as do patients assigned to a Hb target of 14 g/dL compared with those assigned to a target of 10 g/dL.108 Native the studies. intervention. Grade B study. arteriovenous ﬁstulae fare no better than arterio- c. st Lack of adequately powered f. 1 Grade A study (with a statistically si d. 2 Grade A 1 B studies. e. Results are probably not gene a. 2 Grade A, 2 Grade B Grade Footnotes: 2 A, ESA High vs. * Low or Treatment vs. Placebo/Control Grade a. 2 b. Results are probably not gene g. Difference scales/tests us scales/tests g. Difference h. QOL was a primary endpoint of only 1 study. i. Trigger or threshold for tr venous grafts.49,108 S50 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

Fig 15. Target and achieved Hb levels in 18 RCTs comparing higher against lower (or placebo/control) Hb targets for ESA therapy. Whiskers denote the assigned (intent-to-treat) Hb target, data points denote achieved Hb level (see legend). Not shown is Lim 1989, Berns 1999, Abraham 1991 (see Tables 12 and 14). For literature citations, see Table 12 through Table 21.

Substantial blood pressure increases requiring fusions less frequent among patients with HD- intervention are seen in ESA-treated patients CKD in higher-target compared to lower-target regardless of Hb target compared with those treatment groups (Table 20). assigned to placebo or no-ESA control groups. In patients with ND-CKD, the effect of Hb Increased antihypertensive use to achieve equiva- target on progression of kidney disease is un- lent blood pressure also has been seen in patients clear. Eight RCTs enrolling more than 500 total assigned to higher Hb targets compared with patients yielded inconsistent results. Individual lower Hb targets.109,110 However, use of placebo trials showed either prolongation of kidney sur- and no-treatment control groups is confined to vival, acceleration of progression to kidney fail- early trials in which patients entered with low ure, or no effect, but many of the trials were baseline Hb levels and, with treatment, showed underpowered to detect potentially relevant ef- large relative differences between achieved Hb fects in either direction. Because of this limita- target (upper or lower) and baseline Hb levels tion, the unavoidable use of compound interven- (Fig 16). Blood pressure effects between upper tions (early versus late plus lower versus higher and lower Hb targets appear less prominent in Hb targets), and the lack of consistency in re- recent trials compared with earlier trials; how- sults, the Work Group concluded that consider- ever, it is unclear whether this difference is ations of kidney disease progression currently caused by relatively high baseline Hb levels, should not influence the determination of Hb small differences between baseline and either targets in patients with ND-CKD. upper and lower target achieved Hb levels, small QOL outcomes—particularly those reflecting differences between upper and lower achieved vitality, fatigue, and physical function—show a target levels, or differences in patient selection consistent positive relationship to level of Hb between early and more recent trials. target that appears to be continuous within the Hospitalizations and seizures appear not to be examined Hb level range of 6 to 16 g/dL. Pa- affected by target Hb level in patients with either tients report greater vitality, less fatigue, less HD-CKD, PD-CKD, or ND-CKD (Table 20 and depression, and improved physical symptoms at Table 21). Dialysis adequacy is lower and trans- higher compared with lower Hb targets. These ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S51

Fig 16. Baseline mean pretreatment Hb levels compared with achieved mean Hb levels in the upper and lower target. Results for achieved lower target Hb levels are not shown for placebo/control groups. Data for Lim 1989 (4 not shown. In recent trials, baseline Hb (14 ؍ and Kleinman 1989 (n ,(24 ؍ total), Berns 1999 (n 14 ؍ target groups, n level approaches or exceeds the achieved target Hb level of early trials, and differences between baseline and target upper achieved Hb levels are relatively small. For literature citations, see Table 20 and Table 21. results appear to be instrument dependent be- tice, (2) are ill suited to the purpose of measuring cause differences are more readily discernable clinical performance, and (3) if adopted as a with testing instruments specific for CKD. Other CPG, would discourage flexibility in meeting the dimensions of QOL show less consistent results needs of individual patients. Moreover, because regardless of the instrument used. Although ef- the relationship between QOL and Hb level that fect size of QOL outcomes and the quality of emerges from RCT results is continuous, there is evidence are lower in ND-CKD compared with no specific threshold Hb value to distinguish HD-CKD and PD-CKD trials, the direction of between the presence and absence of benefit. effect is similar in all populations tested. The The Work Group considered, but rejected, Work Group rated QOL as highly important to identifying a discrete Hb value (eg, 11.0 g/dL) as patients (Table 20 and Table 21). a Hb target. A discrete target Hb level affords In developing the statement Hb level should clarity and simplicity, but is almost universally be 11.0 g/dL or greater, the Work Group con- impossible to achieve, renders implementation cluded that—when comparing higher with lower as a performance measure difficult, and discour- Hb targets—QOL is a sufficient and, apparently, ages flexibility in using tradeoff decision making the sole determinant of treatment benefit. How- to meet the needs and preferences of individual ever, both the design of the RCTs supporting that patients. conclusion and the continuous nature of the Similarly, the Work Group considered, but relationship between QOL and Hb level pose rejected, identifying a target Hb level bounded challenges to translating evidence into a guide- by narrow upper and lower values (eg, 11.0 to line statement. Hb targets in RCTs are designed 12.0 g/dL). Such a target affords neither clarity to ensure separation of treatment results in higher nor simplicity, is possible to achieve in only a Hb target groups compared with lower Hb target minority of patients,111-113 discourages flexibil- groups to maximize the possibility that between- ity in managing individual patients, and likely group treatment effects will be detected. These promotes cycling of Hb results greater than and targets, whether represented as discrete Hb val- less than the target.114 ues (eg, 10.0 versus 14.0 g/dL108) or a range of The Work Group chose to define a broad Hb values (eg, 9.0 to 10.0 versus 12.0 to 13.0 therapeutic range of Hb levels bounded by a g/dL38), (1) were not intended to serve as Hb discrete lower threshold above which Hb level targets for anemia management in clinical prac- should be maintained and a upper threshold S52 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS above which Hb should not be routinely main- target (9.5 to 11.5 g/dL).109 Two additional trials tained. This approach makes clear to the practitio- compared the efficacy and safety of Hb targets in ner that no patient should be maintained intention- patients with ND-CKD, Cardiovascular Risk Re- ally at less than the lower threshold, and patients duction by Early Anemia Treatment With Epo- should not be routinely maintained at greater etin Beta Trial (CREATE; Hb level of 10.5 to than the upper threshold. We chose 11.0 g/dL as 11.5 versus 13.0 to 15.0 g/dL)115 and Correction the lower Hb level threshold. Hb values at or of Hemoglobin and Outcomes in Renal Insuffi- near 11.0 g/dL comprise an approximate water- ciency (CHOIR; Hb level of 11.3 versus 13.5 shed, defining the upper Hb targets of early RCTs g/dL).115A The CREATE trial has been com- and the lower targets of more recent trials pleted, whereas the CHOIR trial has been termi- (Fig 15). Thus, evidence to support the lower nated early by decision of the trial safety commit- limit of Hb statement draws on the full range of tee. Results from either trial are not published at RCTs available and the full range of Hb targets the time of this writing. examined. This evidence supports the conclusion In short, efforts to gain QOL benefits by treat- that patients treated to a Hb target greater than ing to Hb targets of 13 g/dL or greater are 11.0 g/dL likely will experience measurable QOL countered by increased risk for life-threatening benefits with little or no increase in AEs com- or disabling AEs. The statement that There is pared with treatment at lower Hb levels. insufficient evidence to recommend routinely Hb values greater than 11 g/dL are readily maintaining Hb levels at 13.0 g/dL or greater achievable. Greater than 80% of prevalent pa- recognizes the concern that, for most patients, tients with HD-CKD in the United States cur- the known risks outweigh the known benefits of rently demonstrate Hb levels of 11.0 g/dL or treating to higher Hb level. The phrase insuffi- greater.28 Among patients with a Hb level less cient evidence reflects the finding that published than 11.0 g/dL, greater than 99.6% achieve Hb results from comparative trials large enough to values of 11.0 g/dL or greater within 6 months examine safety are relatively limited. when prescribed sufficient ESA. Thus, the guide- Moreover, the word routinely reflects the cru- line as stated can be implemented readily in cial limitations of currently available evidence clinical practice and provides a rational founda- and the need for further investigation. For ex- tion for developing clinical performance mea- ample, no clinical trial conducted to date has sures for anemia management. examined target Hb safety and efficacy in selec- In the opinion of the Work Group, there is tive subgroups that may especially benefit from insufficient evidence to recommend routinely higher Hb targets, including, for example, pa- maintaining Hb levels at 13.0 g/dL or greater in tients who live at high altitude, are employed, ESA-treated patients. have pulmonary disease, have high levels of Evidence reviewed in the foregoing rationale physical performance, or undergo daily or noctur- supports the conclusion that QOL benefits associ- nal HD. In the absence of information in these ated with treating to a higher Hb target extend and other potential subgroups, medical decision throughout the full range of Hb targets examined making requires weighing known risks and ben- to date, spanning 6 to 16 g/dL. However, the efits in mixed-study populations against un- same evidence suggests that treating to Hb tar- known risks and unknown benefits in individual gets greater than 13 g/dL may increase the risk patients. for serious AEs. The only trial that had sufficient The statement Routinely maintaining patients power to examine safety assigned patients to a at Hb levels of 13 g/dL or greater makes it clear target Hb level of 14.0 g/dL in the higher-target that the recommended upper limit of Hb level treatment arms108 and compared outcomes with applies only to patients considered for maintain- those seen in patients assigned to a lower Hb ing a Hb level at 13 g/dL or greater as a target (10.0 g/dL). That trial was discontinued treatment target, not patients for whom Hb level for safety concerns. In a second trial, patients is likely to transiently equal or exceed 13 g/dL assigned to a higher Hb target (13.5 to 14.5 g/dL) while undergoing treatment for a lower Hb tar- showed an increased incidence of cerebrovascu- get, and not patients who do not require ESA lar AEs compared with those assigned to a lower therapy. Given the biological variation in Hb ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S53 levels in patients with HD-CKD, if the intent of comparative safety and efficacy of lower versus treatment is to maintain Hb levels at 11.0 g/dL or higher Hb targets in selected subgroups of pa- greater, a sizeable fraction of patients will neces- tients that may especially benefit from higher Hb sarily show Hb values transiently at 13 g/dL or targets. greater. Judicious ESA dose adjustments appear sufficient to correct Hb values above intended QOL/QOL Trade-Off 112 targets within 3 to 6 months. QOL, like quantity of life, is a hard patient LIMITATIONS outcome. Available evidence confirms that setting Hb targets for individual patients requires Limits of Available RCTs consideration of the potential trade-off between The strength of an RCT is to study a specific QOL and survival, MI, cerebrovascular event, or question by studying the efficacy of an interven- access thrombosis. However, although we know tion on specific outcomes in a well-defined popu- that risk assessment varies widely by culture, lation. Findings from RCTs thus are often limited personal values, level of education, and level of in their generalizability to diverse populations or medical literacy, we lack information on how to to questions related to effectiveness of combina- assist patients with CKD to interpret risk, judge tions of interventions on combinations of patient risk-benefit trade-offs, and express informed pref- outcomes. These latter questions are what guide- erences. Nevertheless, in everyday medical deci- lines attempt to address. For example, consider a sion making, it is important to incorporate in- hypothetical study designed to compare the ef- formed patient preference.116 fect of 3 Hb treatment targets: Hb level less than 9 g/dL, Hb level of 9 to 12 g/dL, and Hb level Differences in Treatment Strategies Used To greater than 12 g/dL. Patient populations should Achieve Defined Hb Targets be separated into patients with or without CVD and: (1) ND-CKD, (2) HD-CKD, (3) PD-CKD, The duration of anemia before initiating ESA or (4) transplant-associated CKD. Each discreet therapy and the size and rate of Hb level correc- Hb target, CVD status, and each CKD category tion each plausibly contributes to risk, benefit, or deserves scrutiny. However, this is unfeasible: both in anemia therapy. Information on each the hypothetical RCT would collapse under the would be directly helpful to clinicians. weight of a 3 ϫ 2 ϫ 4 (24-group) study design. Similarly, it is possible that the risks and Imagine instead constructing a 24-cell matrix benefits associated with a target Hb level depend to represent all possible combinations of Hb in part on which therapeutic agents were used in targets, CVD status, and CKD categories that we the treatment arm. All available hard-outcome wish to examine. If we added all the results from trials to date have used both ESAs and iron all available RCTs to the specific cells addressed agents together in patients assigned to both treat- by the available RCTs, we would have to leave ment and control arms. No trial using ESA alone, the vast majority of cells empty at this writing. iron alone, or noniron adjuvants alone has been Given the small number of RCTs and the small adequately designed and powered to confirm that size of most of the available RCTs, and given the similar Hb targets afford similar risks and ben- potential differences between target populations, efits regardless of how they are attained. The filling in empty cells by extending results from possibility that each potential class of agents, the few completed cells seems unwarranted and including ESAs, oral iron, parenteral iron, and likely to involve substantial error. non-iron adjuvants, shows differing patient out- Another critical limitation of the evidence, as comes for equivalent Hb outcomes has not been discussed, is the lack of information about the explored. CPR 3.1. USING ESAs

ESAs are critical components in managing the ment setting, efficacy, safety, anemia of CKD. Available ESAs are each effec- and class of ESA used. tive in achieving and maintaining target Hb lev- 3.1.3.2 In the opinion of the Work els. Aspects of administration may differ be- Group, convenience favors tween short-acting and long-acting agents. subcutaneous (SC) administration in non–HD-CKD pa- 3.1.1 Frequency of Hb monitoring: tients. 3.1.1.1 In the opinion of the Work 3.1.3.3 In the opinion of the Work Group, the frequency of Hb Group, convenience favors in- monitoring in patients treated travenous (IV) administra- with ESAs should be at least tion in HD-CKD patients. monthly. 3.1.4 Frequency of administration: 3.1.2 ESA dosing: 3.1.4.1 In the opinion of the Work 3.1.2.1 In the opinion of the Work Group, frequency of adminis- Group, the initial ESA dose tration should be determined and ESA dose adjustments by the CKD stage, treatment should be determined by the setting, efficacy consider- patient’s Hb level, the target ations, and class of ESA. Hb level, the observed rate of 3.1.4.2 In the opinion of the Work increase in Hb level, and clini- Group, convenience favors less cal circumstances. frequent administration, par- 3.1.2.2 In the opinion of the Work ticularly in non–HD-CKD Group, ESA doses should be patients. decreased, but not necessarily withheld, when a downward BACKGROUND adjustment of Hb level is The decision to start ESA therapy in a pa- needed. tient with CKD-associated anemia invokes an 3.1.2.3 In the opinion of the Work additional series of decisions that must be Group, scheduled ESA doses made before the first dose is administered. that have been missed should Each of the following issues must be consid- be replaced at the earliest pos- ered before moving forward: choosing an ESA, sible opportunity. an initial ESA dose, and the route and fre- 3.1.2.4 In the opinion of the Work quency of ESA administration; planning a Hb Group, ESA administration level monitoring schedule; predicting a desired in ESA-dependent patients rate of increase in Hb level; and anticipating should continue during hospi- ESA dose adjustments for Hb results, AEs, or talization. intercurrent hospitalizations. Making informed 3.1.2.5 In the opinion of the Work decisions requires understanding the method Group, hypertension, vascu- of action of ESAs, then selecting those which lar access occlusion, inad- best fit the patient, the patient-care setting, and equate dialysis, history of the constraints of the health care delivery sys- seizures, or compromised nu- tem from the range of choices that are known tritional status are not contra- to be safe and effective. indications to ESA therapy. 3.1.3 Route of administration: RATIONALE 3.1.3.1 In the opinion of the Work Group, the route of ESA ad- Definitions ministration should be deter- The term ESA applies to all agents that aug- mined by the CKD stage, treatment erythropoiesis through direct or indirect

S54 American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S54-S57 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S55 action on the erythropoietin receptor. Currently ESA once weekly, and consideration for dose available ESAs include epoetin alfa, epoetin beta, adjustment every 2 weeks, 44% to 49% of and darbepoetin. Epoetin alfa and beta have been patients will actually require dose adjust- designed to resemble closely the endogenous ment.138 molecule and have similar pharmacokinetics. They are considered “short-acting” in compari- ESA Dosing son to darbepoetin, a second-generation mole- The initial ESA dose and ESA dose adjust- cule with a prolonged half-life, which is consid- ments should be determined by the patient’s Hb ered “long-acting.” level, the target Hb level, the observed rate of increase in Hb level, and clinical circumstances. Frequency of Hb Monitoring This statement aims to guide medical decision The statement The frequency of Hb monitor- making so that the initial rate of increase in Hb ing in patients treated with ESA should be at levels is in keeping with the patient’s current least monthly is intended to provide sufficient status, the treatment setting, and the patient’s surveillance information to assist in achieving need for anemia treatment. In general, the objec- and maintaining target Hb levels safely and expe- tive of initial ESA therapy is a rate of increase in ditiously during ESA therapy. This recommenda- Hb levels of 1 to 2 g/dL per month. Evidence for tion, which fits common practice, follows a chain the safety of initial ESA therapy derives princi- of reasoning. The minimum interval between pally from interventional trials in patients not ESA dose adjustments is 2 weeks because the previously treated with ESAs. In patients with effect of most dose changes will not be seen CKD-associated anemia and initial Hb levels within a shorter interval. Consideration of an less than target range, these trials have shown ESA dose adjustment is based on the next pro- the mean initial rate of Hb level increase to be in jected Hb level. Because the accuracy of projec- the range of 0.7 to 2.5 g/dL in the first 4 weeks. tion (extrapolation) increases with the number of The rate of increase varies greatly and depends contributing data points, the frequency of Hb in part on the patient population, initial dose, monitoring is likely to be an important determi- dosing frequency, and route of administration nant of the accuracy of ESA dose adjustment. (Table 22). Hypertension and seizures were noted Evidence to support this line of reasoning is in the first 3 months after initiating therapy in indirect. Several RCTs have randomized HD severely anemic patients, but it is unclear whether patients with target-range Hb levels to a change these events occurred within the first 4 weeks in frequency of ESA administration, a change in and whether they were related to the rate of ESA class, or both. RCTs that have monitored increase in Hb levels. Hb values weekly and adjusted ESA doses as In general, ESA dose adjustments were not frequently as every 2 weeks have achieved stable made in the first 4 weeks of trials for which Hb levels early after randomization.136-138 A results are available. The frequency of ESA dose single RCT that monitored Hb levels and consid- adjustment thereafter should be determined by ered ESA dose adjustment monthly required 6 to the rate of increase in Hb levels during initial 9 months to stabilize Hb levels after randomiza- ESA therapy, the stability of Hb levels during tion,139 but mean Hb level remained within the maintenance ESA therapy, and the frequency of target range for that trial. Hb testing. The minimum interval between ESA Within the recommended ranges for monitor- dose adjustments in the outpatient setting gener- ing and dose adjustment, unstable Hb level, ally is 2 weeks because the effect of most dose out-of-target Hb level, and HD favor shorter changes will not be seen within a shorter inter- intervals, whereas stable Hb level, within- val. target Hb level, PD, ND-CKD, and minimizing ESA doses should be decreased, but not laboratory resource utilization favor longer necessarily held, when a downward adjust- intervals. The frequency of ESA dose adjustment of Hb level is needed. Withholding ESA ment is unaffected by length of action: during doses, particularly for long periods, may lead an 8-week period, given Hb monitoring weekly, to a delayed decrease in Hb levels to less than short-acting ESA thrice weekly or long-acting target range. Such a decrease may initiate S56 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

periodic cycling of Hb levels at greater than ESA therapy improves patient outcomes in these and less than the target Hb range.114 This settings. finding is in keeping with the mechanism of Route of Administration action of ESA in preventing apoptotic death of CFU-Es and early erythroblasts. Should ESA The route of administration should be deter- doses be withheld in an ESA-dependent pa- mined by the CKD stage, treatment setting, effi- tient, a prolonged loss of erythropoietic precur- cacy considerations, and the class of ESA used. sors may result. Accordingly, the Work Group Among patients with HD-CKD, either SC or IV recommends that ESA doses not be withheld administration is possible. In the outpatient set- routinely for Hb levels greater than target ting, SC administration is the only routinely range, hospitalization, poorly controlled hyper- feasible route of administration for patients with PD-CKD and ND-CKD. Among patients with tension, or vascular access occlusion. HD-CKD, either SC or IV administration is The statement Scheduled ESA doses that have feasible, but the risk for pure red cell aplasia been missed should be replaced at the earliest (PRCA) associated with SC administration, al- possible opportunity reflects the concern that beit small, recently prompted the US Food and missed doses of ESA, like held doses, may lead Drug Administration (FDA) to recommend the to a delayed decrease in Hb levels to less than IV route. The relationship between SC ESA target range. This is illustrated most clearly in administration and risk for PRCA experienced patients who receive ESA dosing every 2 to 4 outside the United States is discussed elsewhere weeks, but the principle applies to all patients (Guideline 3.5). undergoing ESA therapy. Among short-acting ESAs, efficacy of SC Similarly, the statement Hypertension, vascu- administration in patients with HD-CKD is supe- lar access occlusion, inadequate dialysis, history rior to that of IV administration. Although many of seizures, or compromised nutritional status trials have achieved similar results, the findings are not contraindications to ESA therapy aims to of a single large multicenter RCT are sufficient discourage the practice of withholding ESA to establish the strength of evidence for this therapy in the presence of conditions potentially guideline statement.140 In this trial, patients with linked to anemia treatment. If these disorders HD-CKD were eligible for randomization only if arise in the course of anemia therapy with ESA, they were currently receiving epoetin alfa and they should be treated appropriately with specific their Hct was within the target range. Upon measures. There is no evidence that withholding randomization (from SC or IV to IV or SC), ESA ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S57 doses were first decreased to allow Hb levels to class of ESA. Selecting frequency of ESA admin- decrease to less than target range. Doses then istration may require a choice between maxi- were titrated upward to again achieve target Hct mum efficacy and optimum convenience and levels, then were adjusted to maintain Hct in the comfort. Maximum efficacy occurs within dos- target range during a 26-week maintenance phase. ing intervals that are ESA class specific. For Among patients who completed the trial, those example, in patients with HD-CKD receiving SC assigned to SC administration showed 27% lower short-acting ESA therapy, ESA efficacy de- ESA doses than those assigned to IV administra- creases when the dosing is extended from thrice- tion. Mean achieved Hcts did not differ between weekly to once-weekly administration.139 When groups. Of the patients who, by assignment, every-2-week administration of long-acting ESAs switched from IV to SC administration, 58% is extended to every 4 weeks, efficacy either showed a decrease in ESA dose and 23% showed remains stable142 or decreases incrementally.143 an increase; among their counterparts who switched from SC to IV administration, the corresponding proportions were 28% and 49%, respec- LIMITATIONS tively. In short, SC administration of short-acting In practice, the use of ESAs is subject to wide ESAs is associated with an approximately 30% variability that reflects broad differences in prac- decrease in ESA dose for the same target Hb tice setting, patient preference, and reimburse- outcome. However, not all patients show a dose ment constraints. The definition of best practice decrease after conversion from IV to SC, and therefore should reflect both available evidence some will show a dose increase. and prevailing conditions, preferences, and priori- Among long-acting agents, efficacy of SC ties. The evidence base for CPRs on the use of compared with IV administration appears to be equivalent at examined dosing frequencies.137,141 ESAs requires information generated from quality improvement programs and anemia manage- Frequency of Administration ment protocols that examine multiple interven- Frequency of administration should be deter- tions and explicitly state underlying goals and mined by the CKD treatment setting and the assumptions. CPG AND CPR 3.2. USING IRON AGENTS

Anemia therapy in patients with CKD requires weigh ESA responsiveness, Hb and effective use of iron agents, guided by appropri- TSAT level, and the patient’s clinical ate testing of iron status. Efficacy of iron therapy status. appears not to be limited to patients with evi- 3.2.5 Route of administration: dence of iron deficiency. (See Guideline 1.2 for 3.2.5.1 The preferred route of diagnosis of iron deficiency.) Thus, the goals of administration is IV in iron therapy are to avoid storage iron depletion, patients with HD-CKD. prevent iron-deficient erythropoiesis, and achieve (STRONG RECOMMEN- and maintain target Hb levels. DATION) 3.2.1 Frequency of iron status tests: 3.2.5.2 In the opinion of the Work In the opinion of the Work Group, Group, the route of iron ad- iron status tests should be performed ministration can be either IV as follows: or oral in patients with ND- 3.2.1.1 Every month during initial CKD or PD-CKD. ESA treatment. 3.2.6 Hypersensitivity reactions: 3.2.1.2 At least every 3 months dur- In the opinion of the Work Group, ing stable ESA treatment or in resuscitative medication and person- patients with HD-CKD not nel trained to evaluate and resuscitate treated with an ESA. anaphylaxis should be available when- 3.2.2 Interpretation of iron status tests: ever a dose of iron dextran is In the opinion of the Work Group, administered. results of iron status tests, Hb, and ESA dose should be interpreted to- BACKGROUND gether to guide iron therapy. 3.2.3 Targets of iron therapy: The goal of iron therapy in a patient with In the opinion of the Work Group, anemia and CKD is to achieve and maintain a sufficient iron should be administered target-range Hb level. Iron agents may serve as to generally maintain the following primary therapy for selected patients (particu- indices of iron status during ESA larly those with ND-CKD) or as adjuvant therapy treatment: for those also undergoing treatment with an ESA. 3.2.3.1 HD-CKD: Administered as adjuvants to ESAs, iron agents ● Serum ferritin >200 ng/mL prevent iron deficiency and serve to minimize the AND dose of ESA needed to achieve target-range Hb ● TSAT >20%, or CHr >29 levels. Designing appropriate iron therapy for pa- pg/cell. tients with CKD-associated anemia requires an 3.2.3.2 ND-CKD and PD-CKD: understanding of the interpretation of iron status ● Serum ferritin >100 ng/mL test results, the therapeutic significance of iron AND status levels, and the therapeutic and safety limits ● TSAT >20%. of iron administration. Selecting a route of adminis- 3.2.4 Upper level of ferritin: tration, dose, and class of iron agent requires an In the opinion of the Work Group, understanding of the efficacy, safety, and tolerabil- there is insufficient evidence to recom- ity of available iron therapeutics. The stage of CKD mend routine administration of IV iron and treatment setting (HD-CKD and PD-CKD) if serum ferritin level is greater than deserve consideration not only for practical rea- 500 ng/mL. When ferritin level is sons, but also because the quantity, quality, and greater than 500 ng/mL, decisions re- strength of evidence vary among patients with garding IV iron administration should ND-CKD, HD-CKD, and PD-CKD.

S58 American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S58-S70 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S59

RATIONALE In this case, iron therapy may be indicated to replace iron deficits. Conversely, a decreasing Frequency of Iron Status Tests Should Be ferritin level in the presence of an increasing Hb Every 1 to 3 Months level signifies an internal shift in iron from stores The purpose of iron status testing is to either to Hb, as would be expected in a patient respond- evaluate anemia or guide the use of iron agents to ing to ESA therapy. If iron status remains within achieve and maintain targets of iron therapy the target range, additional iron administration during anemia management. Use and interpreta- may not be required. If the Hb level is greater tion of iron status tests in the initial evaluation of than the target range and is being corrected by anemia are described in Guideline 1.2. decreasing the ESA dose, ferritin level may in- Frequency of iron status testing during anemia crease as the Hb level corrects. Again, no iron management should be determined by the pa- may be needed. Finally, an increase in ferritin tient’s Hb level relative to the target range and by level accompanied by a decrease in TSAT sug- the likelihood of blood loss. Initiation of ESA gests inflammation-mediated reticuloendothelial therapy, correction of a less-than-target Hb level blockade and may be accompanied by a decrease during ongoing ESA therapy, bleeding, and surgi- in Hb level and increase in ESA dose. Whether cal procedures each are likely to promote deple- and to what extent iron administration is effec- tion of iron stores and thereby may call for tive and safe in this setting remains unresolved. increased frequency of iron status testing. Simi- In short, medical decision making in using larly, iron status testing is appropriate after treat- iron agents—including initiation of iron therapy, ment with a course of IV iron, to monitor safety choice of route of iron administration, and deter- and adequacy of therapy, and in a patient with mination of dose and duration of iron therapy— hyporesponsiveness to ESA. should be guided by results of iron status tests Clinical settings in which more frequent iron taken together with Hb levels and ESA dose, testing may be necessary include the following: combined with a trend analysis of how each has changed over time. 1. Initiation of ESA therapy 2. Correction of a less-than-target Hb level Targets of Iron Therapy during ongoing ESA therapy In this section, we distinguish the therapeutic 3. Recent bleeding use of iron status tests from their diagnostic use. 4. After surgery As diagnostic indices, iron status tests are helpful 5. After hospitalization in determining the likelihood that iron deficiency 6. Monitoring response after a course of IV is contributing to a low Hb level in patients with iron anemia and CKD (Guideline 1.2). In all patients 7. Evaluation for ESA hyporesponse. with CKD, as in non-CKD patients, evidence of iron depletion and iron deficiency should prompt Interpretation of Iron Status Test Results consideration of blood loss, and further evalua- Should Incorporate Hb and ESA Dose tion may be needed (Guideline 1.2). In a patient undergoing ESA therapy, interpre- By contrast the statement Targets of iron tation of the results of iron status tests should therapy refers to results of iron status tests as incorporate consideration of the Hb level and treatment targets for the safe and effective use of ESA dose. Together, these results provide infor- iron agents. In patients receiving ESA therapy, mation important to medical decision making iron status treatment targets serve to minimize because they elucidate the status of both external the ESA doses required to maintain target-range iron balance (net loss or gain of iron) and internal Hb levels. In patients not receiving ESA therapy, iron balance (disposition of iron between stores iron status treatment targets serve to maximize and circulating red blood corpuscles). For ex- the Hb level and minimize the need to initiate ample, a decreasing ferritin level in the presence ESA therapy to achieve target-range Hb levels. of a stable or decreasing Hb level may signify Targets of iron therapy reflect the treat-to-target external iron loss (GI bleeding or, in patients goals for the use of iron agents in patients with with HD-CKD, dialysis-associated blood loss). HD-CKD, ND-CKD, and PD-CKD with anemia. S60 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

Table 23. RCTs for Ferritin and TSAT Targets in the HD-CKD Population

esaB l eni eD s oitpirc n fo smrA emoctuO mrA 1 v .s rA m 2 Author, Target N Co-intervention Intervention Year Arm 1 (Achieved) (mo)

Definition Result Quality Follow-up Follow-up Applicability Ferritin Ferritin P-value (ng/mL) Arm 2 Hb (g/dL) TSAT (%) ESA Dose IV iron dextran 100 mg for 3,625 IU 4-6 HD sessions 30-50 Arm 1 23 10.6 2 58 6.42 giH h T TAS )%( TIW followed by (30.9-34.9) a approximately Besarab, ESA adjusted to 6 25-150 mg/wk ESA dose 40% less than 0.004 2000 148 maintain baseline Hb Weekly doses of IV iron arm 2 at 3,782 IU 20-30 19 10.5 287 23.9 Low TSAT (%) dextran months 4-6 b TIW (24.5-27.6) 25-150 mg/wk 312 400 19 9.8 661 nd ESA adjusted to H gi h Ferr niti n( g )Lm/ IV iron dextran DeVita IU/kg/wk (469) Mean decrease 154 vs. 31 5 maintain Hb of 10.8- if below target <0.001 2003 149 250 Low Ferritin 200 in ESA dose IU/kg/wk 17 10.2 204 nd 12 g/dL 100 mg for 10 HD sessions IU/kg/wk (ng/mL) (299) Footnotes: a. Range of mean TSAT for months 4-6. b. Arm 1: 52% vs. Arm 2: 21% had at least a 20% reduction in ESA dose (P <0.05).

Iron Targets in Patients With HD-CKD Other available evidence reflects the likeli- Evidence to support the development of an hood of responsiveness to IV iron, but affords iron target guideline in patients with HD-CKD is little information on the comparative safety and drawn from tissue iron studies (semiquantitative efficacy of iron status targets. Patients with high iron in bone marrow and quantitative iron in baseline ferritin levels (mean, 930 ng/mL) were liver), iron challenge tests (response to IV iron administered IV iron if TSAT was less than 50% administration), nonrandomized trials (efficacy and ferritin level was less than 1,000 ng/mL. of single-target IV iron intervention), and RCTs After receiving an average IV iron dose of 38 mg (comparing efficacy of 2 different iron targets). per week for 12 months, patients showed an No iron intervention trial has been sufficiently increase in ferritin (mean, 1,383 ng/mL) and powered to assess safety. Therefore, the Work TSAT (from 27% to 36%) values and a 25% ESA Group sought to recommend iron targets that dose decrease.150 Bone marrow examination in balance efficacy with assumptions regarding patients with HD-CKD may reveal little or no safety. stainable iron in patients with ferritin levels Serum ferritin level greater than 200 ng/ greater than 100 ng/mL,98 particularly in patients mL. Earlier versions of KDOQI guidelines rec- with evidence of inflammation.99 It has been ommended a target ferritin level greater than 100 reported that IV iron challenge is associated with ng/mL. Two RCTs provide comparative informa- a positive response (an increase in reticulocyte or tion on iron status targets for IV iron therapy Hb values or a decrease in ESA dose) in many (Table 23). In 1 RCT, patients with HD-CKD patients with serum ferritin levels greater than randomized to maintain TSAT at 20% to 30% 100 ng/mL (Table 24).151-156 achieved a mean serum ferritin level of 297 Taken together, the evidence suggests that ng/mL compared with those randomized to a greater efficacy may be achieved if the lower TSAT of 30% to 50%, who achieved 730 ng/ limit of the ferritin target were greater than the mL.148 Patients in the higher TSAT target group 200 ng/mL threshold. However, given the ab- showed a mean 40% reduction in ESA dose sence of reliable safety information in clinical compared with those in the lower group. How- trials, uncertainty about the relevance of results ever, the trial was small; iron status targets were from uncontrolled trials, bone marrow iron exami- defined by TSAT, not ferritin level; and ferritin nation, IV iron challenge studies to iron treat- level in the high-target TSAT group increased ment targets, and the concern for the risk for iron throughout the trial and had not reached a steady overload, the statement that serum ferritin levels state by the end of study. In the second trial, generally should be maintained at greater than patients with HD-CKD assigned to a 200-ng/mL 200 ng/mL reflects acceptable—if not maximal— ferritin target compared with those assigned to a efficacy and a cautious consideration for patient 400-ng/mL target achieved mean ferritin levels safety. of 299 compared with 469 ng/mL, respectively. TSAT greater than 20%. The recommended Patients in the higher serum ferritin group had treatment target of TSAT is 20%, similar to final ESA doses 28% lower than those in the previous KDOQI recommendations. In a single lower ferritin group.149 RCT comparing higher to lower TSAT targets, ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S61

elbaT 42 . ulaV e o f lesaB eni eF rr ti ni f or A s isses ng iL ek doohil fo esnopseR ot I V I nor Th ypare i n eht H D DKC- P alupo t noi

aR gn e c Challe gn e tuO oc em Re srednops ernoN s srednop b m( )nae fo naR eg aem( n) noitinifeD( ,rohtuA Y rae N aB les ine of ilesaB ne Ave ar eg esoD fo F itirre n T AS T %( ) A eg tn Du itar on tnemtaerterP P er emtaert tn Ferritin Re )esnops a a (ng/mL) (ng/mL) C italumu ev soD e reF itir n /gn( mL) Fe irr t ni n( /g m )L Threshold Specificity Sensitivity

F irre t ni < 05 0 01 0% %0 001 mg H≥tc %5 ro %66 34% 003< 9 %0 1 8% abhsiF n ,e 6< 00 01 H D E AS d ose 02< 0 %77 %73 74 (22 1. d) IV ir no nartxed 19 69 152 41( 1 d) noisses s ≥%01 051< 71% %96 m 000,1 m g nI 2 om 120 ± 611 1 82 ± 21 1 < 001 %84 7 %5 %73 05< %73 7 %5 004< 9 %6 %71 40 m g %18 %91 005-001 21 .3- 6.54 42 DH bH ≥1 Ld/g 03< 0 8 %4 %81 niL , 2 0 10 61 1 3 2 IV irref c o ix ed (210 ± )011 2( 7 9. ± 1 0. )8 sses ions ni 8 kw 960 m g 2 41 ± 701 91 2 ± 21 8 <200 5 %0 33 %

,namttiM <6 00 5< 0 1 HD IRC e ≥ %1 88% 21 % 51 5 77 d d IV nori de tx nar 5 00 gm 001< 3 %8 5 3% 7991 31( 9 ) (2 6.3 ) isses on ni 2 kw 21 5 ± 001 1 41 ± 1 36 VI chondro ti ni tiM us iki , 002-01 f 1 54-0 f 04 gm eW ylke rof Hb ≥1 Ld/g %36 %73 72 sulf nori-eta 001< 6 %5 %06 02 03 951 48( d) 7.72( d) 8 w k in 8 kw loc l io d 023 gm 78 ± 36 39 ± 7 4 rreF i it n + Ad id oit n la I idn c se dna 05< dna %28 8 %9 muidos VI muidos opyH g > 6% o 13 o r 6 2 gm E ca h DH %14 %95 T rotisse e, cirref Hb ≥15% 1< 00 a nd 521 ( 102 d) .22( 3 d) ses s noi 69% 16 % 02 10 154 lg u tanoc e ni 91-6 kw ST AT %02< w 91-6 w k oc elpm x < 01 0 %53 %87 68 -2 1,770 mg 165 ± 11 5 522 ± 127 2 05< 2 0% %59 3< 00 %38 57% T WI rof 4 gm 001 gm H≥tc %3 ro %56 %53 3< 00 a dn Chu ,gna < 05 0 VI i ron wk th ne 63% 39 % 65 .83( 1 d) ASE ≥%03 rHC < 28 pg 3002 351 (2 03 d) sac hc ara et eve ry 2 wk 6 ni 6 mo 003> na d ,2 200 mg f ro 5 om 91 5 ± 137 2 93 ± 131 %25 9 %8 HC r > 82 gp tooF on et s: .a eM na s re mu ref ir tin o f s bu uorg p prior ot ri on c lah l ne ge. b. Sen tis i ytiv er ef rs t o eht r ita o o f i( n id iv ud als iw th t eh o ut oc me hw o h va e ht e test t rh e lohs di) di iv d de by a( ll ni div laudi s w ht eht moctuo e). ctuO o em srefer ot a er ps on es as defined yb e ca h ts udy. pS .c pS e yticific er sref ot t he oitar fo dividni( u sla w i tuoht ht e emoctuo w oh do ton evah eht set t )dlohserht d ivi ded by ( lla idni slaudiv iw ht tuo ht e o octu em .) uO ct emo srefer ot a r e ops nse sa d nife ed b y hcae .yduts d. aC l .detaluc e. C IR = er t ci colu y et c uo nt um l ilpit ed b y eht H tc vid id de by 40. .f tsE im ta de f or m g raph. g. cimorhcopyH rhtyre o setyc patients randomized to a target TSAT of 30% to and ferritin level was less than 1,000 ng/mL. 50% demonstrated a 40% reduction in ESA dose After receiving an average IV iron dose of 38 compared with those assigned to a target of 20% mg/wk for 12 months, patients showed an in- to 30% (Table 23).148 As is the situation for crease in ferritin (mean, 1,383 ng/mL) and TSAT serum ferritin level, other sources of evidence (from 27% to 36%) values and a 25% ESA dose are indirect and of uncertain value in setting decrease.150 Patients with a TSAT of 20% or TSAT targets. In an uncontrolled prospective greater may still show absent bone marrow iron trial described previously, patients with high base- (Table 25) or respond to IV iron challenge (Table line ferritin levels (mean, 930 ng/mL) were ad- 26).98,99,151-156 The statement that TSAT generally ministered IV iron if TSAT was less than 50% should be maintained at greater than 20% therefore

Table 25. Value of Ferritin in Assessing Iron Storage Excess or Deficiency

Range (mean) of Test Range Test Iron Outcome Most Recent Author, Year N CKD Concurrent Result Patient Population (mean) of Organ Determination Definition Iron Therapy Stage Ferritin (ng/mL) Scale Concurrent Ferritin (ng/mL) Threshold TSAT (%) Sensitivity Specificity Iron Excess >340 57% 75% None for 6 mo SQUID µg/g Iron excess: c Relatively unselected, Canavese, 7-86 14-886 (n = 10) >450 46% 82% 40 HD-CKD Liver (quantitative wet >400 µg/g c mean Hb 11.4 g/dL, 2004 163 (33 ± 15) (361 ± 224) None for 2 mo >500 39% 92% hepatic iron) weight hepatic iron mean albumin 3.6 g/dL (n = 30) >600 25%c 100% Iron Deficiency <600 98% 14% Iron <300 92% 36% Fernandez- Stainable iron Relatively unselected, mean HD-CKD Bone deficiency: No IV iron <200 86% 55% Rodriguez, 63 (18.1 ± 16.3) (273.1 ± 357) in aspirate 0 to +3 Hb 10.6 g/dL, PD-CKD marrow 0 stainable for 3 mo <150 81% 65% 1999 98 (qualitative) mean albumin 3.9 g/dL iron <100 72% 80% <50 50% 95% Selected for unexplained HD-CKD Iron Kalantar- Stainable iron Hb <11 g/dL; mean PD-CKD Bone deficiency: None for Zadeh, 25 4-50 19-1,840 in aspirate 0 to +5 <200 50% 87% albumin 2.8 g/dL, only 4 ND-CKD marrow 0 to +1 13-30 d 1995 99 (qualitative) patients with albumin Tx-ND-CKD stainable iron >3.4 g/dL Footnotes: a. Sensitivity refers to patients with the Test Threshold Ferritin who havethe Outcome Definition / (patients with Test Threshold Ferritin who havethe Outcome Definition plus those with Test Threshold who do not have the Outcome Definition). b. Specificity refers to patients without the Test Threshold Ferritin who do not have the Outcome Definition / (patients without the Test Threshold Ferritin who do not havethe Outcome Definition plus those with the Test Threshold who do not have the Outcome Definition). c. Estimated from graph. S62 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

o eulaV .62 elbaT .62 eulaV o f aB s enile TAST f or gnissessA L i doohilek fo esnopseR ot VI norI yparehT ni eht DKC-DH noitalupoP c R egna R gna e egnellahC srednopseR srednopsernoN b emoctuO naem( ) of aem( )n fo N raeY ,rohtuA raeY N aB es l eni noitinifeD( of

aB s enile egarevA esoD Pr tnemtaerte terP r tnemtae (%) nitirreF tnegA itaruD no )esnopseR TSAT )%( TAST )%( TAST )%( a TAST )%( a Threshold Threshold Specificity )Lm/gn( italumuC ve soD e Sensitivity %11 %69 03< %69 %11 001 mg %66 %43 2< 7 %29 22 % tcH ≥%5 ro nabhsiF ,e 01 DH <24 88 % %44 1.22( 74 1.22( d) 006< IV nori nartxed ASE esod 91 96 51 2 s se sio sn ≥01 % ni 2 om 12< %18 %36 00,1 0 gm 4.91 ± 8.11 4.72 ± 4.91 81< %85 %57 %88 %61 51< %61 %88 53< 7 %7 %0 gm 04 gm 8 %1 %91 12.3 6.54- 100 005- 42 DH H≥b 1 Ld/g 03< %26 %71 002 ,niL 002 1 161 23 VI cirref edixo )8.01 ± 9.72( ± )8.01 902( ± )901 ses snois ni 8 kw 52< %64 %05 9 06 gm 2 .8 5 ± .11 7 25.1 ± 5.1 %88 %13 02< %13 %88 Mi mtt an, 05< 006< 1 HD RC I e ≥%1 %11 %68 79 VI ori n xed tran 05 0 mg 2< 0 5 0% 6 %0 7991 551 6.32( d) 31( 9 d) es s is on ni 2 kw 21 ± 21 42 ± 01 IV dnohc r tio in 04 gm %36 %73 03< 28 % %07 tiM us ,iki 01 - 54 f 10 022- f ylkeeW rof bH ≥1 Ld/g 27 lus f i-eta r no 02 03 51 9 7.72( c) 8( 4 d) 023 gm 8 kw ni 8 kw 23 7. ± 4.6 .43 5 ± .7 8 02< %53 %001 loc l dio VI s muido 13 or 6 2 gm hcaE DH 14 % 59% Te otiss r ,e f cirre bH ≥%51 .22( 521 .22( 3 d) 02( 1 d) noisses 91< %95 %87 1002 154 ulg c etano ni 91-6 w k 86 2 077,1- gm 91-6 kw 7.71 ± 6 6. 2 4.5 ± 4.8 co xelpm tooF n seto : .a aeM n T SAT of s gbu puor p roir t o ori n hc ella .egn b. neS is tivity r efe sr t o t he itar o of ( ni ivid duals w hti the out moc e w oh evah t eh t e ts rht lohse d) vid i ded by ( all ni dividu la s iw ht t eh outc emo ). uO t oc me er ef rs to a ser op sn e a s fed i en d by e ca h s ut .yd c. icepS f ici ty er fers to eht r a it o fo dni( udivi sla iw th uo t eht uo tc mo e w oh d o ton h va e t he et ts t hre hs lo d) d vi id de by (a ll ni di iv d au ls iw tho tu t eh uo t .)emoc uO ct o em re ef sr to a r se po sn e as d fe ined yb each ts udy. d. detaluclaC dethgiew egareva enilesab seulav fo srednopser dna .srednopsernon er = IRC .e IRC = er it cu tycol e c tnuo pitlum lied by eht H ct divid ed yb .04 .f E ts i detam morf rg a hp . reflects acceptable, if not maximal, efficacy and a In the second RCT, patients assigned to a CHr cautious consideration for patient safety. treatment target of 32.5 pg/cell or greater showed CHr greater than 29 pg/cell. Evidence to no change in ESA dose, whereas those assigned support a treatment target for CHr is drawn from to a TSAT treatment target of 20% or greater 2 RCTs comparing the use of CHr with TSAT as showed a 38% decrease in ESA dose. However, an iron target. In the first, patients were assigned between-group ESA differences did not achieve to receive IV iron for a CHr less than 29 pg/cell. significance.158 Functional iron studies have A control group received IV iron for either TSAT shown that the clinical utility of CHr as a base- less than 20% or ferritin level less than 100 line index of response to IV iron challenge varies ng/mL. Ferritin and TSAT were followed up in (Table 27).151,153-155,159 Results of CHr also may both groups, and IV iron was withheld in any be instrument dependent. The current recommenda- patient with a TSAT greater than 50% or ferritin tion of CHr greater than 29 pg/cell is based on the level greater than 800 ng/mL. Patients assigned Advia 120 instrument (Bayer Diagnostics Inc). to CHr targets showed a lower cost of anemia Other iron status tests have been studied, includ- treatment and a decrease in IV iron use compared ing percentage of hypochromic red blood cells, with those assigned to TSAT/ferritin targets.157 zinc protoporphyrin, and soluble transferrin recep-

issessA rof rHC enilesaB fo eulaV .72 elbaT .72 eulaV fo enilesaB rHC rof issessA gn ilekiL dooh fo esnopseR ot VI norI ehT ra yp

egnaR egnellahC pseR edno sr nopsernoN sred c b o )naem( o f ( egnaR ( m )nae noitinifeD fo ohtuA r, raeY N enilesaB fo esaB l eni vA e ar ge oD es esnopseR tnemtaerterP erP mtaert e tn )gp( rHC )gp( reF r niti egA tn ruD a oit n )gp( rHC )gp( a rHC )gp( a CHr (pg) Threshold Threshold Specificity )Lm/gn( lumuC evita oD es Sensitivity

%87 82< %87 %17 ,namttiM 006< IRC %1> %21 %88 1.92( 77 1.92( d) VI nori nartxed 005 gm 1 DH noisses 72< 6 7% %28 991 7 155 931( d) ni 2 kw 62 4. 5.92 62< 44 % 88 % nitiordnohc VI nitiordnohc 04 m g 63% %73 tiM us i ,ik 43-62 e ylkeeW rof H tc %3> 27 022-01 e s lu i-etaf ron < 23 %001 %09 3002 159 (3 2.1 d) 023 gm 8 kw ni 8 kw 3.03 ± 5.1 6.23 ± .0 7 diolloc ≤30 %96 7 8% os dium cirref 13 ro 26 gm hcaE DH 14 % %95 Te ss i ot r ,e bH ≥%51 ≤29 %75 %39 03( 521 03( .5d) 102( d) etanoculg noisses 1002 451 ni 6 1- 9 weeks ≤82 73 % %79 xelpmoc ,1-286 7 07 mg 6-19 wk 0.92 ± .2 6 31.6 ± 2 0. ≤26 10% 001 % 1 00 gm WIT rof 4 kw cH t by 3% ro 6 %5 35% hC uang, VI i or n 65 82( .0d) <500 neht yreve 2 ASE yb %03 2< 8 %87 78 % 02 03 351 as c hc ra ta e 2, 002 gm 2.72 ± 6.1 6.92 ± 7.1 m 5 rof kw rof 5 m o in 6 om hsiF ab ,en CRI ≥1% ni %22 87 % 32 72( .0d) 006< VI nori nartxed 1 , 000 gm 1 DH s e noiss < 62 100% %08 7991 151 2 kw 7.32 9.72 F oo nt o set : a. aeM n rHC of us b org up p oir r ot ir no hc a ell gn .e b. S e ivitisn ty refer s to norI Re ops nders w ti h eht T tse T AS T erhT shold / ( rI on eR s op sredn with t eh Te ts ST AT T erh ohs ldsulp I r no srednopseR ohw do n to m eet Te ts ST TA ht r .)dlohse .c S ep c fi tici y r e sref t o oN n pser o dn ers wit oh tu ht e T tse T TAS Torh ohse dl / N( no er spo dn sre iw uoht t thte seT t TAST rhT es oh dl p lus N rn ednopse sr w ti h t eh T se TS TA t rh se ho dl ). d. C alc .detalu E .e E sti am t de from arg p .h ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S63

lbaT e 28 . naeM ylhtnoM esoD fo norI ni ludA t erP elav tn DKC-DH itaP stne no ASE ehT r ypa

fo epyT fo eM an F e rr ti ni gnirud ohtuA ,r raeY oF pu-woll norI G nevi Na In evret n noit mrA 1 noitnevretnI mrA 2 uO t emoc ydaetS etatS S ydut iseD ng ( )om b naeM oD es )Lm/gn( )om/gm( anam aimena dezilocotorP aimena anam tnemeg erac lausU ac er yb yramirp tsigolorhpen and norI s esorcu 002 ,elbmirB 002 3 81 3 yramirP : rP noitropo fo stneitap htiw 167c 8 yb yduts et am n ru es c il n ci ian 002 .sv 022 R TC bH seulav ni egrat t egnar 841 .sv 251 egraT t :bH 1 1 5.21- d/g L egraT t :bH 1 1 1- 2.5 d/g L s norI s esorcu 394 5002 ,namloT 5002 31 9 CS -niteopebrad 1 kw/x CS -niteope /x1 kw 162 9 :yramirP ASE esod ta 9 om 042-041 d naidem( rof rA m 2 R TC aT r teg :bH 1-11 2 Ld/g Ta tegr :bH 21-11 Ld/g )egnar( skeew )9-3 rP im :yra rI on anam emeg tn I I f irref t ni 001< m/gn L o r TAST < 2 ,%0 If C rH < 29 ,gp F hsi enab , 2 100 157 cH( t, E AS ,esod nori esod norI nartxed 31 8 6 VI nori ed ,nartx 01 0 gm no 10 VI ori n ,nartxed 001 gm no 01 013 .sv 082 TCR er uq ire em n ,t se ur m fer tir ni , ST AT, tucesnoc evi H D snoisses evitucesnoc DH snoisses )rHC 062 021.sv eC rv ille , 5002 81 4 SC brad e teop in- /x1 wk IV pebrad o -nite 1 x kw/ Da br e teop ni sod e dna esod norI ylop malt so e 42 6 e 507 .sv 395 sorC sov re R CT tegraT H b: 31-11 Ld/g tegraT :bH 1-11 3 Ld/g airav b ytili 1 84 v s. 8 8 dedeen-sA )tnettimretnI( nori 546 .sv 13 2 ( ecnanetniaM ( c uounitno s) nori norI xed t nar Be ,baras 91 99 061 orI n areht py if TAST < 2 %0 ro (A :1mr w ee k 1 5 1 42 1 8 Ta tegr TS :TA %03 -50% amirP yr : ASE od es N no TCR- nitirref 002< Lm/gn dna ;retal :bH tegraT :bH 10 1- 1 g Ld/ f 2 80 061.sv Ta tegr H :b 01 - 11 Ld/g g mrA :2 a ll )skeew :setontooF .a N ta oc m noitelp . b. feR le tc s ht e a rev ga e om ylhtn od es fo VI i or n ad im n deretsi . ( One m no ht = 4 keew .)s m 5> dellorne stneitaP .c stneitaP dellorne 5> m o shtn d. E sti detam f mor arg ph .e ehT f i tsr 4 m tno sh of e hca nom-6 th ts udy ep ri do erew d gise tan de sa od es t noitarti ahp s ,se t eh if an l 2 om tn sh erew sed detangi sa aulave t noi rep doi s rof d gniso eff ci ien .yc nori VI .f VI nori d nartxe -003 5,0 00 m g t o AST T ,%03> lof ol dew yb 2 001-5 gm yreve 2-1 skeew ot atniam in AST T neewteb .%05-%03 M a mumix mo≥ylhtn esod 004 .gm oN ri on nevig i f f e nitirr 008 n g Lm/ . g. VI nori d nartxe 001 gm no 01 cesnoc u vit e H D es s snois nehw T SAT 02< % ro tirref in level 002< .Lm/gn tors. Evidence to support these tests as targets for Upper Level of Ferritin iron treatment is poor or unavailable. The statement There is insufficient evidence to Treat-to-target approaches. There are 2 recommend routine administration of IV iron if widely used and effective approaches to IV iron ferritin level if greater than 500 ng/mL reflects treatment: (1) periodic iron repletion, consisting the findings of the Work Group that: (1) no RCTs of a series of IV iron doses administered episodi- have compared the safety and efficacy of ferritin cally to replenish iron stores whenever iron sta- targets greater than 500 ng/mL with the safety tus tests decrease to less than target range; or (2) and efficacy of lower ferritin targets, (2) few continuous maintenance treatment, consisting of studies have examined the efficacy of IV iron at smaller doses administered at regular intervals to ferritin levels greater than 500 ng/mL, (3) no maintain iron status tests stable within target. No study has examined either efficacy or safety RCTs are available to compare the efficacy and beyond surrogate outcomes, (4) no information safety of these 2 approaches; however, the total 160 from interventional trials is available about the cumulative dose may not differ between them. safety of ferritin targets greater than 500 ng/mL, In studies that achieved steady-state ferritin lev- and (5) sufficient evidence exists to suggest that els and reported iron doses administered, the tissue iron stores in patients with ferritin levels average IV iron dose needed to maintain a stable greater than 500 ng/mL are normal to greater ferritin level (signifying achievement of the goal than normal. of neutral iron balance) appears to be in the range No evidence is available to compare ferritin of 22 to 65 mg/wk (Table 28). Higher doses may targets greater than 500 ng/mL with lower fer- lead to progressive increases in serum ferritin ritin targets. That is, no RCTs have assigned levels in some patients.148,156 patients to a ferritin target greater than 500 Iron Targets in Patients With ND-CKD and ng/mL and compared results with those in pa- PD-CKD tients assigned to a ferritin target less than 500 Evidence to support an iron treatment target in ng/mL. In the absence of comparisons between patients with ND-CKD and PD-CKD is lacking. groups in intention to treat, no conclusions can There are no RCTs comparing iron targets in be drawn to assess RRs and benefits of IV iron these patient populations. Dialysis-related blood therapy at the higher ferritin level. loss renders evidence from patients with HD- The efficacy of IV iron in patients with base- CKD unsuitable for application to patients with line ferritin levels greater than 500 ng/mL has ND-CKD or PD-CKD. Thus, in these patients, been examined in a single uncontrolled study.150 the Work Group recommends iron treatment tar- Described previously, this trial administered IV gets that reflect a conservative estimate of effi- iron to patients for 12 months, withheld iron for cacy and a cautious approach to patient safety. patients with a ferritin level greater than 1,000 S64 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS ng/mL or TSAT greater than 50%, and found that patients with CKD. Among patients with CKD mean ferritin level increased from 930 ng/mL at undergoing bone marrow examination, no pa- baseline to 1,383 ng/mL at 12 months and ESA tient with a ferritin level greater than 500 ng/mL dose decreased by 25%. However, the lack of a showed absent marrow iron stores.98,99 Finally, control group renders conclusions about the ESA among patients with HD-CKD undergoing tissue dose change uncertain. Analysis of uncontrolled iron determination by means of magnetic suscep- anemia interventional trials is subject to target tometry, ferritin levels greater than 500 ng/mL bias. That is, patients who are most likely to were associated with hepatic nonÐheme-iron con- survive and complete the trial likely bear a low centrations greater than the upper limit of nor- disease burden and are, in turn, most likely to mal.163 In short, ferritin values exceeding 500 show higher Hb and lower ESA doses. Con- ng/mL likely represent supraphysiological re- versely, patients who are least likely to complete sults in the general population and tissue iron the trial bear a high disease burden and therefore sufficiency or excess in patients with HD-CKD. are most likely to show worsening anemia and Thus, no current evidence is available to sup- higher ESA doses. Successful trial completion port routine treatment of patients with serum therefore is associated with higher Hb levels and ferritin levels greater than 500 ng/mL. Rather, lower ESA doses, regardless of intervention. clinicians should judge the individual patient’s Other efficacy trials have examined the func- clinical status and ESA responsiveness and base tional relationship between baseline ferritin level iron treatment decisions on this assessment. In and likelihood of a surrogate hematologic re- particular, when serum ferritin level is greater sponse to IV iron challenge (Table 24). However, than 500 ng/mL while the concurrently measured none of these studies examined patients with TSAT is less than 20%, iron deficiency may be baseline ferritin values greater than 500 ng/mL. present and a course of iron treatment may be Moreover, because efficacy of IV iron in the considered. available interventional and functional iron stud- In short, no current evidence is available to ies is limited to surrogate outcomes (reticulocyte support treating most patients with serum ferritin increase, Hb level increase, or ESA dose de- levels greater than 500 ng/mL. A therapeutic crease),152-155,161,162 evidence of direct patient response to a 1,000-mg IV iron challenge in a benefit (including improved QOL, health, or sur- patient with a ferritin level greater than 500 vival) of IV iron administration in patients at ng/mL is unlikely. Because iron stores are ad- ferritin values greater than 500 ng/mL is lacking. equate or elevated in patients with ferritin values Information on potential harm to patients greater than 500 ng/mL, the long-term safety of similarly is lacking. Again, evidence to permit iron therapy at target ferritin values exceeding reliable assessment of risks and benefits can 500 ng/mL is untested, and efficacy results of IV only come from RCTs assigning patients to iron therapy are limited to surrogate outcomes intent-to-treat ferritin targets greater than 500 rather than direct patient benefits, the Work Group ng/mL compared with less than 500 ng/mL, favors limiting routine iron treatment to patients with adequate power to assess both efficacy with serum ferritin values less than 500 ng/mL. and safety. Accordingly, the conclusion of the Work Group Evidence that tissue iron stores in patients is that there is insufficient evidence to recom- with ferritin levels greater than 500 ng/mL are mend routine administration of IV iron if ferritin adequate to greater than normal includes results level is greater than 500 ng/mL. from surveys in the general US population and The finding of a TSAT less than 20% coupled studies of bone marrow iron or liver iron in with a ferritin level greater than 500 ng/mL patients with CKD. Among randomly selected poses a particularly difficult problem for clini- individuals in the general US population, ferritin cians.157 This situation may be caused by iron values greater than 500 ng/mL lie at or greater test variability,157 falsely low TSAT results, than the 95th percentile among males aged 15 to inflammation, or reticuloendothelial iron block- 59 years and far exceed the 95th percentile among ade. Evidence on the risks and benefits of IV females of all ages.163 However, findings in the iron therapy in these patients is almost entirely general population do not necessarily apply to lacking. However, the statement that There is ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S65

Table 29A. Comparative RCTs of IV versus PO Iron Administration and Efficacy of Anemia Management in the HD-CKD and PD-CKD Populations

ESA Co- Baseline Intervention Intervention Clinical Outcome a Result Author, N IV Agent Year Statistical

(mo) Actual Value

Dose Frequency Description Definition Signifi- Quality

CKD Stage CKD Stage IV vs. PO

Applicability PO Agent Ferritin Ferritin cance (ng/mL) Hb (g/dL) Hb (g/dL) Mean Follow-up TSAT ( %) ESA Dose

7,100 100 mg Mean Hb 10.8 23 191 Iron dextran 2x/wk 11.5 vs. 10.6 + IU* over 2 min (g/dL) Fishbane, Ferrous sulfate IV ESA 52 HD-CKD 4 325 mg TID 1995 156 6,750 (N = 21) adjusted b Mean ESA dose 10.6 21 179 4,050 vs. 7,563 + IU* Iron polysaccharide (IU/treatment) 150 mg BID (N = 11) Iron sodium Each HD Fudin, 7.8 0 21 c 204 62.5 mg Mean Hb 30 HD-CKD 26 gluconate complex session None 11.0 vs. 6.5 c nd 1998 167 (g/dL) 6.8 0 20 c 268 Ferrous sulfate 160 mg Daily 5 mL (~250 mg) Mean Hb 7.3 0 26 345 Iron dextran Every 2 wk 11.9 vs. 10.2 + Macdougall, HD-CKD over 25–30 min SC ESA (g/dL) 25 4 1996 166 PD-CKD adjusted d Mean ESA dose 7.2 0 27 309 Ferrous sulfate 200 mg TID 1,202 vs. 1,294 NS (IU/kg/16 wk) * Per treatment. Footnotes: a. Primary outcomes of study given unless otherwise noted. b. Adjusted to target Hct 30%-34%. c. Estimated from graph. d. ESA dose kept stable for at least 8 weeks. Dose decreased if Hb >12 g/dL or dose doubled if Hb had not increased by 1 g/dLt 8 a weeks.

Coding of Outcomes: + Statistically significant benefit seen for comparison of arm 1 vs. arm 2, unless otherwise noted insufficient evidence to recommend routine ad- Route of Administration ministration of IV iron if ferritin level is greater Evidence to support the statement The pre- than 500 ng/mL does not preclude IV iron ferred route of iron administration is IV in pa- administration in selected patients when, in the clinician’s judgment, a trial of iron therapy is tients with HD-CKD consists of results from 3 warranted. RCTs comparing IV iron with oral iron adminis- It is important to reiterate that the statement tration (Table 29AÐTable 31A), including 2 that There is insufficient evidence to recommend rou- incorporated a placebo or nontreatment control tine administration of IV iron if ferritin level is arm (Table 30). A fourth RCT compared oral iron greater than 500 ng/mL refers to ferritin targets, administration to no-iron control (Table 30). In 2 not to achieved or acquired ferritin levels. Fer- of these 4 RCTs, patients were assigned to IV or ritin levels exceeding 500 ng/mL are often oral iron treatment while undergoing ESA 159,166 achieved as a result of iron therapy given to therapy. By study completion, those as- patients with lower baseline ferritin levels or signed to IV iron showed a greater Hb level, may be acquired in the course of infection, lower ESA dose, or both compared with those inflammation, or other illness. Achieved or ac- assigned to oral iron (Table 29A, Table 31A). In quired ferritin levels are described elsewhere,164 the third RCT, patients not receiving ESA showed but are not the subject of this guideline state- an increase in Hb levels from 7.8 to 11.0 g/dL ment.165 after IV iron, but no change in Hb levels was

moC .B92 elbaT .B92 moC p evitara sTCR fo VI .sv OP norI noitartsinimdA adn ycaciffE fo aimenA naM a eg tnem ni eht DKC-DN noitalupoP

SE A -oC Basel ni e tnI e vr e oitn n evretni n it no nilC aci l uO tcome a stluseR ,rohtuA N IV gA ne t Ye ra S itat stic la

(mo) lautcA laV ue

esoD rF uqe e ycn tpircseD i no eD f noitini -ifingiS Quality

CKD Stage IV v s. OP Applicability Applicability

Ferritin Ferritin PO Age tn ecnac (ng/mL) Hb (g/dL) Mean Follow-up TSAT ( %) ESA Dose

500 m g x2 or 5x % p tneita s htiw 39 61 dn 2.01 dn 61 39 orI n s u esorc ro N = 3 2 p ta i tne s 4 .4 %0 vs. 2 .8 0% + Van yW ck, vo er 14 d bH ≥1 Ld/g 161 -3 5 2 02 0 gm on E SA 02 05 1 96 detsujdanu aeM n Hb 10.1 dn 17 104 Ferro su s lu f ta e 3 25 gm T DI 0 7. .sv 4.0 + 4 yad yb yad 4 2 d/g( L) 9.8 0 71 521 norI esorcus 200 gm keew ly aeM n H b Ld/g( ) 0.1 .sv 7.0 SN iw stneitap % stneitap iw ht hC arytan, b CS E SA 172 96 -3 4 1.5 bH ≥0.8 Ld/g dna %0.56 .sv %0 2005 9 7. 0 61 301 uorreF s etaflus 523 gm DIT un da j su det + c eF eF r ≥160 n /g mL 44( .0% v .s )%0 dna( TS TA )%5> gm 003 gm CS ASE 99. 0 dn 1 00 norI esorcus tnom h yl evotS s, evo r 2 rh da j tsu ed ot naideM ASE esoD 4-3 54 4-3 d 2.5 6.14 .sv 5.33 SN 1002 171 eveihca bH (I )kw/gk/U 9 7. 0 n d 47 Fe rr uo s sulfate e 002 m g TDS Ld/g 21 Ld/g A ,lawragg 8.5 0 46 181 norI nartxed 2 Lm 001( )gm om/x2 naeM bH 40 5-4 f 3 ESA anu d tsuj ed 10.0 sv . 8 9. + 2003 170 3.6 0 06 91 0 Ferrous s lu fa et 002 gm DIT d/g( L) F too :seton a. yramirP semoctuo fo yduts nevig sselnu o esiwreht .deton b. nitaerC i en ecnaraelc < 04 m/Lm ,ni n to req gniriu ylaid s .si c. P = .0 00 f or b hto a an ly es s. d. I ividn d slau w hti S rC 8.2> 4 m ,Ld/g ton uqer iring laid y is s. er. itanretlA v ,yle eht p a stneit viecer ed na qe u tnelavi esod fo aen etla r evitan nori p e ap r noita taht dah neeb rp e suoiv yl tarelot .d initaerc naem neewteb secnereffid tnacifingis yllacitsitatS .f yllacitsitatS tnacifingis secnereffid neewteb naem initaerc n2e ecnaraelc fo VI nori puorg .6( 47 )nim/Lm dna OP nori puorg 1( 20. )nim/Lm ta enilesab .)50.0

oC id gn of uO t :semoc + St ta i cits a yll cifingis tna feneb it ees n f ro c mo parison fo mra 1 .sv mra ,2 sselnu to he esiwr n eto d S66 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

Table 30. Comparative RCTs of PO Iron Administration versus Placebo/Control and Efficacy of Anemia Management in the HD-CKD and PD-CKD Populations

Baseline Intervention Results

Author, Year N PO Agent Actual Value Statistical CKD ESA Clinical Outcome (mo) Dose Frequency Arm 1 vs. Signifi- a,b (IU)

Stage Follow-up Cointervention Definition

Ferritin Arm 2 cance (ng/mL) Hb (g/dL) TSAT ( %) ESA Dose Placebo/Control

d Fudin, 6.8 0 20 204 Ferrous sulfate 160 mg Daily Mean Hb HD-CKD 12-26c 24 None at 12 mo 6.5 vs. 6.0d nd 1998 167 6.3 0 16d 190 Control — — (g/dL) Mean Hb HD-CKD 7.2 0 27 309 Ferrous sulfate 200 mg TID 10.2 vs. 9.9d NS Macdougall, SC ESA (g/dL) PD-CKD 4 25 1996 166 adjustede Mean ESA dose ND-CKD 7.3 0 23 458 Control — — 1,294 vs. 1,475 NS (IU/kg/16 wk) 24 iron- Iron polysaccharide 10.1 5,714 26.5 40 051 m g IB D ESA naeM Hb )Ld/g( g 10.7 vs. 10.5 NS deficient complex adjusted Markowitz, patientsf 10.4 4,850 42.6 39 Placebo 150 mg BID Mean ESA dose (IU)g 6,142 vs. 4,650 nd HD-CKD 3 1997 185 25 iron- Iron polysaccharide 11.2 4750 36.4 221 051 gm DIB ESA eM an bH )Ld/g( g 7.3 vs. 8.5 NS replete complex adjusted patientsf 10.7 4,692 35.4 228 Placebo 150 mg BID Mean ESA dose (IU)g 4,417 vs. 4,615 nd Footnotes: a. Primary outcomes of study given unless otherwise noted. b. Outcomes reported at final follow-up, unless otherwise noted. c. 12-month follow-up for control group; 26-month follow-up for oral iron group. d. Estimated from graph. e. ESA dose kept stable for at least 8 weeks. Dose decreased if Hb >12g/dL or dose doubled if Hb had not increased by 1 g/dL 8 at weeks. f. Iron Deficient: Ferritin <100 µg/L; ZPP <90 mmol/mL heme; Hb>10g/dL. Iron Replete: Ferritin <100 µg/L; ZPP >90 mmol/mL heme; Hb >8.3. g. Primary outcome was change in serum ferritin and ZPP. Results reported at 3-month follow-up before final 2-month period without treatment. seen in patients assigned to oral iron. Between- procedure, disproportionately high baseline ESA group comparisons were not reported.167 Of the doses in the oral iron treatment group, a substan- 4 available RCTs, 3 included a placebo-treatment tial dropout fraction confined to the oral iron or no-iron-treatment arm. Results showed no treatment group, a significant decrease in serum difference in final Hb level or ESA dose between ferritin levels during 6 months of oral iron treat- oral iron treatment and placebo/control. In sum- ment, and an undefined IV iron treatment proto- mary, evidence in patients with HD-CKD sup- col render the results of this trial difficult to ports 2 conclusions: (1) oral iron administration interpret. Further studies will be required to is not demonstrably more effective than either provide a basis for recommendations on the use placebo or no treatment; and (2) IV iron adminis- of this agent. tration is superior to oral iron administration. Evidence to support the statement The route A newer form of oral iron that has not been of iron administration can be either IV or oral studied widely is heme iron polypeptide. A single in patients with ND-CKD or PD-CKD consists RCT of patients with HD-CKD compared treat- of results from 4 RCTs comparing IV iron with ment with IV iron with treatment with heme iron oral iron administration in patients with ND- polypeptide.168 Uncertainty in the randomization CKD (Table 29B, Table 31A). The Work Group

Table 31. Serious AEs of Iron Agents in Patients Naïve to Tested Iron Agent (N > 100)a

Serious ADE Rate Dose Size IV Rate Doses (n) Author, Year Study Design N Population (%) Definition of Serious AE (mg) (mg/min) [Doses/pt] Per Patient [Dose] b Iron Dextran McCarthy, Retrospective, 3,578 0.8 254 HD 300 5-10 Suspected anaphylactic reaction comprised of wheezing/dyspnea 2000 186 database review [nd] [1st course] Retrospective, 20,213 0.035 Requiring parenteral administration of epinephrine, Walters, 2005 176 20,213 c HD and PD — — database review [1] [Test or 1st ] corticosteroids, or antihistamines on the same dialysis day Ferric Gluconate Drug intolerance, which included anaphlyactoid reaction, hypotension, chills, back pain, nausea, dyspnea/chest pain, facial RCT, 2,514 0.6 Michael, 2002 175 2,534 HD d 125 12.5 flushing, rash and cutaneous symptoms of porphyria, and 1 life- crossover [1] [1st] threatening event requiring parenteral antihistamine, hydrocortisone, and epinephrine. Iron Sucrose e Any event that is fatal or life-threatening, results in or prolongs Prospective, 8,583 0 hospitalization, results in significant disability or incapacity, is Aronoff, 2004 187 665 f HD 100 50 cohort [nd] [NA] unusual, or in opinion of the investigator, presents a significant hazard to the patient. Prospective, Stage 3-5 266 Not specified. No adverse reactions requiring parenteral Blaustein, 2003 188 107 500 3.2 0 cohort (1 PD, 2 HD patients) [2] medications or hospitalizations Prospective, ND 0 Hypotension, wheezing, respiratory stridor, shortness of breath, Charytan, 2004 189 130 f HD d 100-200 50-100 cohort [nd] g [NA] hypertension, anaphlyaxis, chest tightness Footnotes: a. Modified for K/DOQI use from reference190. See Methods for selection criteria. b. Majority of ADEs occurred during this/these dose(s). c. Incident subpopulation of 48,509 patient study. d. <10% of patient with iron dextran allergic history. e. Included patients with previous intolerance to either iron dextran or ferric gluconate. f. All patients with previous intolerance to iron dextran and/or ferric gluconate. g. Median cumulative dose: 1,000 mg; range: 100-5,000 mg. ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S67

rated the evidence from these trials as strong or The 4 dosing regimens included 1,000 mg admin- moderately strong. There are no RCTs in pa- istered in 2 or 5 divided doses over 2 weeks,169 tients with ND-CKD comparing oral iron 200 mg monthly,170 300 mg monthly,171 or 1,000 therapy with placebo or no iron treatment. mg in 5 doses over 5 weeks.172 There are no RCTs in patients with PD-CKD. Finally, the 4 available RCTs differ substan- The 4 available in ND-CKD RCTs differ sub- tially in severity of anemia at baseline, ranging stantially in the use, timing, and adjustment of from a mean Hb level of 5.8 to 6.3 g/dL in 1 ESA therapy (Table 31B). Three initiated ESA trial170 to 9.7 to 10.2 g/dL in the remaining 3 therapy in all study patients at randomization, trials. whereas 1 continued ESA therapy in patients Between-group differences showing superior- previously treated, but did not initiate ESA ity of IV iron over oral iron were seen in 2 of the therapy in previously untreated patients. One 4 available RCTs (Table 31B). The 2 studies that adjusted ESA during the trial to achieve a target initiated ESA therapy in patients with moderate Hb level, whereas 3 permitted no ESA dose anemia found no difference between patients adjustment or required removal from the trial if assigned to IV iron compared with oral iron.171,172 ESA therapy initiation or ESA dose increase was The 2 trials that showed differences either initi- needed. ated ESA therapy simultaneously with iron The 4 available RCTs also differ in the dosing, therapy in patients with severe anemia170 or frequency, and duration of IV iron administered. maintained prior ESA status without adjustment S68 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS in patients with moderate baseline anemia.169 In Hypersensitivity Reactions the latter study, superiority of IV iron over oral There currently are 3 forms of IV iron that are iron was greatest among patients with ESA widely available: iron dextran, sodium ferric therapy and baseline Hb level less than 9.0 gluconate, and iron sucrose. All forms of IV iron 169 g/dL. may be associated with acute AEs, occasionally Other outcomes compared in the 4 available severe, comprised of hypotension with or with- RCTs included rate of decline in kidney function, out other symptoms and signs. The cause of the QOL, adverse GI effects, dietary protein and reactions is incompletely understood. Immune energy intake, and adherence to prescribed mechanisms (including mast cellÐmediated pro- therapy. Compared with patients assigned to oral cesses leading to a clinical syndrome resembling iron therapy, those assigned to IV iron treatment anaphylaxis) may have a role in some cases. In showed no difference in rate of decrease in others, the iron agent may release bioactive, 169-171 169 kidney function, no difference in QOL, partially unbound iron into the circulation, result- 169-172 fewer GI symptoms, no difference in pro- ing in oxidative stress and hypotension (labile or 171 tein and energy intake, and better adherence free iron reactions). The pathogenesis may differ 169 to prescribed therapy. Five patients in 2 trials, depending on the type of IV iron. Anaphylactoid all women with low body mass, experienced reactions appear to occur more frequently with adverse reactions to administration of IV iron, iron dextran,173 and labile or free iron reactions including hypotension, cramping, arthralgia, and occur more frequently with nondextran forms of myalgia.169,171 iron.174 Given the variability in trial design, inconsis- The reported frequency of acute drug AEs tency of results, potential for adverse effects of depends in large part on the structure and rigor of IV iron, lack of information on potential ad- the experimental design used. Direct patient ob- verse effects of oral iron, surrogate nature of servation after administration of an IV iron agent tested outcomes in iron intervention trials, permits the most reliable assessment. For ex- paucity of information on patient preference ample, in a double-blind placebo-controlled trial and adherence to treatment recommendations, of sodium ferric gluconate in 2,534 HD pa- and lack of information on the efficacy of oral tients,175 investigators directly observed patients iron compared with placebo or no-iron treat- for evidence of reactions after blinded IV admin- ment, the Work Group concluded that the istration of drug or placebo. If, in the opinion of strength of available evidence is insufficient to the investigator, a reaction was demonstrated, support a guideline statement on the use of IV serum tryptase levels were measured to evaluate versus oral iron in patients with ND-CKD. for mast cellÐmediated hypersensitivity. However, in the opinion of the Work Group, Prospective single-arm clinical trials also con- the route of iron administration can be either tribute information on drug-related AEs in the IV or oral in patients with ND-CKD or PD- target population (Table 31). In reports of these CKD. trials, observation for drug-related AEs was per- Because no RCTs are available to compare IV formed by the study nurse or a staff nurse or, iron with oral iron in patients with PD-CKD, the alternatively, the method of observation was not Work Group considered whether to regard the specified. PD-CKD patient population as similar to ND- Retrospective study designs use review of CKD, similar to HD-CKD, or too dissimilar medical records or large-scale electronic data- from either to justify a recommendation. On the bases to identify potential drug-related AEs after basis of the assumption that patients with PD- the fact.176,177 CKD do not experience ongoing dialysis-associ- Finally, pharmacovigilance surveillance stud- ated blood loss, the Work Group reasoned that ies use national voluntary reporting data sources, patients with PD-CKD resemble the ND-CKD including the Freedom of Information Act data- population and differ from patients with HD- base administered by the FDA, to evaluate AE CKD. Thus, patients with PD-CKD are included profiles of IV iron agents. Pharmacovigilance in the current opinion-based CPR. studies yield unreliable information about abso- ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S69 lute AE rates and are susceptible to uncontrolled exposures to iron dextran in patients with HD- observer bias, but afford the advantage of detect- CKD (rate per exposure, 0.02%). There was 1 ing rare events, shedding light on common char- fatal event. Although many AEs followed admin- acteristics of low-frequency events, and suggest- istration of a test dose or first dose, this study ing relative rates of AEs. However, they do not identified serious AEs in patients who had suc- permit statistical comparison of AE rates for cessfully received previous test or treatment different IV iron agents. doses. The majority of the non-naïve patients The frequently nonspecific nature of IV ironÐ who experienced AEs after iron dextran adminis- related AEs, the substantial overlap between drug- tration did so at the time of the first dose of a related AEs and dialysis-related AEs (eg, dizzi- planned series, suggesting that the risk for first- ness, dyspnea, cramps, pruritus, nausea, dose AE may recur in prevalent patients after an constipation, diarrhea, and hypotension), and the interval free from iron dextran exposure.177 low anticipated event rate for the most serious Walters and Van Wyck 2005.176 The data- AEs pose significant challenges for each study base of a large dialysis provider was examined to design described. The single greatest problem is identify episodes of iron dextranÐinduced anaphy- the absence of generally accepted criteria to laxis sufficiently severe to require use of resusci- identify IV ironÐrelated AEs and distinguish hy- tative medications. They found 7 events in a total persensitivity from nonhypersensitivity reac- of 48,509 patients treated. However, all 7 events tions. The absence of criteria for IV ironÐrelated occurred in previously unexposed patients after AEs introduces potential observer bias into both first or second exposure to the drug, yielding a prospective and retrospective trials. Pharmaco- slightly greater true incidence of 0.035%. The vigilance studies also are subject to observer bias lower figure reported in this study compared because only AEs thought to be drug related are with the 2 studies using chart review probably reported. As long as neither descriptive criteria reflects the decision to study only the incidence nor objective markers are available, designation of AE requiring in-center use of resuscitative of an AE as drug related likely will remain medication.176 Of interest, the prevalence of sus- somewhat subjective. Although blinding the in- pected hypersensitivity in that study matched the tervention so that neither the investigator nor the findings of the previous retrospective reviews: patient is aware of the identity of the agent 0.69% of prevalent patients were recorded as administered improves the objectivity of AE re- sensitive to iron dextran. porting, event rates are sufficiently low to pre- Pharmacovigilance surveillance studies yield clude comparative trials that are logistically fea- information on deaths potentially related to IV sible. For example, to conduct an RCT that iron. Deaths thought to be IV ironÐrelated have would be adequately powered to compare seri- been reported to the FDA for each of the IV iron ous AE rates between 2 iron agents would re- agents available in the United States. In an analy- quire enrolling more than 10,000 patients, an sis of US and European surveillance databases, unlikely prospect. Accordingly, comparative as- 31 deaths occurring in association with 196 cases sessment of safety is likely to continue to rely on of iron dextran anaphylaxis were recorded in the retrospective chart review, analysis of large- United States between 1976 and 1996 compared scale medical databases, and pharmacovigilance with none with similar exposure rates with so- reports. dium ferric gluconate.180 Another group studied Evidence is available from 2 retrospective the FDA Freedom of Information surveillance reports, including 1 in patients with CKD. Both database to compare severe AE rates for IV iron analyses concluded that the rate of life-threaten- drugs.181 They confirmed higher anaphylaxis and ing reactions to iron dextran administration is fatality event rates for iron dextran than for the 0.6% to 0.7%.178,179 Two additional studies used nondextran irons. These results should be inter- large medical databases to further identify the preted with caution given that AEs for drugs in nature and consequences of life-threatening reac- the marketing phase probably are grossly under- tions after IV iron dextran. reported for all agents; definitions for AEs are Fletes et al 2001.177 This study identified not generally accepted, as we have noted; and the 165 suspected drug-related AEs after 841,252 primary intention of the database is to generate S70 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS signals of unexpected AEs, not to compare is uncertain. In short, iron status targets as goals drugs.182 Nonetheless, these results add to the for treatment require treat-to-target RCTs to pro- body of evidence indicating a trend toward more vide comparative evidence of efficacy and safety. frequent and more severe reactions with iron However, available evidence is of limited quality dextran. or is altogether lacking. Taking the pharmacovigilance information together with evidence that iron dextran AEs may Upper Level of Ferritin occur in patients who have successfully received The limitations of available evidence, de- a previous test dose or series of therapeutic scribed previously, provide the rationale for the doses, we recommend that resuscitative medica- cautious use of IV iron at high ferritin levels. tion and personnel trained to evaluate and resuscitate anaphylaxis accompany each administra- Route of Administration tion of iron dextran. In patients with HD-CKD, available RCTs are small, thereby limiting their potential applicabil- LIMITATIONS ity to unselected HD-CKD target populations. As in all IV iron studies to date, outcomes are Targets of Iron Therapy limited to the surrogate Hb levels and ESA There are a number of serious limitations to doses. Although these outcomes have financial the evidence supporting iron status targets. ESA implications for the total cost of anemia manage- dose and Hb level response as surrogates for iron ment, the comparative cost of anemia manage- efficacy are universal among iron intervention ment using IV versus oral iron treatment has not trials (IV iron challenge studies), RCTs, and been examined, and the relationship, if any, be- uncontrolled prospective trials. No iron treat- tween surrogate outcomes and outcomes that are ment trials have been designed or are sufficiently important to patients has not been determined. powered to yield information on outcomes, in- Moreover, although managing anemia requires cluding the crucial issue of safety, that are di- chronic and ongoing care, most clinical trials rectly important to patients. Of the 2 available involving therapeutic iron intervention are of RCTs comparing iron status targets, 1 used TSAT relatively short duration; thus, the long-term ef- targets,108 1 used ferritin targets,149 both were fects of iron therapy remain unknown. Resolu- small, neither examined safety, and the TSAT tion of these issues, particularly in the ND-CKD trial showed increasing nonÐsteady-state ferritin and PD-CKD patient population, clearly is needed levels in the higher TSAT target arm. Con- before a conclusive assessment of net benefit can versely, IV iron challenge studies yield informa- be made. tion that is limited to the acute hematologic Other Limitations response to a dose of IV iron—results that bear little relevance to optimum treatment targets for 1. Absence of generally accepted criteria for ongoing iron administration. Moreover, even identifying IV ironÐrelated AEs. when limited to predicting the acute response to 2. Absence of a distinction in the literature IV iron challenge, available iron status tests between reactions caused by labile iron and perform poorly, yielding relatively low sensitiv- those caused by hypersensitivity. ity and specificity over a range of cutoff values, 3. No adequately-powered trials directly com- flat receiver-operator characteristic curves, and paring different IV iron agents. low area under the curve. Finally, although stainable bone marrow iron may yield important CLINICAL IMPLICATIONS information on the likelihood of storage iron AEs thought to be related to labile iron require depletion or iron excess or on the relationship a decrease in the dose or rate of infusion or both. between iron stores and results of serum iron AEs thought to be related to hypersensitivity to status tests, the relevance of this information to the agent require stopping the agent and preclude setting optimum treatment targets for iron status further administration. CPG AND CPR 3.3. USING PHARMACOLOGICAL AND NONPHARMACOLOGICAL ADJUVANTS TO ESA TREATMENT IN HD-CKD

Several pharmacological agents and nonphar- fatty acids into the mitochondria, where they are macological manipulations of the HD prescrip- oxidized to produce energy. L-carnitine is also tion have been examined for potential efficacy as thought to be involved in the metabolic conver- adjuvants to ESA treatment. Studies are not avail- sion of acyl coenzyme A, which accumulates in able to address the use of pharmacological or patients with renal failure and is toxic to cells, to nonpharmacological adjuvants to ESA treatment the less toxic acyl carnitine.191 Deficiency of in patients with ND-CKD and PD-CKD. carnitine in patients on maintenance HD therapy 192 3.3.1 L-Carnitine: In the opinion of the was demonstrated nearly 30 years ago. L- Work Group, there is insufficient evi- carnitine, which has been studied primarily when dence to recommend the use of L- administered IV to HD patients, has been postu- carnitine in the management of ane- lated to have beneficial effects on ESA-hypore- mia in patients with CKD. sponsive anemia, HD-related hypotension, myo- 3.3.2 Vitamin C: In the opinion of the cardial dysfunction, impaired exercise tolerance Work Group, there is insufficient evi- and performance status, muscle symptoms, and dence to recommend the use of vita- impaired nutritional status. However, no patho- min C (ascorbate) in the management genic mechanism by which carnitine deficiency of anemia in patients with CKD. might contribute to anemia or provoke ESA hypo- 3.3.3 Androgens: Androgens should not be responsiveness has been conclusively eluci- used as an adjuvant to ESA treatment dated. Furthermore, the therapeutic mechanism in anemic patients with CKD. by which L-carnitine administration might im- (STRONG RECOMMENDATION) prove anemia or enhance ESA responsiveness has not been determined. Finally, no consistent BACKGROUND relationship between baseline plasma carnitine levels, anemia, and response to L-carnitine admin- For the purposes of these recommendations, istration has been demonstrated.193,194 we consider adjuvants to ESA therapy to be In considering L-carnitine administration, the therapeutic agents or approaches that aim to Work Group confined evaluation of efficacy out- enhance responsiveness to ESA therapy in iron- comes to RCTs in which the effect of IV L- replete patients. The target patient population carnitine on Hb level and ESA dosing had been may include both ESA-hyporesponsive and rela- reported in patients with CKD. Thus, Work Group tively responsive patients, although the focus of interest often is the hyporesponsive patient. A conclusions and the resulting guideline statement positive response to an adjuvant treatment con- 3.3.1 address the limited use of L-carnitine as an sists of either an increase in Hb level at a given ESA adjuvant in patients with CKD. Guideline ESA dose or the attainment and maintenance of a 3.3.1 does not address use of L-carnitine for specific Hb level at a lower ESA dose (see potential nonhematologic indications. Executive Summary). In the United States, Medicare coverage for L-carnitine is available for patients who have RATIONALE been on dialysis therapy for at least 3 months, have a plasma free carnitine level less than 40 ␮mol/L, and have “erythropoietin-resistant ane- L-Carnitine mia...that has not responded to standard eryth- In the opinion of the Work Group, there is insuf- ropoietin dosage...andforwhich other causes ficient evidence of efficacy to recommend use of have been investigated and adequately treated.”195 L-carnitine in the management of anemia in pa- Hyporesponsiveness to ESAs is not specifically tients with CKD. The role of carnitine deficiency in defined other than having a persistent Hct less the pathogenesis of the anemia of CKD, if any, is than 30% despite erythropoietin dosage that “is unclear. Levocarnitine (L-carnitine) is a carrier considered clinically appropriate to treat the par- molecule involved in the transport of long-chain ticular patient.”

American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S71-S78 S71 S72 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

The statement There is insufficient evidence to quality, important inconsistencies, major uncer- recommend use of L-carnitine in the manage- tainty about the directness and applicability of ment of anemia in patients with CKD is sup- results, imprecise data, and a probability of bias ported by results from 6 available RCTs of IV rendered the overall quality of evidence very low L-carnitine administration to ESA-treated HD (Table 33). None of the studies specifically en- patients (Table 32).196-200 No RCTs are available rolled patients with anemia and ESA hyporespon- in patients with PD-CKD or ND-CKD (Table 32). siveness or identified a specific subset of patients Anemia was the primary study outcome in only 3 particularly likely to respond to L-carnitine ad- of these studies.196-198 No available RCTs were ministration. Therefore, whether L-carnitine en- judged to be of high quality. Five of the 6 RCTs hances the effect of ESA therapy in such patients were judged to be of low quality. The RCTs were is not known. The Work Group found no specific characterized by small numbers of enrolled pa- evidence of adverse drug effects associated with tients, short duration of observation, concomi- IV L-carnitine in patients with CKD. The ab- tant use of IV and oral iron, adjustments in ESA sence of high-quality evidence for efficacy and dosage, high dropout rates, and uncertainty about safety supports the opinion of the Work Group specific ESA and/or iron dosing during the study. that there is insufficient evidence to recommend Between-group comparison, the result least sub- use of L-carnitine in the management of anemia ject to bias, was available in only 2 studies; there in patients with CKD. was no difference in either Hb level or ESA dose The conclusion of the Work Group differs outcomes in 1 of these studies, and in the other, from those of selected previous reports. A meta- the comparison was only of erythropoietin resis- analysis concluded that L-carnitine administra- tance index (ERI). Serious limitations in method tion was associated with higher Hb levels and ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S73

Table 33. Use of L-Carnitine as an Adjuvant to ESA Treatment in the HD-CKD Population

Directness of Summary of Findings Methodological Evidence, Quality of No. of Studies Total Quality of including Other Evidence for Qualitative Description of Importance Outcome & Study Design N Studiesa Consistency Applicability Considerations Outcome Effect Size of Outcome Uncertain benefit due to limited Hb/Hct Level/ Serious Important Major Imprecise data evidence. 1 study with between-arm 6 RCTs 171 Very low Moderate ESA Dose limitationsb inconsistenciesc uncertaintyd Probability of biase comparison with partial evidence of reduced ESA dose. CKD population: AEs 171 oN tnacifingis sEA detroper hgiH 6 RCTs Balance of Benefit and Harm: Quality of Overall Evidence: No Net Benefit Very Low for Benefit

Footnotes: a. Theoretically, adjusting the ESA dose will make any subsequent change in Hb irrelevant since Hb change is a likely response to an adjustment in ESA. b. Quality varied from Grade B to C. c. N/A given between-group comparison was available in only 2 of 6 and only 1 showed statistically significant reduction in ESA dose. d. The RCTs were characterized by small numbers of enrolled patients, short duration of observation, concomitant use of IV andoral iron, and adjustments in ESA dosage; 2 studies have primary outcomes other than Hb/Hct or ESA dose. e. At least 3 studies sponsored by pharmaceutical companies (1 unstated). lower ESA doses in ESA-treated patients with range,206,207 whether this reflects a clinically HD-CKD.201 However, the meta-analysis insignificant deficiency is uncertain; in other stud- cluded only 3 of the 6 RCTs examined by the ies, ascorbic acid levels have been normal or Work Group and incorporated results from stud- elevated.208 It has been suggested that 150 to 200 ies outlined in 3 abstracts, but that were never mg of vitamin C daily is needed to normalize published in peer-reviewed journals. An NKF vitamin C levels in most HD patients.206 Carnitine Consensus Conference recommended Several anecdotal reports, small case se- the use of IV L-carnitine in selected ESA- ries,194,209 and nonrandomized studies (primar- 202 hyporesponsive dialysis patients. However, ily in HD patients with iron overload, elevated the consensus conference lacked systematic data serum ferritin levels, and functional iron defi- abstraction and analysis of method quality. ciency) using IV vitamin C in doses of 100 to Previous guideline development processes us- 500 mg 3 times weekly have suggested a pos- ing systematic evidence review and rigorous sible beneficial effect.210-213 Plasma levels of evaluation for method quality have reached con- ascorbic acid were not measured in most of the clusions consistent with the current guideline studies in which vitamin C was administered as statement, there is insufficient evidence to recom- an adjuvant to ESA therapy. mend the use of L-carnitine in the management Four RCTs of vitamin C in ESA-treated HD of anemia in patients with CKD . These include patients have been reported (Table 34). Although NKF-KDOQI Guidelines for Nutrition in the uncontrolled studies mentioned tended to CKD,203 previous NKF-KDOQI anemia guide- focus on a possible role of IV vitamin C in lines,2 and the Revised EBPGs for Anemia in patients with HD-CKD with iron overload and CKD.16 functional iron deficiency, only 1 RCT included Finally, although oral L-carnitine has also been patients with functional iron deficiency.214 studied as an ESA adjuvant, no RCTs are avail- One RCT included patients with iatrogenic iron over- able. Direct comparison of IV to oral L-carnitine 215 has not been reported in HD or other CKD load. None of the studies specifically included patients. patients with ESA hyporesponsiveness. These studies have not shown consistent benefit of IV Vitamin C (Ascorbate) vitamin C in either within-group or between- In the opinion of the Work Group, there is group comparisons. insufficient evidence to recommend use of vita- Oral vitamin C, which can augment absorp- min C (ascorbate) in the management of anemia tion of iron from the GI tract, has been evaluated in patients with CKD. Vitamin C has been re- as an adjunct to ESA therapy in small uncon- ported to increase the release of iron from ferritin trolled studies.216-218 Oral and IV vitamin C and the reticuloendothelial system and increase were compared in 1 recent RCT in a small iron utilization during heme synthesis.204,205 Al- number of HD patients for 8 weeks; in a larger though many HD patients may have plasma number of patients, oral vitamin C or no treat- ascorbic acid levels less than the normal ment were compared for 3 months.219 There was S74 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

no significant difference within or between groups high-dose vitamin C, either directly or through in Hb levels or ESA doses in either comparison. its effects on mobilization of iron, has been The long-term safety of IV ascorbic acid in reported.223,224 HD patients remains undefined, with secondary In summary, the Work Group concluded that oxalosis being the primary concern,220,221 al- serious method limitations and important incon- though this was not described with short-term sistencies render the quality of available informa- use in any of the studies described. Plasma tion on vitamin C efficacy very low (Table 35). oxalate levels increase with IV vitamin C admin- At the same time, information on the potential istration. A risk for calcium oxalate supersatura- serious adverse effects (oxalosis) of chronic vita- tion in plasma recently was reported in HD min C administration is altogether lacking. Thus, patients administered IV vitamin C.222 Aside the Work Group concluded that there is insuffi- from the potential for systemic oxalosis, concern cient evidence to recommend use of vitamin C is also warranted because a pro-oxidant effect of (ascorbate) in the management of anemia in

Table 35. Use of Ascorbic Acid as an Adjuvant to ESA Treatment in the HD-CKD Population

Directness of Summary of Findings Methodological Evidence Quality of No. of Studies & Total Quality of including Other Evidence for Importance Outcome Study Design N Studies a Consistency Applicability Considerations Outcome Qualitative Description of Effect Size of Outcome No consistent benefit. 2 of 4 reports Hb/Hct Level/ Serious Important Imprecise datad (both crossover studies) showed benefit 4 RCTs 154 Direct Very low Moderate ESA Dose limitationsb inconsistenciesc Sparse data for Hb/Hct; 1 of 4 showed benefit for ESA. CKD population: 4 RCTs None described in RCTs. Theoretical concern given isolated rare reports of toxicity in patients with ARF or CKD, including secondary AEs 154+ systemic oxalosis, retinal oxalosis, calcium oxalate arthopathy, myocardial calcium oxalate deposition. Toxicity in general population with High General population: intermittent IV doses as used in HD-CKD studies is not known nor is applicability to HD-CKD. observational data Balance of Benefit and Harm: Quality of Overall Evidence: No Net Benefit Very Low for Benefit

Footnotes: a. Theoretically, adjusting the ESA dose will make any subsequent change in Hb irrelevant since Hb change is a likely response to an adjustment in ESA. b. 1 Grade B and 3 Grade C studies. c. Inconsistencies with regard to the direction of the results for treatment and placebo/control groups. d. References for statistical comparisons varied between studies (within arm or between arms). ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S75

Table 36. RCTs Evaluating Effects of Treatment With IM Androgens on Hb Levels in the HD-CKD Population

Result/ Baseline Cointervention Intervention Outcomesa Metric Comparison AEs Author, N Arm 1 Year Definition of (mo) Description of

Clinical Description of Event Quality

Follow-up Cointervention Hb (g/dL) Arms Applicability Ferritin Ferritin

Outcomes in Dose (ng/mL)

Arm 2 Between Hb (g/dL) TSAT (%) Reduction Within Arm vs. Baseline Baseline vs. D/C of Drug or D/C of SC ESAb Nandrolone decanoate Transient flu-like symptoms, c Sheashaa, 16 7.5 34 403 Mean Hb 10.4 4 6 1,000 IU TIW; 50 mg IM 2x/wk NS distressing hirsutism, and/or 2005 259 IV iron at 6 mo 16 7.3 35 394 Control 10.0 elevated liver function tests 0 protocol nd Nandrolone decanoate- Mild side effects attributed to 9 8.4 41 364 11.0 0 100 mg IM weekly nandrolone decanoate: IV ESA Acne N = 1, Gaughan, 1,500 IU TIW d; Mean Hb at injection site pain N = 2, and 6 + 1997 260 Oral and/or IV iron 26 wk injection site hematoma N = 1. 10 8.3 32 301 Control 9.4 0 adjusted e Significant in serum glutamic oxaloacetic transaminase from baseline in androgen group IV ESAb Nandrolone decanoate 6 7.8 nd 1,322 0.12/wk dn 0 Berns, 40 IU/kg TIW; 2 mg/kg IM weeklyf Rate of 4 NS Unacceptable acne 1992 261 Oral and/or IV iron increase in Hb 6 7.8 nd 776 Control 0.11/wk nd 2 adjusted e Footnotes: a. Primary outcome of study unless otherwise noted. b. All ESA administered per study protocol at start; none with prior ESA treatment. c. Subsequent improvement in symptoms after discontinuation of drug. d. Per study protocol at start; washout until Hb 8.7g/dL in those previously on ESA. e. Iron adjusted to maintain ferritin ≥100 mg/dL and TSAT ≥20%. f. Patients randomized to nandrolone decanoate received nandrolone for 2 months before starting ESA.

Coding of Outcomes: Statistically significant benefit for arm 1 vs. arm 2 (with reference to benefit to patient) Statistically significant increase from baseline patients with CKD. None of the RCTs enrolled anemia in dialysis patients despite the need for patients with ESA hyporesponsiveness, and only intramuscular (IM) injection and a variety of 1 enrolled patients with functional iron defi- AEs, including acne, virilization, priapism, liver ciency214; therefore, whether IV vitamin C en- dysfunction, injection-site pain, and risk for pe- hances the effect of ESA therapy or iron metabo- liosis hepatis and hepatocellular carcinoma. Pro- lism in such patients is not known. posed mechanisms of action of these drugs in- In addition to the lack of definitive evidence of clude increased erythropoietin production from efficacy and safety, there is no evidence that renal or nonrenal sites, increased sensitivity of clinical outcomes—such as reduced hospitaliza- erythroid progenitors to the effects of erythropoi- tions, improved cardiovascular status, and re- etin, and increased red blood cell survival. duced mortality—are improved in patients for Three RCTs explored a possible role for andro- whom IV vitamin C treatment is initiated as an gens in combination with ESA therapy in HD ESA adjuvant. Whereas the putative mechanism patients (Table 36). All were small short-term of action of IV vitamin C as an ESA adjuvant is studies, currently recommended Hb levels were an increase in the release of iron from ferritin and not achieved, and in 2 of the studies, the ESA the reticuloendothelial system and increased iron doses used were lower than those used in most utilization during heme synthesis, none of the patients with HD-CKD on chronic ESA treat- studies reviewed showed reduced utilization of ment. The studies did not enroll patients with administered iron therapy. IV vitamin C has not ESA hyporesponsiveness, so it is not known been evaluated in patients with PD-CKD. Therefore, given the low quality of evidence what effect, if any, androgens would have in such for efficacy and unresolved concerns for serious patients. It is unclear whether any enhanced AEs of chronic administration, the Work Group erythropoietic effect caused by androgens would concluded that there is insufficient evidence to confer clinical benefits that outweigh the poten- recommend use of vitamin C (ascorbate) in the tially significant AEs of androgens or the effects management of anemia in patients with CKD. of simply allowing patients to remain at somewhat lower Hb levels without androgens. Short- Androgens term and long-term toxicity of androgens limit Androgens should not be used as an adjuvant their use, especially in women. to ESA treatment in anemic patients with CKD. In short, evidence for efficacy of androgens is Before the availability of epoetin therapy, andro- characterized by serious method limitations, impor- gens were used regularly in the treatment of tant inconsistencies, and sparse data (Table 37). S76 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

Table 37. Use of Androgens as an Adjuvant to ESA Treatment in the HD-CKD Population

Directness of Summary of Findings Methodological Evidence Quality of No. of Studies & Total Quality of including Other Evidence for Importance of Outcome Study Design N Studies Consistency Applicability Considerations Outcome Qualitative Description of Effect Size Outcome No consistent benefit. Only 1 of 3 papers Hb/Hct Level/ Serious Important showing significant between-arm 3 RCTs 63 Directc,d Sparse data Very low Moderate ESA dose limitationsa inconsistenciesb comparison in Hb/Hct (difference in Hb of 1.6 g/dL).c CKD population: 3 RCTs Mild to severe AEs noted in RCTs included severe acne, elevated AST, discomfort at injection site. This is consistent with reported AEs AEs 63+ from androgens in non-CKD populations which include virilization, priapism, peliosis hepatis, liver enzyme abnormalities, and hepatocellular High General population: carcinoma. Mechanism of action and profile of AEs of androgens are believed to be similar in non-CKD and CKD populations. trials, case reports, narrative reviews Balance of Benefit and Harm: Quality of Overall Evidence: No Net Benefit Very Low for Benefit

Footnotes: a. 1 Grade B and 2 Grade C studies. b. Statistically significant effect with only 1 of 3 studies. c. The studies used different ESA and iron protocols and had different definitions for the Hb/Hct outcome. d. The studies used different ESA and iron protocols and had different definitions for the Hb/Hct outcome.

The Work Group, as a result, considered the the utility, if any, of pentoxifylline as an adjuvant quality of evidence to be very low. The Work to ESA treatment in patients with HD-CKD Group judged mild to severe drug-related AEs in remains to be determined. the target and general population as being highly important. Given the very low quality of evi- Vitamin Supplements Other Than Vitamin C dence for efficacy and demonstrated AEs of Although deficiencies of vitamin B12 and fo- androgen therapy, the Work Group concluded late are recognized causes of anemia and rarely that androgens should not be used as an adjuvant are the basis for ESA hyporesponsiveness, there to ESA treatment in anemic patients with CKD. are no RCTs showing that supplementation with (Strong Recommendation) vitamin B12, folate, or other vitamins (in the absence of documented vitamin deficiency) is an Other Pharmacological Agents Not Addressed effective adjuvant to ESA therapy. One RCT of in Guideline Statements ESA-treated HD patients did not find a benefit of 227 Statins pyridoxine (vitamin B6) as an ESA adjuvant. A growing body of literature indicates that Summary there may be clinically important, nonÐlipid- lowering effects of the hydroxymethylglutaryl There was insufficient evidence for efficacy to coenzyme A reductase inhibitors, or statins, in- recommend use of statins, pentoxifylline, and cluding antiproliferative, anticoagulant, immuno- vitamin B12 and folate supplements (other than suppressive, anti-inflammatory, antioxidant, and when used to correct documented vitamin defi- cytoprotective effects. Because a component of ciency) as adjuvants to ESA therapy in patients anemia in patients with CKD may be related to with CKD. Because these therapies are not in underlying inflammatory processes, a potential widespread use as ESA adjuvants, they are pre- role for statins in enhancing epoetin therapy may sented here primarily for general information, be plausible. Only a single small retrospective but were not considered by the Anemia Work study addressed statins as adjuvants to ESA Group for inclusion in Guidelines or CPRs ad- therapy.225 Given the nature of this study, further dressing ESA adjuvants. investigation of the effects of statins as ESA Modifications of Dialysis Treatments Not adjuvants may be warranted, but their utility, if Addressed in Guideline Statements any, remains to be determined. The effects of modifications of the HD dialysis Pentoxifylline prescription and various components of the HD One small uncontrolled open-label study of 16 treatment on anemia in patients with HD-CKD patients with CKD (HD, PD, and transplant have been studied. Unlike the pharmacological recipients) with ESA-resistant anemia evaluated agents discussed, it is not likely that these dialysis the effect of pentoxifylline as an ESA adju- treatment modifications would be undertaken for vant.226 Further studies may be warranted, but the primary purpose of enhancing ESA responsive- ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S77 ness, and they therefore were not considered by the dialyzers were compared.228 Kt/V values were Anemia Work Group for inclusion in Guidelines or similar in the 2 groups at baseline and the end of CPRs addressing ESA adjuvants. the 6-month study. Hb levels were higher and Distinguishing between the effects of in- ESA doses were lower in the high-flux group by creased HD dose (ie, urea reduction ratio [URR] 3 months and remained so to the end of the study. or Kt/V) from effects of concomitant changes in membrane on response to ESA therapy is diffi- Vitamin E cult because most studies compared different Vitamin E has been considered as a potential Kt/V levels using membranes of different types. adjuvant to ESA therapy based on the consider- The effects of dialysate composition, primarily ation that antioxidant properties of vitamin E in comparisons of standard bicarbonate dialysate may prolong red blood cell life span in patients to ultrapure dialysate, also have been evaluated. with CKD and anemia. Oral vitamin E has not No RCTs have been reported that compared HD been studied in prospective controlled trials of and PD (or different doses of PD) on anemia ESA-treated patients. Vitamin EÐbonded dia- outcomes or ESA dose in ESA-treated patients. lyzers were studied in a single RCT in which the primary focus was the effects of a vitamin HD Intensity (“dose”), Membrane Type, and EÐbonded hemodialyzer on carotid artery athero- Other Dialysis Modifications sclerosis and rheological properties of red blood A few studies have examined the relationship cells.237 Mean ESA dose decreased after 1 year between dialysis dose and dialysis membrane on the vitamin EÐbonded dialyzers, but Hb levels type and anemia outcomes,228-236 with conflict- and other important parameters, such as amount ing results. Because in most of these studies, of iron administered, were not reported. The patients in the higher versus lower Kt/V (or safety of vitamin E also must be considered: a URR) groups were dialyzed with membranes of recent meta-analysis in the nonÐkidney-disease different composition and flux, the role of dialy- population suggested that doses of 400 U/d or sis dose cannot be separated from the effects of more of vitamin E were associated with an in- membrane permeability or biocompatibility. One crease in all-cause mortality.238 of these studies, which was not an RCT, compared different levels of Kt/V with the same Ultrapure Dialysate membranes236 and observed an inverse correla- Inflammatory cytokines are proposed to inter- tion between ESA dose and Kt/V, but no relation- fere with the erythropoietic effect of ESAs both ship between Kt/V and Hb level, in HD patients directly and through impaired mobilization and treated with unsubstituted cellulose membranes. utilization of iron. An inflammatory stimulus in In a subsequent report, which also was not an HD patients may be endotoxin or bacterial con- RCT,235 it was suggested that this relationship tamination of dialysate. Standards for bacterial holds only for patients with a Kt/V less than and endotoxin content of water used for dialysis 1.33, and above this level, there was no correla- and for dialysate vary around the world. For tion between Kt/V and ESA dose. dialysate, recently revised voluntary standards Three RCTs were performed in HD patients in (Association for the Advancement of Medical which anemia-related outcomes were compared Instrumentation) include an upper limit for bacte- for high-flux and low-flux dialyzers.229-231 No ria of 200 CFU/mL, and for endotoxin, of 2 difference in Hct or ESA responsiveness was EU/mL.239 In some countries, limits of 100 found between groups. Assessment of these stud- CFU/mL and 0.25 EU/mL for bacteria and endo- ies is complicated by the use of membranes of toxin have been applied, respectively. Ultrapure different composition, attainment of different dialysate has 0.1 CFU/mL or less of bacteria and Kt/V levels, short study duration, and inclusion less than 0.03 EU/mL of endotoxin.239,240 Ultra- of patients not on ESA therapy. In another RCT pure dialysate is produced by generation of mi- of HD patients who were unable to reach a target crobiologically purer water than used for Hb level of 11 g/dL or greater with at least 200 standard dialysate, minimizing potentially con- U/kg/wk of recombinant human erythropoietin taminating biofilm, and use of ultrafilters. Some (rHuEPO), low-flux and high-flux polysulfone uncontrolled observations suggest that the re- S78 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS sponse to ESA treatment may be enhanced by the changing from conventional HD to daily or noc- use of ultrapure dialysate solutions.241 Three turnal HD specifically enhances ESA responsive- RCTs examined the effects of using online- ness in ESA-hyporesponsive patients. produced or filtered ultrapure dialysate on anemia outcomes in HD patients.242-244 ESA doses Peritoneal Dialysis were significantly decreased by up to 33%. The Although observational data have suggested that patients treated with PD may have lower use of ultrapure dialysate typically was associ- 255,256 ated with lower C-reactive protein and interleu- ESA requirements than HD patients, no kin 6 levels compared with standard dialysate, controlled trials have been reported comparing thought to be indicative of reduced inflammatory HD and PD or different doses of PD on anemia responses. outcomes or ESA dose in ESA-treated patients. Hemodiafiltration Summary Hemodiafiltration (HDF) has been evaluated Whereas the Anemia Work Group does not prospectively in a few RCTs with conflicting recommend changing patients with HD-CKD results. A small randomized study that compared from standard bicarbonate dialysate to ultrapure acetate-free biofiltration and low-flux HD in ESA- dialysate for the purpose of enhancing ESA treated patients suggested that ESA doses were responsiveness, studies suggest that the use of lower with HDF.245 In a comparison of online ultrapure dialysate results in lower ESA doses in HDF with high-flux HD, no differences in Hb patients with HD-CKD. There is insufficient evi- levels or ESA dose were observed.246 Another dence at this time that other modifications in the study, in which only about 40% of patients were HD prescription or various components of the being treated with an ESA, compared HDF and HD treatment enhance ESA therapy. high-flux HD. It found no difference in Hb levels LIMITATIONS or ESA doses.247 HDF with online production of pyrogen-free solutions also may have an advan- We acknowledge the limitations of these guide- tage in terms or anemia outcomes compared with lines. For lack of evidence, we do not address the conventional solutions.248 potential adjuvant effect of pharmacological agents or alterations in dialysis prescription on Daily and Nocturnal HD anemia outcomes in patients without concomi- There are no RCTs comparing either daily HD tant ESA therapy. Among patients receiving ESA or nocturnal HD with conventional intermittent therapy, available evidence is restricted to pa- HD for effects on anemia or ESA requirements. tients with HD-CKD. Among available ESA In the most recent report from a small nonrandom- agents, evidence is restricted to use of epoetin ized comparison of short daily, long nocturnal, alfa. As reported elsewhere in this document, a and conventional HD, only nocturnal HD pa- guideline needs to be based on both high- or tients had a statistically significant increase in moderate-quantity evidence and consistently Hb levels during 18 months; ESA doses also demonstrated net medical benefit. Therefore, we tended to increase, although the difference was considered appropriately designed, adequately not statistically significant.249 In 1 study, conven- powered RCTs to be the required foundation. tional HD patients with a baseline single-pool Few such RCTs currently are available. Among Kt/V of at least 1.3 were changed to short-daily available RCTs, only hematologic outcomes were dialysis 6 times per week with the same weekly assessed: no evidence exists to confirm the as- dialysis time.250 Weekly Kt/V increased by 31%. sumption that hematologic outcomes gained with In the patients studied for 12 months, mean ESA adjuvant therapy share the same risk-benefit pro- dose decreased 45% compared with baseline, file as those gained with ESA and iron therapy with stable or increased Hb levels. Doses of ESA alone. Finally, although the primary motivation were trending upward between 6 and 12 months behind adjuvant therapy is to decrease cost by of observation. Other studies, none of which decreasing ESA doses, information on the com- were RCTs, reported variable results.251-254 To parative costs and benefits of ESA with and our knowledge, it has not been studied whether without proposed noniron adjuvants is lacking. CPR 3.4.: TRANSFUSION THERAPY

Red blood cell transfusions should be used judi- If red blood cell transfusions are deemed nec- ciously in patients with CKD, especially because essary for the immediate treatment of patients of the potential development of sensitivity affect- with chronic anemia, the goal should be to attain ing future kidney transplantation. However, dean Hb concentration that will prevent inadequate spite the use of ESA and iron therapy, transfusion tissue oxygenation or cardiac failure. When red with red blood cells occasionally is required, in blood cell transfusions are considered for the particular in the setting of acute bleeding. long-term treatment of patients with chronic ane- 3.4.1 In the opinion of the Work Group, no mia, treatment goals (other than to maintain a single Hb concentration justifies or certain Hb concentration) should be determined requires transfusion. In particular, in advance and assessed serially to ascertain the target Hb recommended for whether the goals are being met. In this setting, chronic anemia management (see the physician and patient must consider such Guideline 2.1) should not serve as a questions as: What symptoms and signs are transfusion trigger. caused or aggravated by the anemia? Can these symptoms and signs be alleviated by red blood RATIONALE cell transfusions? What is the minimum level of Anemia commonly is observed in patients Hb at which the patient can function satisfacto- with CKD. The degree of anemia is a reflection rily? Do the potential benefits of red blood cell of the severity of CKD. Anemia impacts on transfusions outweigh the risks (and possibly the cardiac function and is associated with increased inconveniences) for any given patient? In deter- cardiovascular morbidity and mortality and de- mining the risk-benefit ratio for a given patient, creased QOL.262 such factors as lifestyle, the presence of other Typically, the anemia of CKD is chronic, and medical conditions, the likely duration of the patients compensate for the anemia through a anemia, and the patient’s overall prognosis should number of mechanisms. Thus, in determining the be considered. For example, a patient may be need for red blood cell transfusions, it is impor- willing to tolerate a very limited capacity for tant to evaluate the state of compensation of the exertion if anemia is likely to be temporary, but patient. In general, otherwise healthy individuals not if the anemia will be permanent. display few symptoms or signs of anemia at rest In general, risks per unit of red blood cells when Hb level is greater than 7 to 8 g/dL, although transfused are the same as in any setting. A they may show dyspnea with exertion. At 6 g/dL, number of retrospective studies have identified most patients will report some weakness and, risks related to aggressive transfusion support. A with progressive decreases in Hb values, dys- review of patients with acute coronary artery pnea at rest and congestive heart failure (CHF) syndromes revealed a greater mortality rate in occur.263 Also, the risk for relevant tissue hyp- transfusion recipients.264 In the presence of se- oxia increases, particularly in the presence of vere chronic anemia, transfusion may lead to vascular disease. Therefore, in general, decisions CHF, particularly in the elderly. In such cases, concerning transfusion are not acute, and there is red blood cell transfusions must be administered an opportunity to consider the risks and benefits very slowly, and, in patients with HD-CKD, of transfusion as treatment. transfusion during hemofiltration may be re- Before considering transfusion of red blood quired. The administration of many red blood cells for the treatment of chronic anemia, it is cell transfusions over a prolonged period can essential to assess signs and symptoms and deter- eventually lead to iron overload. mine the cause of the anemia so that, when The use of ESAs can greatly reduce the need for appropriate, treatment other than red blood cell red blood cell transfusions in patients with anemia transfusions may be used. Classic examples of of CKD when target Hb concentrations are reached anemias that may be severe, but correctible by and maintained.265,266 With the advent of new alternative therapies, are iron-deficiency anemia immunosuppressant regimens after 1995, the ben- and pernicious anemia in adults. efits of pretransplantation transfusion have been

American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S79-S80 S79 S80 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS rendered largely obsolete. There is some evi- should be necessary only for patients with acute dence that donor-specific transfusion with living bleeding (usually GI), acute hemolysis, or severe donor transplantation improves survival, but the inflammation or blood loss through surgery, and decision to perform donor-specific transfusion then only in an emergency or if the patient must still be made on a case-by-case basis. Blood exhibits a rapid decline in condition. Interna- transfusions can induce antibodies to histocom- tional guidelines provide criteria for deciding patibility leukocyte antigens that can reduce the when transfusion is necessary.269-271 success of kidney transplantation; thus, transfu- Patients with CKD on HD therapy are more sions generally should be avoided in patients likely to need blood transfusions than those on awaiting a renal transplant.267 If deemed essen- PD therapy because of the HD procedure itself. tial, red blood cell transfusions in this patient Patients on HD therapy lose blood from frequent group should be conducted in line with published blood tests, trapped blood in the dialyzer and recommendations.268 tubing,272 and increased risk for GI bleeding If therapy with an ESA is started at the Hb from anticoagulants. However, aggressive iron concentration recommended in these guidelines replacement has largely eliminated the need for and Hb levels are maintained at the recom- red blood cell transfusions, even for patients on mended target concentrations, blood transfusions HD therapy. CPR 3.5. EVALUATING AND CORRECTING PERSISTENT FAILURE TO REACH OR MAINTAIN INTENDED HB

Although relative resistance to the effect of ESAs fested by persistent, below-target Hb levels de- is a common problem in managing the anemia of spite substantial ESA doses or by within-target patients with CKD and is the subject of intense Hb levels attained only at very high ESA doses. interest, the bulk of available information sug- In patients with HD-CKD undergoing ESA therapy, gests that—in the absence of iron deficiency— Hb levels less than 11 g/dL are associated with there are few readily reversible factors that con- increased mortality and hospitalization rates, and tribute to ESA hyporesponsiveness. failure to achieve an Hb level greater than 11 g/dL is a poor prognostic sign. Given the dispro- 3.5.1 Hyporesponse to ESA and iron portionate burden of morbidity and mortality that therapy: the hyporesponsive patient population bears and In the opinion of the Work Group, the the ESA expense that hyporesponsiveness engen- patient with anemia and CKD should ders, hyporesponsiveness to ESAs deserves more undergo evaluation for specific causes scrutiny than it has received. Although most of hyporesponse whenever the Hb disorders associated with hyporesponsiveness are level is inappropriately low for the readily apparent, a review of available informa- ESA dose administered. Such condition on patients with coexisting hematologic or tions include, but are not limited to: oncological disorders may be worthwhile. Simi- ● A significant increase in the ESA larly, a rare disorder, PRCA, deserves special dose requirement to maintain a consideration. certain Hb level or a significant decrease in Hb level at a constant Antibody-Mediated PRCA ESA dose. ● A failure to increase the Hb level to Rarely, patients undergoing ESA therapy greater than 11 g/dL despite an develop antibodies that neutralize both ESA ESA dose equivalent to epoetin and endogenous erythropoietin. The resulting greater than 500 IU/kg/wk. syndrome, antibody-mediated PRCA, is char- 3.5.2 Evaluation for PRCA: acterized by the sudden development of severe In the opinion of the Work Group, transfusion-dependent anemia. Rapid recogni- evaluation for antibody-mediated tion, appropriate evaluation, and prompt inter- PRCA should be undertaken when a vention can be effective in limiting the conse- patient receiving ESA therapy for quences of this life-threatening hyporesponse more than 4 weeks develops each of condition. the following: Antibody-mediated PRCA, although rare in ● Sudden rapid decrease in Hb level patients administered ESAs, received urgent at- at the rate of 0.5 to 1.0 g/dL/wk, or tention after 1998. Between 1989 and 1998, three requirement of red blood cell trans- reports described the development of PRCA in a small number of patients with CKD adminis- fusions at the rate of approximately 273,274 1 to 2 per week, AND tered ESAs. Reports of PRCA increased ● sharply in 1998 and reached a peak in 2002 Normal platelet and white blood 273,275 cell counts, AND (Fig 17). These reports were associated ● Absolute reticulocyte count less with SC administration of an epoetin alfa formu- than 10,000/␮L. lation not available in the United States. In 1998, in this formulation, polysorbate 80 was used to BACKGROUND replace human albumin. Preparations of this product included single-dose syringes fitted with un- Hyporesponsiveness to ESAs coated rubber stoppers and prefilled with epoetin Hyporesponsiveness to ESAs is a common alfa. Results of an intensive investigation indi- finding of grave significance whether it is mani- cate that in the presence of polysorbate 80 and

American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S81-S85 S81 S82 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

The incidence rate of PRCA in patients who were exposed to SC-administered ESA from syringes with uncoated stoppers and polysorbate 80 is estimated at 4.23 cases/10,000 patient- years.275 The incidence with SC use of all other forms of SC-administered ESA is estimated to be 0.5 cases/10,000 patient-years.275 The finding that antibody-mediated PRCA develops only rarely, even among patients exposed to adversely Fig 17. Exposure to Eprex® and case counts of PRCA. Relationship between reporting rate of new equipped syringes, suggests that additional fac- cases of PRCA and route of administration (SC versus tors must have been involved to render ESA IV), stabilizer (human serum albumin versus polysor- immunogenic, potentially including differences bate 80), and coated versus uncoated stoppers in preparations of Eprex®, a form of epoetin alfa mar- in host immunoreactivity or product storage and keted outside the United States. Reprinted with permis- handling. sion.275 RATIONALE uncoated rubber can release organic compounds that may act as immunoadjuvants, thereby in- Hyporesponse to ESA and Iron Therapy creasing the immunogenicity of SC-administered epoetin alfa.273-275 The patient with anemia and CKD should Between 2001 and late 2003, single-dose sy- undergo evaluation for specific causes of hypore- ringes with polysorbate 80 and uncoated stop- sponse if Hb level is persistently less than 11 pers were replaced by syringes with fluoro-resin- g/dL AND if ESA doses are equivalent to epoetin coated stoppers.275 In addition, SC administration greater than 500 IU/kg/wk. Results from the of the agent had been prohibited in Europe and USRDS national data system show that the distri- discouraged in Canada. By 2004, the incidence bution of epoetin alfa doses is quite broad, the of new antibody-mediated PRCA had decreased number of administrations per percentile range is to pre-1998 levels. relatively constant over the full spectrum of Isolated cases of PRCA have been observed in doses, and the relationship between percentile association with the use of other ESAs.273,274,276-279 range and mean epoetin dose per administration No case of antibody-associated PRCA has been is distinctly nonlinear (Fig 18). In these uns- documented in patients treated with only IV elected patients, the 99th percentile doses are 30 administration of ESAs.274 An increase in anti- times greater than the 1st percentile doses, and body-mediated PRCA has not been seen among the top 20% of patients seem to be using a patients in the United States, where the immuno- disproportionate amount of ESA compared with genic formulation has not been available. the lower 80%.

Fig 18. Mean epoetin dose per patient per administration by percentile of dose (1st, 5th to 95th, and 99th). Data are for December 2004, courtesy of USRDS. ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S83

Available information on hyporesponsiveness In short, the available evidence suggests that is weakened by the lack of a validated quantita- approximately 20% of patients with HD-CKD tive measure of ESA resistance. Although a resis- in the United States are administered ESA tance index—calculated by dividing the weight- doses in excess of an epoetin equivalent of adjusted ESA dose by Hb level—has been 30,000 IU/wk, or 428 IU/kg/wk for a 70-kg proposed, precisely how the ESA dose and Hb patient (Fig 2). Approximately 10% of patients level should be determined (area under the curve, with an Hb level less than 11 g/dL will persis- single value, time averaged) has not been stan- tently fail to attain a target Hb level of 11 g/dL dardized and reference values have not been or greater. However, among US patients with validated. an Hb level less than 11 g/dL and ESA doses in Factors most commonly associated with persis- excess of 30,000 IU/wk epoetin equivalents, tent failure to achieve target Hb levels for at least less than 1% will remain at less than the Hb 6 months despite ESA therapy include the follow- target for 6 months. 280 ing : The patient with anemia, CKD, and a preexist- ● Persistent iron deficiency ing hematologic disorder represents an uncom- ● Frequent hospitalization mon, but challenging, cause of ESA hyporespon- ● Hospitalization for infection siveness and deserves special consideration. The ● Temporary catheter insertion quality of reviewed material is insufficient to ● Permanent catheter insertion provide specific recommendations. However, a ● Hypoalbuminemia brief review of the available literature may prove ● Elevated C-reactive protein level. helpful to medical decision making in the treatment of these patients. In general, these problems do not pose a diagnostic challenge or yield to simple solutions. Management of anemia in patients with CKD A second set of disorders also may be identified with preexisting hematologic disorders associ- among hyporesponsive patients. Unfortunately, ated with anemia may present specific prob- they also are found among those who do not lems because of multifactorial causes. In some meet criteria for hyporesponsiveness, and they patients, anemia may result predominantly from also represent neither frequent nor elusive diag- low endogenous erythropoietin levels and can noses280: be corrected readily by administration of ESAs. In other disorders, impaired marrow function, ● Pancytopenia/aplastic anemia ineffective erythropoiesis, and shortened red ● Hemolytic anemia blood cell survival may contribute to anemia ● Chronic blood loss and ESA hyporesponsiveness. ● Cancer, chemotherapy, or radiotherapy We identified a small number of publica- ● Inflammatory disease tions that evaluated ESA responsiveness in ● Acquired immune deficiency syndrome patients with CKD and preexisting hemato- ● Infection. logic disorders (Table 38). The publications Only a small percentage of patients with a are predominantly observational, taking the Hb level less than 11 g/dL fail to respond to form of individual case reports or reports of a ESA therapy. Persistency analysis shows that small series of patients. Although the available approximately 10% of patients with HD-CKD information is insufficient to support guide- who enter a 6-month period with a Hb level lines or CPRs, the following summary state- less than 11 g/dL remain at less than that ments may be helpful: threshold for 6 consecutive months.112 Among Thalassemia patients with a Hb level less than 11 g/dL administered high ESA doses (epoetin alfa Ͼ ● Patients with thalassemia have a poor 30,000 IU/wk), only 0.6% of patients re- response to ESA.281-283 mained at less than that threshold for 6 consecu- ● Higher doses of ESA are required to achieve tive months. target Hb levels.281-284 S84 ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS

Table 38. Published Experience in Patients With Anemia, CKD, and a Preexisting Hematologic Disorder

Adult/Ped/ Author, Year Location Mean Age N Diagnosis Type of Study EPO Response Higher doses of ESA to achieve Thal-t (n = 4 ) Matched controls Different doses Cheng, 1993 281 China Adult 4 target Hb 1 HbH tnatsiseR ot ASE Lai, 1992 283 China Adult 4 α-Thal-T, β-Thal-m (n = 36) Randomly selected Different doses Poor response of Thal patients ESA dose requirement higher in Di Iorio, 2003 282 Italy Adult 10 β-Thal-m Lab quantification of Hb chains Different doses patients with high levels of anomalous Hb chains CKD Stage 5 National logitudinal cohort Epo requirement Less response, higher ESA ym elpitluM ym e amol Powe, 1993 288 US Adult 59,462 requirement Less response, higher ESA elkciS c lle esaesid requirement No sickle crisis on ESA, Hb S Tomson, 1992 289 UK Adult 3 Sickle cell disease Case reports Increasing doses increased, no Tf for 12 weeks Tarng, 1997 296 Taiwan Adult 1 Hb J-Meinung Case reports Poor response to ESA Drueke, 1990 297 ecnarF weiveR Ataga, 2000 298 SU weiveR Drueke, 1991 299 cnarF e R ive ew Patients became Tf independent on Di Iorio, 2004 284 Italy 12 β-Thal-m Case series Increasing doses high doses of ESA, sustained effect Need for ESA increases with Di Iorio, 2003 282 Italy 10 β-Thal-m Case series HbA2 levels and Epo presence of high levels of Hb A2 chains (see Ataga, 2000 298) 5 β-Thal-t Lab for endogenous Epo Lab study β-Thal major high levels of ESA Paritpokee, 2002 285 Thailand Children 51 bH E t- 55 β-Thal-m/Hb E Endogenous Epo 3 Homocyg sickle cell disease Intervention with HU ESA levels increase on HU Papassotiriou, 2000 290 Greece levels 10 Hb S/β-Thal Both are increased in patients with Papassotiriou, 1998 287 erG ce e 02 bH H C esa es r sei Epo dna RfT slevel HbH/ineffective erythropoiesis/hemolysis High ESA levels in S-β-Thal Katopodis, 1997 286 Greece 16-59 17 S-β-Thal Comparison with normal controls Epo levels, Hb compared to normal, correlation with Hb

● Patients can become independent of trans- Multiple Myeloma fusions on high doses of ESA.284 ● Higher levels of ESA are required to ● The need for ESA increases in patients achieve target Hb levels in patients with with high levels of Hb chains in the A2 HD-CKD.288 serum.282 ● Therapy with ESAs reduces transfusions ● Patients with ␤-thalassemia major show and improves QOL in anemic patients with high endogenous erythropoietin levels.285 or without kidney disease.291 ● Patients with HbS/␤-thalassemia have a higher-than-normal endogenous erythropoi- General Comments 286 etin level. Patients with ␤-thalassemia and Hb S disease Hb H may require higher doses of ESA compared with patients with CKD who do not have hematologic ● 281,287 Patients are resistant to ESA. disorders. There is no apparent contraindication ● Endogenous serum erythropoietin levels to increase the ESA dose in either disorder. Some and serum transferrin receptor levels are patients may require doses usually used in hema- 287 both increased. topoietic disorders (40,000 to 60,000 U/wk). ● The mechanism of anemia is likely related to ineffective erythropoiesis and Evaluation for PRCA 287 hemolysis. Evaluation for antibody-mediated PRCA should be undertaken when a patient receiving Hb S ESA therapy for more than 4 weeks develops ● Patients show higher ESA dose require- each of the following: sudden rapid decline in ment to achieve target Hb level.288,289 Hb level at the rate of 0.5 to 1.0 g/dL/wk, or ● There may be a lower incidence of sickle requirement of red blood cell transfusions at the cell crises while on ESA therapy.289 rate of approximately 1 to 2 per week; normal ● Endogenous erythropoietin levels increase platelet and white blood cell counts; and abso- with hydroxyurea therapy.290 lute reticulocyte count less than 10,000/␮L. ANEMIA IN CHRONIC KIDNEY DISEASE IN ADULTS S85

Syndrome Recognition Management and Treatment The characteristic sign of PRCA is almost It is prudent to discontinue the administration complete cessation of erythropoiesis. Accord- of any ESA product in patients with suspected ingly, a patient affected with PRCA evidences a and confirmed diagnosis because the antibodies decrease in Hb level of about 0.1 g/dL per day are cross-reactive and continued exposure may and a reticulocyte count less than 10,000/␮L, lead to anaphylactic reactions.292 Patients likely consistent with the normal rate of red blood cell will require transfusion support. Treatment with destruction and an absence of red blood cell immunosuppressive approaches is effective in a production. Because nonerythroid marrow is un- significant number of patients.293 Renal allo- affected, leukocyte and platelet counts are ex- grafts usually result in a rapid decrease in anti- pected to be normal.274,276 body titers associated with therapeutic benefit.293 It is not clear whether treatment with ESA can be Diagnostic Evaluation resumed safely after clearance of the antibody. Recommendations based on expert opinions have Based on very preliminary data, this might be been published to guide the workup and therapy feasible; however, caution is advised.294 A single of patients suspected to have antibody-mediated case also has been reported in which Hb levels PRCA.274,277,279 The definitive diagnosis is de- could be restored in the presence of antierythro- pendent upon demonstration of the presence of poietin antibodies by switching to a different neutralizing antibodies against erythropoietin. product.295 III. CLINICAL PRACTICE RECOMMENDATIONS FOR ANEMIA IN CHRONIC KIDNEY DISEASE IN CHILDREN

CLINICAL PRACTICE the best evidence available regarding issues re- RECOMMENDATIONS FOR ANEMIA IN lated to the identification, diagnosis, initial evalu- PEDIATRIC PATIENTS WITH CKD ation, and strategies for treatment and monitoring of children with anemia and CKD stages 1 to INTRODUCTION 5, including those treated with dialysis. Our review is not exhaustive and our intent is not to Presentation of clinical practice recommenda- substitute for textbooks. Specific details on poten- tions for anemia in pediatric patients with CKD tial dosing regimens, therapeutic choices, and is warranted because the pediatric patient popula- practice options in caring for children with ane- tion, from newborn through adolescence, differs mia and CKD are found elsewhere.301 substantially from the adult population in key Please note that to be consistent with the use metabolic, growth, developmental, and psycho- of Hb levels, as opposed to Hct, in these new logical factors.300 Nevertheless, providers caring guidelines, all Hct values from studies quoted for adult and pediatric patients with CKD largely have been converted to Hb equivalents by a share the same topics of concern regarding the diagnosis and management of anemia. More- factor of 0.3 g/dL per percent of Hct; eg, an Hct over, the bulk of information to support CPRs of 33% is converted to an Hb level of 9.9 g/dL. and the only evidence of sufficient strength to All statements in the pediatric section assume support evidence-based guidelines are available the preface In the opinion of the Work Group, and from studies of the adult patient population. all statements are provided as CPRs because Given the distinct needs of pediatric patients, there is insufficient evidence in pediatric patients shared topics of concern among providers for to support evidence-based guidelines. When, in both pediatric and adult patients, the generally the opinion of the Work Group, the adult guide- low quality of evidence in pediatric patients, and line statement applies equally well to adults and the unavoidable need to generalize from evi- children, the statement is accompanied by the dence in adults, the Work Group chose to present following: CPRs in pediatric patients as a separate section, (FULLY APPLICABLE TO CHILDREN) using adult guidelines as a frame of reference, When, in the opinion of the Work Group, the changing recommendations when appropriate, adult guideline statement needs modification or and describing available evidence in pediatric adjustment for children, the following instruc- patients under the guideline rationale. tion is given, followed by the pediatric-specific Here, we address issues pertinent to children guideline statement. with anemia and CKD at all stages of disease. (APPLICABLE TO CHILDREN, BUT Our goal is to offer recommendations based on NEEDS MODIFICATION)

S86 American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: p S86 CPR FOR PEDIATRICS 1.1: IDENTIFYING PATIENTS AND INITIATING EVALUATION

Identifying anemia is the first step in evaluating function decreases. From USRDS 2004 data, we the prognostic, diagnostic, and therapeutic signifi- know that the mean eGFR at initiation of dialysis cance of anemia in patients with CKD. therapy in children is 10.3 mL/min/1.73 m2 and 1.1.1 Stage and cause of CKD: (FULLY that, at this point, 35% to 40% of children are APPLICABLE TO CHILDREN) already receiving ESA therapy.28 One study of In the opinion of the Work Group, Hb 35 children with various degrees of renal disease testing should be carried out in all examined the role of erythropoietin and potential patients with CKD, regardless of stage inhibitors of its action. The study suggested that or cause. in this group, the risk for anemia increased at a 1.1.2 Frequency of testing for anemia: GFR less than 35 mL/min/1.73 m2.302 Unfortu- (FULLY APPLICABLE TO CHIL- nately, unlike the adult literature, there is no DREN) better direct or observational evidence that clearly In the opinion of the Work Group, Hb delineates a particular cutoff level of function for levels should be measured at least which the risk for CKD-associated anemia in- annually. creases significantly in children. 1.1.3 Diagnosis of anemia: (APPLICABLE TO CHILDREN, BUT NEEDS MODI- Frequency of Testing for Anemia FICATION) ADULT CPR This guideline is considered applicable to chil- In the opinion of the Work Group, dren because there is no direct evidence to sup- diagnosis of anemia should be made port a different recommendation. However, it is and further evaluation should be un- often the case that children (especially younger dertaken at the following Hb concen- ones) will receive more frequent laboratory moni- trations: toring of their CKD status caused, in part, by ● <13.5 g/dL in adult males. expected changes in values during growth. ● <12.0 g/dL in adult females. PEDIATRIC CPR Diagnosis of Anemia In the opinion of the Work Group, in In terms of defining a lower limit of “normal” the pediatric patient, diagnosis of ane- Hb level before initiating a workup for anemia in mia should be made and further evalu- a child with CKD, it seems reasonable to apply ation should be undertaken whenever the same approach, although not the same val- the observed Hb concentration is less ues, as in the adult Guidelines. In other words, than the fifth percentile of normal anemia in a child with CKD should be diagnosed when adjusted for age and sex. and evaluated at such time that the child’s Hb level decreases to less than the fifth percentile for RATIONALE their age and sex. Adjustment in normal Hb Stage and Cause of CKD levels—and hence the definition of anemia—for There is no evidence in the pediatric CKD children living at higher altitudes likely is reason- population to contradict the principles as out- able, as in the adult Guidelines, although age- lined in the Adult Guideline 1.1 with respect to specific pediatric data are not available. the use of Hb level, not Hct, to define anemia and The normative values for this definition in the value of obtaining this at a standard time (eg, children older than 1 year of age have been taken before dialysis during the midweek run) in pa- from the NHANES III reference data, as in tients on HD therapy. adults (Table 39), whereas the values for children It also is clear that children with CKD are from birth to 1 year are taken from data compiled more likely to develop anemia as their renal elsewhere (Table 40).303

American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S87-S88 S87 S88 Table 39. Hb Levels (g/dL) in Children Between 1 and 19 Years for Initiation of Anemia Workupa Anemia Definition Met if Value is <5th All Races/Ethnic Groups Number of Subjects Mean Standard Deviation Percentile BOYS 1 ry na d revo 21 326, .41 7 4.1 21 1. 1 2- y r 931 12 0. 8.0 .01 7 ry 5-3 ry 1 2, 81 1 .2 4 8.0 1 .1 2 6-8 ry 709 9.21 8.0 5.11 ry 11-9 ry 377 3.31 8.0 0.21 ry 41-21 ry 045 1.41 1.1 4.21 91-51 yr 8 63 51 .1 1.0 31 .5 GIRLS revo dna ry 1 ry dna revo 947,31 2.31 1.1 4.11 ry 2-1 ry 58 8 21 .0 8.0 01 .8 ry 5-3 ry 733,1 4.21 8.0 1.11 ry 8-6 ry 576 8.21 8.0 11 .5 ry 11-9 ry 437 1.31 8.0 9.11 12-14 yrb 26 1 31 .3 0.1 11 7. 15-19 yrb 059 2.31 0.1 5.11 a. Based on NHANES III data, United States, 1988-94; data abstracted from Tables 2 & 3.40 b. Menstrual loses contribute to lower mean and 5th percentile Hb values for group. NMAI HOI INYDSAEI CHILDREN IN DISEASE KIDNEY CHRONIC IN ANEMIA

Table 40. Hb Levels (g/dL) in Children Between Birth and 24 Months for Initiation of Anemia Workup a A eg naeM bH 2– DS b mreT ( oc rd lb )doo 61 .5 5.31 -1 3 d 1 5.8 14 5. kw 1 kw 5.71 5.31 kw 2 kw 16.5 5.21 om 1 om .41 0 0.01 m 2 m o 5.11 0.9 om 6-3 om 5.11 5.9 42-6 mo 21 .0 01 .5 a. Data taken from normal reference values.303 b. Values 2 standard deviations below the mean are equivalent to <2.5th percentile CPR FOR PEDIATRICS 1.2: EVALUATION OF ANEMIA IN CKD

Anemia in patients with CKD is not always adequacy of iron for erythro- caused by erythropoietin deﬁciency alone. Initial poiesis. laboratory evaluation therefore is aimed at iden- PEDIATRIC CPR tifying other factors that may cause or contribute In the pediatric patient, serum to anemia or lead to ESA hyporesponsiveness. TSAT to assess adequacy of 1.2.1 In the opinion of the Work Group, iron for erythropoiesis. initial assessment of anemia should include the following tests: (APPLI- RATIONALE CABLE TO CHILDREN, BUT NEEDS There is no evidence to support any different MODIFICATION) recommendations in children compared with 1.2.1.1 A CBC including—in addi- adults with respect to statements 1.2.1.1 through tion to the Hb concentration— 1.2.1.3. However, tests other than TSAT that are red blood cell indices (MCH, included in the recommendation for assessment MCV, MCHC), white blood of iron available for erythropoiesis in adults have cell count and differential and not been well studied in the pediatric CKD platelet count. population, with only 2 studies examining the 1.2.1.2 Absolute reticulocyte count. use of CHr in HD patients304,305 and none evalu- 1.2.1.3 Serum ferritin to assess iron ating the use of PHRCs. This would seem to stores. suggest that until further data are available, these 1.2.1.4 ADULT CPR tests remain research-based methods in children Serum TSAT or CHr to assess with CKD.

American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: p S89 S89 CPR FOR PEDIATRICS 2.1: HB RANGE

Treatment thresholds in anemia management de- development, school attendance, exercise capac- scribe the intended goal of current treatment for ity, and family support. To assist medical deci- the individual patient. The Hb treatment range sion making, the Work Group provides the follow- represents the intended goal of ESA and iron ing review of the literature. therapy. 2.1.1 Lower limit of Hb: (FULLY APPLI- Mortality CABLE TO CHILDREN) Observational evidence relating Hb level to In patients with CKD, Hb level should mortality is available. Children in the North Ameri- be 11.0 g/dL or greater. (MODER- can Pediatric Renal Transplant Cooperative Study ATELY STRONG RECOMMENDA- (NAPRTCS) database from 1992 to 2001 with an TION) Hb level less than 9.9 g/dL compared with those 2.1.2 Upper limit of Hb: (FULLY APPLI- with an Hb level greater than 9.9 g/dL showed an CABLE TO CHILDREN) elevated risk for mortality: adjusted RR, 1.52; In the opinion of the Work Group, 95% CI, 1.03 to 2.26; P Ͻ 0.05.306 The relation- there is insufficient evidence to recom- ship between Hb level and mortality, when exam- mend routinely maintaining Hb levels ined at other cutoff values for Hb, appeared at 13.0 g/dL or greater in ESA-treated continuous. Patients with more severe anemia patients. also experienced increased risk for hospitalization (17.2% Ϯ 1.8% versus 12.3% Ϯ 2.1%, RATIONALE respectively; P Ͻ 0.01). Determination of Hb targets in pediatric patients resists definitive recommendation. QOL, The Heart so significant to the development of the child A single RCT provides evidence for the ben- and life of the family, lends urgency to the efit of treatment of anemia with ESA compared consideration of higher Hb level thresholds. with placebo. In a blinded crossover trial of 11 However, evidence lacks both quality and quan- children aged between 2.3 and 12.3 years, under- tity, rendering assessment of both benefit and going HD or PD, and with a baseline Hb level risk uncertain. Age-specific variation in normal between 4.3 and 8.1 g/dL, patients were assigned Hb levels introduces further uncertainty. Finally, to either ESA therapy (Hb Ͼ 10 g/dL) or placebo given key metabolic, growth, developmental, and for 24 weeks.126 Seven patients completed both psychological differences between children and trial arms. ESA therapy was associated with partial adults, exclusive reliance on evidence in adults is correction of an elevated cardiac index by 6 inappropriate. months and a significant reduction in left ventric- The Work Group presents lower and upper ular mass by 12 months. targets for Hb levels in children using values in Two observational studies have examined the adults for reference. However, we add 2 signifi- relationship between anemia and LVH in chil- cant qualifications. The first is that both the dren with CKD.307,308 In these studies, patients lower and upper Hb targets serve only as opinion- with severe LVH (left ventricular mass index Ͼ based CPRs, in keeping with the lack of pediatric- 51 g/m2) showed a statistically lower Hb level specific evidence. The second is that medical than those without LVH (Hb, 9.5 Ϯ 1.8 versus decision making to set Hb targets in individual 10.9 Ϯ 2.3 g/dL; P ϭ 0.027). Left ventricular patients should be informed by available evi- compliance also was related to Hb level in chil- dence that is uniquely pediatric. Consideration dren (r ϭ Ð0.65; P ϭ 0.02). The findings suggest should be given, for example, to the potential that severe anemia in children with CKD stage 5 need to make adjustments for the normal age- leads to chronic increases in cardiac workload specific Hb distribution (Table 39 and Table 40). and a consequent increase in both left ventricular In weighing the potential QOL benefits of Hb end-diastolic volume and mass. targets, the available evidence in adults should In this RCT, exercise capacity improved with be enriched by consideration of QOL issues that ESA treatment (mean achieved Hb, 11.2 g/dL; are crucial to children, including neurocognitive range, 9.5 to 14.2 g/dL) compared with placebo

S90 American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S90-S92 ANEMIA IN CHRONIC KIDNEY DISEASE IN CHILDREN S91 control.126 Measures of capacity significantly In 116 children enrolled in a multicenter trial, affected included a 2-minute walking test (n ϭ 7) sleep, activity (in school, at home, or in social and a formal treadmill testing using the Bruce situations), alertness, feelings, and a summary protocol, full (n ϭ 3) or modified (n ϭ 3). score composed of all subscores were assessed at Distance walked, in meters, approached but did baseline and every 6 months during ESA treat- not reach statistical significance in the ESA arm ment to attain a target Hb level of 9.6 to 11.2 of the crossover, P ϭ 0.06; similar results were g/dL compared with baseline pretreatment Hb seen from both the regular or modified treadmill level of 6.7 g/dL (range, 3.4 to 9.5 g/dL).313 data, P ϭ 0.07. Scores at 1 year after ESA increased after treat- In a nonrandomized interventional trial, 18 ment for all aspects examined, but only the 10% children with CKD stage 5 (15 patients, on HD increase in the summary score achieved signifi- or PD) and a Hb level less than 9.9 g/dL were cance (P Ͻ 0.05). administered IV or SC ESA until Hb level was In the previously described randomized cross- 126 greater than 9.9 g/dL; baseline Hb level of 6.5 Ϯ over trial, a 25-question modified parental 0.8 g/dL changed to a final level of 10.0 Ϯ 0.6 questionnaire used a visual analog scale to assess g/dL; P ϭ 0.001.309 Exercise time (treadmill 5 domains: sleep, school performance, diet, psy- with a modified Bruce protocol) increased signifi- chosocial, and a physical performance/health con- cantly (before ESA, 10.3 Ϯ 1.9 minutes; after struct. Only physical performance and general ESA, 11.2 Ϯ 1.9 minutes; P ϭ 0.01), and resting health showed a significant treatment effect oxygen consumption decreased from 7.8 Ϯ 1.8 (again, baseline Hb, 4.3 to 8.1 g/dL versus final to 6.9 Ϯ 1.4 mL/min/kg; P ϭ 0.01 with the Hb, 11.2 g/dL; range, 9.5 to 14.2 g/dL). higher Hb level. However, there was no change A generic health QOL questionnaire has been in stroke volume, blood pressure, or any cardiac administered in a cross-sectional fashion to the parents of children with CKD stages 1 to 5 and indices after the first month at the higher Hb 312 level. transplant recipients. This instrument, which Similarly, a small cohort (n ϭ 7) of HD has shown internal consistency and concurrent validity and has been used in a single-center trial patients showed an improvement in aerobic work 314 capacity and effort tolerance, as evidenced by of children on HD therapy, measures 12 as- statistically significant changes in the workload pects of health-related QOL in the domains of reached, peak oxygen uptake, and average venti- physical functioning; limitations in schoolwork latory anaerobic threshold after treatment of ane- and activities with friends, general health, bodily pain and discomfort, limitations in family activi- mia with ESA (baseline Hb, 6.3 Ϯ 0.9 g/dL ties; emotional/time impact on the parent; impact versus final Hb, 11.2 Ϯ 1.2 g/dL).310 of emotional or behavior problems on school Finally, 10 children undergoing PD were evalu- work and other daily activities; self-esteem; men- ated before and 18 months after limited correc- tal health; behavior; family cohesion; and change tion of anemia with ESA (baseline Hb, 5.9 Ϯ 0.9 in health. When evaluated as a continuous vari- g/dL versus final Hb, 8.7 Ϯ 1.5 g/dL). Patients able, Hct was linked directly to measures of showed a significant slowing of heart rate, P Ͻ improved health, as seen in the single-item gen- 0.01, but no improvement for other cardiac pa- eral health (r ϭ 0.36; P ϭ 0.003), general health rameters.311 (r ϭ 0.29; P ϭ 0.004), and physical functioning (r ϭ 0.35; P ϭ 0.0004) domains.312 QOL and Neurocognitive Effects Further analysis of the same data, adjusting for The relationship between QOL and anemia has eGFR, age, sex, race, dialysis modality, and been examined in children with CKD.126,312,313 transplantation or chronic renal insufficiency and In 2 trials,312,313 QOL instruments were not dividing the patients into groups with Hb level validated in patients with CKD and were com- greater than or less than 10.8 g/dL (defined by pleted by parents. Only 1 study blinded parent- the investigators as anemic), showed an associa- responders to use of ESA or placebo, although 5 tion between anemia and 4 domains: Parental of 7 parents in that study correctly identified use Impact-Time, Family Activities, RoleÐPhysical, of ESA during the appropriate treatment.126 and Physical Functioning. Further multivariate S92 ANEMIA IN CHRONIC KIDNEY DISEASE IN CHILDREN analysis done after division of the group into Hb g/dL to final, 9.3 g/dL), and no effect was seen tertiles, less than 9.9 g/dL, 9.9 to 10.8 g/dL, and for any measured nutrition or growth parameter, greater than 10.8 g/dL, showed a strong dose- including midarm circumference, triceps skin- response relationship between Hb level and fold thickness, weight gain, or the SD scores for health-related QOL as measured by this instru- growth velocity or height.317 ment. It should be noted that differences between the mid and upper tertile of Hb levels did not Hypertension reach significance, whereas those between the The development or worsening of hyperten- lowest and upper tertile in the areas of Physical sion during treatment with ESAs in children is of ϭ ϭ Functioning (P 0.02), RoleÐEmotional (P significant concern. In a study of children aged 4 ϭ 0.05), RoleÐPhysical (P 0.02), Parental Im- months to 21 years, assignment to either high- ϭ ϭ pactÐTime (P 0.004), Family Activities (P dose (epoetin, 450 IU/kg/wk) or low-dose (150 ϭ 0.003), and Physical Summary score (P 0.01) IU/kg/wk) ESA treatment was associated with a did. (Note: these results have been converted significant increase in diastolic blood pressure by from Hct to Hb values.) week 12 compared with baseline (88 Ϯ 6.7 versus The neurocognitive effects of anemia also 68 Ϯ 17 mm Hg; paired t-test, P ϭ 0.01).319 The have been examined. In healthy children aged 6 investigators reported a nonsignificant trend be- to 11 years, impairment in cognition is associ- tween increasing Hb levels and increasing sys- ated with iron deficiency and an Hb level less 315 tolic and diastolic blood pressures despite stable than 11.8 g/dL. Furthermore, in a multicenter or lower ESA dose. single-arm interventional trial evaluating 22 children with CKD aged 4 months through 16 years, Thrombotic Events treatment of anemia was associated with a significant increase in IQ, determined by using the No thrombotic events, including vascular ac- Weschler intelligence test, although the relative cess thrombosis, were reported in either arm of the mentioned high-dose versus low-dose ESA increase in Hb levels was small (Hb baseline, 319 9.2 Ϯ 1.6 versus final, 9.7 Ϯ 1.7 g/dL; P ϭ trial. However, the small number of patients, 0.007).316 relatively short length of follow-up, and relatively few patients with HD preclude any reliable Nutrition and Growth conclusions about safety. No other larger prospec- No reliable studies have been published that tive trials of pediatric patients treated with ESAs examine the relationship between treatment of to various Hb targets are available to further anemia and nutrition in children. The literature address the issue of clotting or access thrombosis, although it has been reported to occur inter- consists largely of subjective reports by the child 318,320Ð323 or family or short small trials that fail to achieve mittently in children. significant treatment effects. Treatment of anemia apparently fails to re- ESA Therapy and Loss of Renal Function verse growth retardation in children with CKD In patients with ND-CKD assigned to high- undergoing HD or PD.317,318 In a US phase III, dose compared with low-dose ESA treatment, no randomized, double-blinded, placebo-controlled significant difference in creatinine levels was trial of ESA therapy in children, 81 children on seen; in the same trial, between-group compari- HD or PD therapy showed a substantial increase son showed no change in Kt/V among patients in Hb levels with treatment (baseline, 6.3 to 6.6 with HD-CKD.319 CPR FOR PEDIATRICS 3.1: USING ESAs

ESAs are critical components in managing the 3.1.3.1 ADULT CPR anemia of patients with CKD. Available ESAs In the opinion of the Work are each effective in achieving and maintaining Group, the route of adminis- target Hb levels. Aspects of administration may tration should be determined differ between short-acting and long-acting by the CKD stage, treatment agents. setting, efficacy considerations, and the class of ESA 3.1.1 Frequency of Hb monitoring: (FULLY used. APPLICABLE TO CHILDREN) PEDIATRIC CPR 3.1.1.1 In the opinion of the Work In the opinion of the Work Group, the frequency of Hb Group, in the pediatric patient, monitoring in patients treated the route of administration with ESAs should be at least should be determined by the monthly. CKD stage, treatment setting, 3.1.2 ESA dosing: (FULLY APPLICABLE efficacy considerations, the TO CHILDREN) class of ESA used, and the 3.1.2.1 In the opinion of the Work anticipated frequency and pain Group, the initial ESA dose of administration. and ESA dose adjustments 3.1.3.2 In the opinion of the Work should be determined by the Group, convenience favors SC patient’s Hb level, the target administration in non–HD- Hb level, the observed rate of CKD patients. increase in the Hb level, and 3.1.3.3 In the opinion of the Work clinical circumstances. Group, convenience favors 3.1.2.2 In the opinion of the Work IV administration in patients Group, ESA doses should be with HD-CKD. decreased, but not necessarily 3.1.4 Frequency of administration: (APPLI- held, when a downward ad- CABLE TO CHILDREN, BUT NEEDS justment of Hb level is needed. MODIFICATION) 3.1.2.3 In the opinion of the Work 3.1.4.1 ADULT CPR Group, scheduled ESA doses In the opinion of the Work that have been missed should Group, frequency of adminis- be replaced at the earliest pos- tration should be determined sible opportunity. by the CKD stage, treatment 3.1.2.4 In the opinion of the Work setting, efficacy consider- Group, ESA administration in ations, and class of ESA. ESA-dependent patients PEDIATRIC CPR should continue during hospi- In the opinion of the Work talization. Group, in the pediatric patient, 3.1.2.5 In the opinion of the Work the frequency of administra- Group, hypertension, vascu- tion should be determined by lar access occlusion, inad- the CKD stage, treatment set- equate dialysis, history of ting, efficacy considerations, seizures, or compromised nu- and class of ESA; as well, tritional status are not contra- consideration should be given indications to ESA therapy. to the anticipated frequency of, 3.1.3 Route of administration: (APPLI- and pain on administration of CABLE TO CHILDREN, BUT NEEDS each agent and their potential MODIFICATION) effects on the child and family.

American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S93-S98 S93 S94 ANEMIA IN CHRONIC KIDNEY DISEASE IN CHILDREN

3.1.4.2 In the opinion of the Work Currently, the most robust evidence for dosing Group, convenience favors less ESA products in children is related to erythropoi- frequent administration, par- etin alfa and beta products, with information on ticularly in non–HD-CKD darbepoietin alfa dosing just now becoming avail- patients. able. All clinicians are advised to carefully evaluate RATIONALE the individual patient’s particular issues related to their Hb level and likely response before Frequency of Hb Monitoring deciding on a particular ESA product, dosing This guideline is considered applicable to chil- regimen, and frequency of monitoring before the dren because there are some data supporting this initiation of or changes in ESA and/or iron thera- assumption and there is no reason for a different pies. recommendation. The following section is divided into dialysis There are 2 reasons that may justify closer versus nondialysis patients and short-acting ver- monitoring of all pediatric patients with 1- to sus long-acting ESA products. 2-week Hb levels when initiating and/or making significant change to the ESA dose. The first is to Dialysis Population (HD and PD) ensure that the patient is responding to the cur- 1. Short-acting ESAs. Data from the latest rent dose. Although not as well described in NAPRTCS annual report highlight the variation children as in adult literature, it likely is true that in initial dose reported in children on dialysis a patient who will reach an intended 1-g/dL therapy, with the younger child consistently re- increase in Hb level after 1 month of a given ceiving more rHuEPO on a per-kilogram-per- ESA/iron regimen will have that increase occur week basis.325 The doses referred to next are not relatively evenly over each of the 4 weeks. In derived from RCTs, but rather from registry data other words, blood work weekly or every 2 and therefore should be interpreted with caution. weeks will allow the clinician to institute an They represent approximations of the dose re- increase in dosing if a rate of increase of approxi- quired in an average cohort of children and are mately 0.25 g/dL per week is not seen in the early not represented as recommendations for any spe- part of the month. (In adults, the rate of increase cific child. (Note that 93% of these patients were has been estimated to be 0.3 g/dL per week [0.2 administered erythropoietin alfa as Epogen¨;Am- to 0.5 g/dL] on appropriate doses of rHuEPO.324) gen, Thousand Oaks, CA.) Similarly, a child for whom the rate of in- This variation in dosing can be seen in relation crease appears rapid, eg, greater than 0.5 g/dL to: per week, could have the dose adjusted before a) The dialysis modality, with PD patients overshooting at the end of that month of therapy. initially requiring approximately 225 U/kg/wk The greater frequency of monitoring likely is compared with nearly 300 U/kg/wk of erythropoi- beneficial until the patient has reached a target etin alfa products to achieve target Hb levels for Hb level and is on a stable dose of ESA. At this those on HD therapy.325 Interestingly, this differ- point, less frequent monitoring may be indicated, ence in dosing disappears during the first 18 to eg, every 4 weeks. 30 months of follow-up. It should be noted that the initial dosing difference persisted even when ESA Dosing both groups of patients were administered the In the opinion of the Work Group, this guide- ESA SC326 and may be caused in part by such line is applicable in children, but needs some factors as blood loss and the use of IV ESAs in modification or adjustment. HD patients as opposed to SC ESAs in PD or nondialysis patients.319 Initial Dose and Dose Adjustments b) Age, with patients younger than 1 year Although there are many dosing guidelines requiring an average of 350 U/kg/wk; those 2 to for the use of ESAs in children, it is important 5 years, approximately 275 U/kg/wk; those 6 to to realize that, as in adults,123 there is a large 12 years, 250 U/kg/wk; and finally those older variation in pediatric dosing of these drugs.325 than 12 years needing only slightly more than ANEMIA IN CHRONIC KIDNEY DISEASE IN CHILDREN S95

200 U/kg/wk at the time of starting ESA poietin is more likely to be closer to 0.5 ␮g for therapy.325 One report speculates that this appar- every 200 U of erythropoietin alfa, rather than ently increased “clearance” of erythropoietin in a the 1 ␮g/200 U conversion recommended by the young child is caused by the presence of nonhe- manufacturer. The authors themselves suggest a matopoietic binding sites for the erythropoietin wide range of doses, 0.25 to 0.75 ␮g/kg/wk, as molecule,327 which—in the fetus or developing being reasonable for the initial switch between child—appears to function as a nonhematopoi- therapies. etic cellular growth factor.328 Because internaliza- The most recent report prospectively looked at tion and, potentially, degradation of some of the the use of darbepoetin alfa in 8 children on PD erythropoietin dose would limit the amount of therapy, 6 children on HD therapy, and 12 chil- hormone available for stimulating erythropoi- dren with ND-CKD.331 All patients weighed esis, one could explain the need for higher abso- more than 8 kg and were younger than 18 years. lute doses to achieve a given marrow response or Therapy with ESAs was initiated, if the patient Hb production. Furthermore, the authors specu- was ESA-naïve, when Hb level was less than late that as the child ages, the number of these 10.0 g/dL, and only patients judged iron-replete, sites decreases as development slows, with an TSAT greater than 20%, were included. The attendant decrease in the absolute hematopoietic 327 primary outcomes were: (1) the proportion of dose of erythropoietin required. patients with a mean Hb level greater than 10.0 No major differences were noted in relation to g/dL between weeks 8 and 12 and then weeks 20 either race or sex with respect to initial ESA dose to 28 of the study, and (2) the percentage of all required. There was a general tendency for all Hb values greater than 10.0 g/dL at each time doses to decrease over time, presumably on the point. basis of a lower amount of ESA being required The initial dose of darbepoetin alfa was 0.45 to maintain—as opposed to reach—a given Hb ␮g/kg/wk, administered IV in the HD patients target. and SC in the PD and ND-CKD patients. Much 2. Long-acting ESAs. Currently, there are only of the data are presented in aggregate and include 3 published articles329-331 and 1 abstract332 in the both the 26 patients enrolled prospectively and 7 peer-reviewed literature related to long-acting ESA products and children with CKD, including more patients enrolled retrospectively. Statistical those on dialysis therapy. It is expected that there analysis by the investigators did not show a difference in terms of drug dose between the will be other results in due course to assist in ϭ guiding the pediatric practitioner in using longer beginning and end of the study (P 0.77), and acting ESA agents more effectively. there was no difference in doses between the ϭ The first study evaluated the pharmacokinetic HD, PD, and ND-CKD groups (P 0.62) regard- properties of darbepoetin alfa in children with less of their being in the prospective or retrospec- ϭ ϭ CKD and compared the results with prior studies tive arm of the study (P 0.92 and P 0.73, published in adults.330 They showed that clear- respectively). ance, drug half-life, and bioavailability of darbe- Data from 23 patients in the prospective co- poetin alfa, whether administered IV or SC, were hort, 18 of whom were available at weeks 20-28 similar between pediatric and adult patients. The for analysis, showed a statistically significant 1 difference was that absorption of the drug increase in Hb levels at 8 to 12 weeks of 11.7 Ϯ administered SC appeared to be more rapid in 1.1 g/dL compared with the baseline of 10.5 Ϯ pediatric than adult patients. 1.0 g/dL (P ϭ 0.002), which was still present at The second study described the clinical use of week 20 to 28 when Hb level was now 11.4 Ϯ darbepoietin alfa in children.329 The investiga- 0.9 g/dL (P ϭ 0.01 compared with baseline). In tors switched 7 long-term HD patients aged 11.5 this group, 96% (CI, 0.90 to 1.0; P ϭ 0.03) years (range, 7 to 15.2 years) with a stable Hb compared with baseline at 8 to 12 weeks and level on erythropoietin therapy to darbepoietin 94% (CI, 0.72 to 0.99; P ϭ 0.06) compared with IV therapy. Results from this small study suggest baseline at 20 to 28 weeks had mean Hb values that (at least in the pediatric HD population) the greater than 10.0 g/dL. Similarly, proportions of conversion dose of erythropoietin alfa to darbe- patients with total Hb values greater than 10.0 S96 ANEMIA IN CHRONIC KIDNEY DISEASE IN CHILDREN g/dL were 94% at 8 to 12 weeks and 90% at 20 to A trial reported on behalf of the Australian and 28 weeks. New Zealand Paediatric Nephrology Association In all ESA-naïve patients (n ϭ 8), the average also looked at the use of SC ESA in predialysis time to reach the target of 10.0 g/dL from base- children.316 Unfortunately, although 10 of the 22 line of 9.0 Ϯ 1.4 g/dL was 3.4 weeks, with no patients were predialysis, their study does not patient requiring a higher dose of darbepoetin allow the data to be abstracted specifically for alfa to achieve this. this group. The overall Hb response to darbepoetin alfa 2. Long-acting ESAs. As discussed in more was related to the age of the patient, higher detail in the preceding section (Long-Acting values for a given dose being achieved in older ESAs in the Dialysis Population), a study showed patients (P ϭ 0.02). similar efficacy of darbepoetin alfa in the 15 Only 1 patient experienced a serious AE that patients, 12 prospectively enrolled, with ND- was thought to be possibly related to the darbepo- CKD treated with a dose of 0.45 ␮g/kg/wk.331 etin alfa. In this case, a girl on home HD therapy was admitted with a worsening of her hyperten- Rate of Increase in Hb Levels sion; this was at a time when her Hb level was A few pediatric studies provide some informa- 13.2 g/dL but concurrent with a withdrawal from tion on which to base recommendations on the clonidine some 18 days before the admission. safety and side effects related to the rate of Her blood pressure subsequently was well con- increase in Hb levels. trolled on fewer medications while remaining on A prospective trial that examined the safety of darbepoetin alfa therapy. rHuEPO therapy in children on HD or PD therapy or predialysis, randomized 44 children (aged 4 Nondialysis Population months to 21 years) to either low (150 U/kg/wk) 1. Short-acting ESAs. In the ND-CKD popula- or high (450 U/kg/wk) doses of epoetin alfa tion, a number of studies looked at the dose of administered in 3 divided doses.319 Patients were rHuEPO required to achieve set targets. Cur- followed up for a total of 12 weeks and targeted rently, there is only 1 published prospective for individual Hb levels of between 2 SD less randomized trial that has looked at rHuEPO than and the mean value for that child’s age. dosing in children with ND-CKD.319 In this trial, Taking the groups as a whole, there was an 25 predialysis patients, CKD stages 3 to 5 but not average increase of 4.2 Ϯ 2.1 versus 2.4 Ϯ 1.5 on PD or HD therapy, were randomized to either g/dL per month in the high-dose versus low-dose 150 U/kg/wk (n ϭ 12) or 450 U/kg/wk (n ϭ 13) groups. Hypertension appeared to be more com- of rHuEPO administered SC in 3 divided doses. mon, but did not reach statistical significance, for In all 12 of the 150-U/kg/wk patients and 11 of those in the high-dose group compared with the 13 of the 450-U/kg/wk patients, the Hb target of low-dose group, 8 of 21 versus 5 of 23 patients, less than 2 SDs less than, but less than the mean respectively; chi-squared P ϭ 0.17. The investi- for age, was achieved. When at target, the dose gators noted that the trend in systolic and dia- was reduced to maintain the target Hb level; on stolic blood pressure was to increase as Hb level average, this required 143 Ϯ 102 U/kg/wk for increased. No other side effects seemed related the group of responders. to the rate of increase in Hb levels. A nonrandomized open-labeled prospective Another study randomized 20 anemic children study took 11 patients aged 0.6 to 17 years with 5 to 16 years of age who were on a stable the equivalent of CKD stages 4 and 5 (all predi- continuous ambulatory PD regimen for 3 months alysis) whose mean Hb level was 7.9 g/dL and to examine the effects of low-dose rHuEPO, 50 treated them with a single dose of 150 U/kg/wk U/kg/wk (group A), versus high-dose rHuEPO, of SC rHuEPO.333 An increase in Hb level greater defined as 50 U/kg 3 times a week (group B).334 than 2 g/dL was seen in all 11 patients in a mean Translating their data to Hb values from Hct, of 45 days, range of 14 to 119 days, and subse- those in group A showed a steady increase over 6 quently, the patients maintained Hb levels be- months from an Hb level of approximately 6.3 tween 11.5 and 13.5 g/dL, with a mean dose of g/dL to 9.9 g/dL compared with the increase seen 133 U/kg/wk, range of 75 to 300 U/kg/wk. in group B, for which Hb level went from 6.4 to ANEMIA IN CHRONIC KIDNEY DISEASE IN CHILDREN S97

10.7 g/dL in only 3 months; an approximately pared with IV, as confirmed in an observational 0.6-g/dL increase per month in group A versus trial in children.325 1.4-g/dL increase per month in group B. Al- As of the 2004 NAPRTCS annual report, 96% though mean arterial blood pressure increased in of children on PD therapy were administered an both groups during the study (group A, 83 to 87 ESA by the SC route as opposed to only 14% in mm Hg; group B, 85 to 101 mm Hg), it only the HD population. reached statistical significance in group B, P Ͻ However, in children, the psychological im- 0.05. This was borne out by the need to increase pact of frequent and/or painful injections also is the baseline antihypertensive medications in 8 of important to assess when deciding on a dosing the 10 group B patients and initiate these medica- route. Currently, the single-dose preloaded sy- ¨ tions in the remaining 2 patients in this group. ringes of both epoetin alfa (Eprex ) and darbepoi- (Note: rHuEPO therapy was discontinued briefly etin alfa available in many parts of the world do in 4 patients in group B; 2 patients because of not contain benzyl alcohol, which acts as a local hypertensive encephalopathy; conversely, no pa- anesthetic, in the epoetin alfa multidose vials. tient in group A required an increase in antihyper- This means that injections with the preloaded ¨ tensive medication or initiation of such therapy Eprex syringes generally are more painful than during the study.) those from the multidose vials. Similarly, in a If one examines a number of recent pediatric Canadian study, 8 of 14 patients with prior expe- ¨ recommendations for the acceptable rate of in- rience using epoetin alfa (Eprex ) reported that crease in Hb values, they vary widely. One group darbepoetin alfa caused more pain on injection; recommends a rate of increase between 1 and 2 the remaining 6 patients did not specifically comment about whether they believed Eprex¨ g/dL per month for all children below the target 331 range.301 Recent guidelines from the European caused more pain on injection. Paediatric Peritoneal Dialysis Working Group Note: Multidose vials of epoetin alfa should recommend an increase of approximately 0.66 be avoided if at all possible in premature infants g/dL per month as minimally acceptable335 and and newborns because of a rare, but well-recog- an increase of more than approximately 2.5 g/dL nized, complication from the use of benzyl alco- per month as unacceptable. An often-quoted study hol in the preparation of the compound. This excipient has been described to cause numerous recommends, without evidence, that the goal serious and potentially fatal reactions, including should be an increase in Hb of 1 g/dL per metabolic acidosis, intraventricular hemorrhage, month.336 and neurological problems. Sixteen neonatal deaths ESA dosing should be decreased, not held, if were reported that were thought to be caused by Hb level is elevated. In the opinion of the Work benzyl alcohol toxicity, generally described as Group, this guideline is fully applicable to chil- the so-called “gasping syndrome.”337,338 dren. The issue of dosing epoetin alfa through the Missed ESA doses. In the opinion of the Work intraperitoneal route in children to eliminate in- Group, this guideline is fully applicable to chil- jection pain also has been studied by various dren. ESA dosing during hospitalizations. In the investigators.339,340 In general, the added costs opinion of the Work Group, this guideline is fully because of a higher dose required to achieve the applicable to children. same target Hb level, the need for at least 1 dry Contraindications to ESA therapy. In the opin- day dwell (or more) per week to achieve the ion of the Work Group, this guideline is fully epoetin alfa absorption required, and the poten- applicable to children. tial for more frequent episodes of peritonitis have lead most centers away from use of this Route of Administration modality routinely. As of the 2004 NAPRTCS As in adults, convenience in an outpatient report, less than 2% of children on PD therapy setting favors use of the SC route for delivery of currently are administered epoetin alfa through ESAs; with the added realization that even in the the peritoneal cavity.325 At present, there are no face of IV access, a short-acting ESA product data available on the use of darbepoietin alfa will be more efficacious administered SC com- intraperitoneally. S98 ANEMIA IN CHRONIC KIDNEY DISEASE IN CHILDREN

Frequency of Administration In 1 study, to appropriately dose patients at As in the adult population, there is a move 0.45 ␮g/kg/wk with the available preloaded dar- toward extending the dosing interval of all ESA bepoetin alfa syringes, the investigators started products to minimize injections while maintain- 63% of their patients on every 10-, 14-, or 21-day ing efficacy of the product; ie, not having to dosing regimens; the remainder were on once- 331 increase dose significantly more than the gain in weekly dosing. Their results showed no im- time between dosing, to maintain or achieve the pact of the various dose intervals on the Hb value target Hb level. over time (P ϭ 0.01), but showed that both HD 1. Short-acting ESAs. At present, there is no and older patients required more frequent dosing RCT evidence to show that this strategy is appro- (P Ͻ 0.0001 and P ϭ 0.01, respectively), al- priate in children with the current short-acting though the dose amount was not different in any ESA products on the market. However, data from group based on age or mode of renal replacement the 2004 NAPRTCS annual report highlight that therapy. 75% to 80% of children on HD therapy will As outlined in CPR 3.1.3.1, the psychological receive short-acting ESA products IV during a impact of frequent and painful injections also is standard (3 times per week) HD regimen, whereas important to assess when deciding on a dosing a smaller number (15% to 20%) receive it twice a frequency in children. Use of less-frequent dos- week, and other dosing regimens are uncommon. ing strategies is, in theory, of benefit by reducing However, 70% of children on PD therapy gener- both the absolute number of injections and the ally receive only once-weekly or perhaps twice- pain associated with injections for the child on weekly injections, with 25% receiving injections ESA treatment. However, this may be mitigated thrice weekly, and approximately 5% receiving by the fact that currently, the only long-acting the drug less than once per week.325 ESA on the market produces a noticeably in- It also may be reasonable on occasion to creased degree of pain at time of injection for consider more frequent SC dosing of the short- many children compared with some of the short- acting ESAs, eg, thrice weekly, when the pa- acting ESA therapies. tient’s Hb level is well less than the target Hb At present, there are no good data from the desired, although the absolute rate of increase pediatric ND-CKD population with regard to the should not exceed 2 g/dL per month. Here, the dosing frequency of various ESAs. more-frequent dosing ensures that the therapeu- Finally, the Work Group also was aware of tic threshold of the hormone is maintained, given results from a soon-to-be-published open-label its short half-life, and—in concert with adequate multicenter noninferiority trial examining the iron stores—may assist in achieving a more use of darbepoetin alfa versus epoetin alfa in the rapid increase in Hb level. treatment of anemia in children with CKD stages 2. Long-acting ESAs. Data on the dosing fre- 4 to 5, including patients on HD and PD therapy quency of darbepoietin alfa in pediatric HD and the nondialysis CKD population (Bradley patients comes from 7 patients, all of whom were Warady, personal communication, June 7, 2005). dosed once weekly after HD.329 From a pharma- Because the results were not yet peer reviewed or cokinetic study,330 it is clear that extended- published at the time the guidelines were final- dosing regimens with darbepoietin alfa in chil- ized and were not blieved to alter the guideline dren should be possible, although currently, few statements, this study is not incorporated in this data have been published outlining strategies for set of guidelines, but will provide data related to dosing frequency. use of darbepoetin alfa in children. CPR FOR PEDIATRICS 3.2: USING IRON AGENTS

Anemia therapy in patients with CKD requires ● Serum ferritin > 100 ng/mL effective use of iron agents, guided by appropri- AND ate testing of iron status. Efficacy of iron therapy ● TSAT > 20%. appears not to be limited to patients with evi- 3.2.4 Upper level of ferritin: (FULLY AP- dence of iron deficiency. (See Guideline 1.2 for PLICABLE TO CHILDREN) diagnosis of iron deficiency.) Thus, the goals of In the opinion of the Work Group, iron therapy are to avoid storage iron depletion, there is insufficient evidence to recom- prevent iron-deficient erythropoiesis, and achieve mend routine administration of IV and maintain target Hb levels. iron if serum ferritin is greater than 3.2.1 Frequency of iron status tests: 500 ng/mL. When ferritin level is (FULLY APPLICABLE TO CHIL- greater than 500 ng/mL, decisions DREN) regarding IV iron administration In the opinion of the Work Group, should weigh ESA responsiveness, Hb iron status tests should be performed and TSAT level, and the patient’s as follows: clinical status. 3.2.1.1 Every month during initial 3.2.5 Route of administration: (FULLY AP- ESA treatment. PLICABLE TO CHILDREN) 3.2.1.2 At least every 3 months dur- 3.2.5.1 The preferred route of admin- ing stable ESA treatment or in istration is IV in patients with patients with HD-CKD not HD-CKD. (STRONG RECOM- treated with an ESA. MENDATION) 3.2.5.2 In the opinion of the Work 3.2.2 Interpretation of iron status tests: Group, the route of iron ad- (FULLY APPLICABLE TO CHIL- ministration can be either IV DREN) or oral in patients with ND- In the opinion of the Work Group, CKD and PD-CKD. results of iron status tests, Hb level, and ESA dose should be interpreted 3.2.6 Hypersensitivity reactions: (FULLY together to guide iron therapy. APPLICABLE TO CHILDREN) 3.2.3 Targets of iron therapy: (APPLI- In the opinion of the Work Group, CABLE TO CHILDREN, BUT NEEDS resuscitative medication and personnel trained to evaluate and resuscitate MODIFICATION) anaphylaxis should be available when- In the opinion of the Work Group, ever a dose of iron dextran is sufficient iron should be administered administered. to generally maintain the following indices of iron status during ESA BACKGROUND treatment: As in adults, the most commonly identified 3.2.3.1 ADULT CPR HD-CKD: reason for poor responsiveness to ESA therapy in ● Serum ferritin > 200 ng/mL, children is iron deficiency.341 AND The need for iron supplementation in children ● TSAT > 20%, or CHr > 29 to maintain iron stores and promote the efficient pg/cell. use of ESAs can be predicted from the knowl- PEDIATRIC CPR edge of mean daily blood—hence iron—losses HD-CKD: in both those who are predialysis and those on ● dialysis therapy. Predialysis patients, and likely Serum ferritin > 100 ng/mL; 2 AND those on PD therapy, lose on average 6 mL/m /d of blood, mainly from the GI tract, which trans- ● TSAT > 20%. lates to a cumulative yearly iron loss and there- 3.2.3.2 ND-CKD and PD-CKD: fore requirement of 0.9 g/1.73 m2.342 This is just

American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S99-S104 S99 S100 ANEMIA IN CHRONIC KIDNEY DISEASE IN CHILDREN more than half of that predicted for pediatric HD required, and there are no adult or pediatric data patients, for whom the mean GI blood loss is 11 on the safety of supraphysiological ferritin lev- mL/m2/d, and when coupled with the dialyzer els. losses of 8 mL/m2 per treatment, translates to a One pediatric study in older children on HD yearly cumulative loss of approximately 1.6 therapy suggested that a TSAT less than 20%, but g/1.73 m2.342 not a ferritin level less than 100 ng/mL, was As in the adult population, several studies predictive of iron deficiency.349 However, other have shown that supplementation of iron in chil- pediatric studies in the CKD population have dren receiving ESA therapy allows a reduction in shown that although a normal ferritin level can- the ESA dose required per unit of Hb level not exclude iron deficiency, absolute or func- 343,344 achieved. However, constipation and nau- tional,348,350 a low ferritin level, less than 60 sea, as well as poor GI iron absorption, often ng/mL, is a specific predictor of its presence.351 limit effective supplementation with oral iron Therefore, we recommend that the current 166,345,346 preparations. These facts and the avail- Guidelines be followed in children with the ex- ability of newer IV iron preparations believed ception that until data are available on the risks less likely to induce AEs recently led to more and benefits of higher ferritin targets in children, studies with IV preparations of iron in children. the targets be left at a ferritin level greater than The majority of these studies examined only 100 ng/mL and TSAT greater than 20% for patients in CKD stage 5, especially those on HD 304,305,344,347,348 patients with HD-CKD, as well as PD-CKD and therapy. ND-CKD populations. RATIONALE With respect to the use of CHr, which has been touted as a valuable screening test for iron defi- Frequency of Iron Tests ciency in “normal” children even before the This guideline is considered applicable to chil- appearance of iron-deficient anemia, the cutoff dren because there are no special data in the normal and abnormal values remain unclear and pediatric population and there is no reason for a wide. For example, a cutoff value of 26 pg only different recommendation. produced sensitivity and specificity of 70% and 78% for iron deficiency and 83% and 75% for Interpretation of Iron Tests iron deficient anemia in this select group of This guideline is considered applicable to chil- non-CKD young children, respectively.352 dren because there are no special data in the In the pediatric CKD literature, CHr has been pediatric population and there is no reason for a examined in a limited fashion in children on HD different recommendation. therapy.304,305 In both studies, an increase from baseline CHr levels was observed in response to Targets of Iron Therapy oral iron, IV iron dextran, and IV sodium ferric HD-CKD gluconate in patients judged to be both iron 304 Caution must be exercised in applying the replete (oral iron and IV iron dextran) and 305 targets or normal ranges as outlined in the guide- iron-deficient (IV sodium ferric gluconate). line for adult patients with HD-CKD because the Unfortunately, cutoff values for use of this marker increase in ferritin target from 100 to 200 ng/mL in the pediatric CKD population are not clear, in adult HD patients is based on evidence to and this is shown clearly in these 2 studies, in suggest that a level of 100 ng/mL underestimates which, in nearly identical patient populations (ie, iron deficiency,98,99,151Ð156 and data from a study children on HD therapy), the baseline value for in which patients with an average serum ferritin CHr was higher in the group judged as iron level of 730 ng/mL showed a 40% reduction in deficient based on current DOQI guidelines305 ESA dose compared with those with an average than in those judged iron-replete by those same ferritin level of 297 ng/mL.108 There is no simi- criteria.304 lar RCT-level evidence for this in children; Therefore, whereas it is clear that CHr may yet namely, that achieving a specific ferritin level have a prognostic or diagnostic role in anemia in will produce a consistent decrease in ESA doses children with CKD, further research clearly is ANEMIA IN CHRONIC KIDNEY DISEASE IN CHILDREN S101 needed to understand and define the value of this Efficacy of IV Iron test in the pediatric CKD population. To date, the largest randomized prospective study to address this issue is a multicenter, inter- ND-CKD and PD-CKD national, double-blind, parallel-group, efficacy This guideline is considered applicable to chil- and safety study that examined the use of sodium dren because there are no special data in the ferric gluconate in anemic children with a mean pediatric population and there is no reason for a age of 12.1 Ϯ 2.6 years on HD therapy.305 All different recommendation. patients at the time of enrollment met 1 or both of the current KDOQI targets for iron deficiency Upper Level of Ferritin while on ESA therapy, ie, a TSAT less than 20% This guideline is considered applicable to chil- or ferritin level less than 100 ng/mL. Of this dren because there are no special data in the group, data from 56 patients were available to pediatric population and there is no reason for a assess the efficacy of the 2 dosing regimens different recommendation. chosen: 1.5 versus 3.0 mg/kg administered for 8 Note: As in the adult guidelines, this refers sequential HD sessions. only to the intentional targeting of a patient’s All patients in both dosing arms showed a ferritin level to greater than 500 ng/mL, not to significant increase in their baseline Hb levels of the individual’s achieved or acquired ferritin 0.9 g/dL at 2 weeks; 0.9 g/dL in the 1.5-mg/kg level, which may be at or greater than this level. dose and 1.0 g/dL in the 3.0-mg/kg dose arm at 4 weeks from the last dose of IV iron, all P Ͻ 0.02. Route of Administration No patient had any change in rHuEPO dose This guideline is considered applicable to chil- during the iron therapy; however, during the dren because there are no special data in the 4-week follow-up period, 5 patients had reduc- pediatric population and there is no reason for a tions in rHuEPO dose of 14%, 20%, 50%, 50%, different recommendation. and 67%, all at the 2-week point. Evaluation of Suggestions for iron supplementation in chil- all other indices studied (TSAT, ferritin level, dren with CKD vary in terms of dose, as well as and CHr) also showed a significant mean in- type of preparation. Many children in the ND- crease from baseline values in both arms, which CKD or PD-CKD populations may benefit from remained significant at 4 weeks after infusion. and receive oral iron therapy,325 whereas those However, only ferritin levels showed a signifi- on HD therapy are likely to require and most cant change between the low-dose and high-dose often receive IV agents to allow for sufficient iron arms, seen at both the 2-week and 4-week iron stores for ongoing erythropoiesis.325,348,350 time points following the last dose of IV iron. Based on their data, the investigators concluded Oral Iron that IV sodium ferric gluconate was efficacious For oral iron therapy, the recommendations in both repleting and maintaining iron stores in are for doses of elemental iron ranging from 2 to children on HD up to 4 weeks after infusion. 3 mg/kg/d up to 6 mg/kg/d, with a maximum of In a separate and earlier randomized prospec- 150 to 300 mg of elemental iron per day in 2 to 3 tive study, the investigators studied 35 pediatric divided doses335,336 taken 2 hours before or 1 patients between 1 and 20 years of age on HD hour after all calcium-containing binders and therapy for their response to oral or IV iron.304 food to maximize GI absorption.353 A prospec- All patients were iron replete, as defined by tive randomized trial of 35 pediatric patients KDOQI, with a TSAT greater than 20% and between 1 and 20 years of age on HD therapy, all ferritin level greater than 100 ng/mL. They were iron replete, examined response to oral or IV iron all well dialyzed and randomized to either weekly (more details discussed next).304 This study IV iron dextran, which was dosed by weight and clearly showed a much better response to IV iron provided as 2 sets of 6 weekly doses with two in terms of an increased serum ferritin level, but 2-week breaks to monitor levels, or daily oral did not show any advantage of IV over oral iron iron at a dose of 6 mg/kg/d as outlined in the in terms of maintaining iron stores. NKF-KDOQI 2000 Anemia Guidelines for 16 S102 ANEMIA IN CHRONIC KIDNEY DISEASE IN CHILDREN weeks.2 Data showed that only the IV iron dex- (95% CI, 3.3 to 8.8) in the study group compared tran produced a significant increase in serum with 14.4 mg/kg/mo (95% CI, 12 to 16.8) in the ferritin levels, P ϭ 0.001. Those treated with IV control arm, P Ͻ 0.001. The investigators showed iron dextran also showed a significant decrease that patients in the study arm achieved and main- in dose of rHuEPO required to maintain target tained a stable Hb value, whereas those treated as Hb levels, P ϭ 0.046, and an increase in CHr, controls had a much more variable increase and although this was not statistically different from decrease in Hb values. those seen in the oral iron group. Both the oral Four other nonrandomized trials in chil- and IV preparations maintained all patients in the dren,344,348,350,354 all involving HD patients and iron-replete state, as when the study was started. 1 including patients with ND-CKD and/or trans- Although there was a significant difference in plant recipients,354 examined the utility of IV mean ferritin levels (259.1 compared with 138.5 iron in maintaining or increasing Hb levels and ng/mL; P ϭ 0.003) and rHuEPO doses (reduc- decreasing the dose of ESA required to do so. tion of 33% from baseline; P ϭ 0.046) in the IV Two trials used iron gluconate,350,354 1 trial used iron group during the study, there was no statisti- iron dextran,348 and 1 trial used iron sucrose344; cally significant difference between the IV ver- doses ranged from 1 to 4 mg/kg/wk of the sus oral iron arms as a whole. The investigators various products as maintenance therapy, and concluded from this short-term study that IV iron time of therapy varied from 2 to 24 weeks. All 4 dextran seemed more effective in improving— trials showed increases in either Hb level or Hct but no more effective in maintaining—iron stores and a decrease in ESA requirements between 5% in pediatric HD patients on ESA therapy com- and 62% per week or per dose of ESA. pared with oral iron. A recent meta-analysis on the use of IV iron in Another prospective randomized trial exam- pediatric HD patients used pooled data that did ined the issue of intermittent versus maintenance not include the 2 most recent trials,304,347 but IV iron in pediatric patients on HD therapy.347 included the other 4 studies described, as well as The study group of 20 patients received IV iron a number of abstracts. Meta-analysis showed dextran targeted to initial ferritin levels and a that in terms of an increase in Hb, Hct, ferritin, calculation of the net projected iron stores re- and TSAT values and decrease in ESA require- quired to target an Hb level of 11.55 g/dL, ments, there was a positive correlation with IV whereas the 20 patients in the control group were iron therapy with an effect size that varied from treated with intermittent courses of 10 weekly 0.62 (95% CI, 0.11 to 1.13) to 1.86 (95% CI, doses of IV iron dextran as defined by body 1.58 to 2.15) when evaluated using a standard- weight (repeated as necessary based on the pres- ized weighted-mean difference approach.343 ence of a ferritin level Ͻ 100 ng/mL, TSAT Ͻ 20%, or Hct Ͻ 33%). All patients had basal Hb Dosing of IV Iron levels of approximately 8 g/dL and reached 10 The use of IV iron preparations and the appro- g/dL by 3 months in both arms. The study’s end priate dosage of each is a complex topic. It is points were ferritin, TSAT, and Hb levels. Suc- made more so because one needs an approach to cess was defined as maintaining an acceptable both the immediate repletion of iron stores in a target range of ferritin between 100 and 800 patient who is deficient and a strategy for main- ng/mL and TSAT of 20% to 50%. The study taining an effective level of iron for ongoing enrolled both absolute and functionally iron- erythropoiesis. deficient patients. A large number of patients, 20, The goal of the initial iron therapy is to replenish were excluded during this study; 13 patients the body store of functional iron and thus assist in because of iron overload, defined as a ferritin the production of red blood cells and Hb in concert level greater than 800 ng/mL (9 from the control with an ESA. The exact dosing regimens, fre- group). Three patients in the control group re- quency of IV iron therapy, and appropriate monitor- quired a blood transfusion and also were ex- ing for effectiveness and safety will be related to cluded. Using the study’s iron and rHuEPO pro- the iron preparation chosen.2,305,344,355 tocols, there was a significant difference in iron Currently published pediatric studies looking dose required during the study, 6 mg/kg/mo at the issue of chronic or maintenance therapy ANEMIA IN CHRONIC KIDNEY DISEASE IN CHILDREN S103 with IV iron have provided 1 to 2 mg/kg/wk of (labile or free iron reactions) are both possible elemental iron and targeted TSATs between 20% mechanisms, and the underlying cause may dif- and 50% and serum ferritin levels of 100 to 800 fer depending on the type of IV iron. Anaphylac- ng/mL, based on the prior Anemia Guidelines,2 toid reactions appear to occur more frequently to decide on further doses and frequency of with iron dextran,173 and labile or free iron administration.344,350 reactions, more frequently with nondextran forms It is important to remember that it is possible of iron.174 (For further discussion, please see the to have acceptable levels of both TSAT and corresponding Adult Guidelines.) ferritin and still benefit from IV iron if the patient Although very rare in pediatrics, of 28 chil- has so-called functional iron deficiency; there- dren enrolled in 2 separate studies involving iron fore, occasionally, after careful assessment of the dextran,304,348 1 child had an allergic reaction risk and benefits, a “trial” of IV iron in an anemic that necessitated stopping the medication.348 Cur- patient—even one who appears iron replete— rent evidence would suggest that the risk for may be indicated. life-threatening reactions is greater with IV dextran products than sodium ferric gluconate175 Use of IV Iron in Patients with ND-CKD and 187 PD-CKD and iron sucrose products. Although both iron sucrose and gluconate prod- The issue around the utility and practicality of ucts seem to have better safety profiles than using IV agents in the non-HD population needs dextran, side effects—presumably caused by to be addressed in children because it is not acute iron toxicity during rapid free iron release— uncommon that these patients either show an also have been described with both of these inability to tolerate or fail to respond to oral products. replacement of iron stores. With respect to sodium ferric gluconate, pedi- Currently, in pediatrics, the lack of easy IV atric safety data from the trial described ear- access hampers the use of the strategies used in lier305 were available in 66 patients administered the HD population, namely, small, but frequent, 8 IV doses. One patient in the group adminis- dosing. In the non-HD population, much higher tered 1.5 mg/kg per dose was reported to have single doses of the various IV irons, such as 356Ð358 359,360 361 isolated episodes of mild nausea, diarrhea, and dextran, sucrose, or gluconate, have been administered at less frequent intervals vomiting, whereas a patient in the 3.0-mg/kg (eg, monthly) in adults to obtain the benefit of IV group had an episode of severe anemia ascribed therapy while minimizing the inconvenience of to the drug by the investigator at that site. How- both the need for IV starts and hospital monitor- ever, no patient had an allergic or anaphylactic ing during the therapy. However, this may carry reaction during the immediate treatment; no de- different risks in terms of either acute or chronic layed reactions occurred in the 4 weeks after the toxicities. Currently there is little comparable last iron dose, and no deaths were reported evidence in children, with only 1 published study during the study. Together with data from 2 other reporting a maximum delivered dose of iron trials in which 21 children were administered sucrose of 200 mg344 and another reporting a this product without serious AEs, this observa- maximum dose of 250 mg of iron gluconate.354 tion offers some proof about the safety of sodium ferric gluconate in children.350,354 Safety of IV Iron Iron sucrose safety data are sparse in the One concern in the use of IV iron in children, pediatric CKD literature, with only 2 studies especially in an outpatient setting, is the poten- reported. One was a retrospective study that did tially fatal acute AEs.179 All forms of IV iron not report serious AEs in the 8 patients who had may be associated with acute AEs, which may received at least 1 dose of IV iron sucrose as include hypotension, anaphylactoid reactions, and Venofer (American Regent, Shirley, NY).344 An- a variety of other symptoms. Immune mecha- other was a prospective trial that looked at only a nisms with activation of mast cells or release of small number of patients (n ϭ 14) divided into 3 bioactive partially unbound iron into the circula- different treatment groups with various iron regi- tion resulting in oxidative stress and hypotension mens and did not specifically report on AEs.362 S104 ANEMIA IN CHRONIC KIDNEY DISEASE IN CHILDREN

The issue of iron overload as a “side effect” of develop iron overload than those with an abso- IV iron therapy also should be addressed. Iron lute iron deficiency, 70% versus 19%; P Ͻ excess is believed to generate cellular oxidative 0.005. Note: There was no direct or indirect stress, and iron stored as ferritin can assist in evidence of organ damage offered by the inves- initiating lipid peroxidation of cell membranes.363 tigators in support of their definition of iron In humans, increased ferritin levels have been overload. linked to a variety of conditions, including in- Finally, the Work Group also was aware of creased severity of strokes364 and acute renal results from a soon-to-be-published prospec- failure,365 although direct evidence for tissue tive multicenter trial of sodium ferric glu- toxicity related to serum ferritin level is lacking conate complex for maintenance iron therapy in the CKD population. in iron-replete children on HD therapy and In children, a study comparing 2 different administered rHuEPO (Bradley Warady, per- dosing strategies of iron dextran showed, by sonal communication, June 7, 2005). Because using Kaplan-Meier analysis, a hazard of iron results were not yet peer reviewed or published overload (defined by the investigators as a serum at the time these guidelines were finalized and ferritin level Ͼ 800 ng/mL or TSAT Ͼ 50% at were not believed to alter the guideline state- any time during the trial) of 20% in the study arm ments, the study is not incorporated in this set versus 100% in the control arm after 6 months of of guidelines, but will provide data related to treatment.347 Patients with functional iron defi- dosing, safety, and efficacy of rHuEPO in the ciency also were statistically more likely to pediatric HD-CKD population. CPR FOR PEDIATRICS 3.3: USING PHARMACOLOGICAL AND NONPHARMACOLOGICAL ADJUVANTS TO ESA TREATMENT IN HD-CKD

Several pharmacological agents and nonphar- Pediatric patients on HD therapy may have low macological manipulations of the HD prescrip- plasma carnitine levels366-368; this is less clear in tion have been examined for potential efficacy as pediatric PD patients, for whom the data are adjuvants to ESA treatment. Studies are not avail- more conflicting.369,370 However, any justifica- able to address the use of pharmacological or tion for use of L-carnitine in children is much nonpharmacological adjuvants to ESA treatment sparser than in the adult literature. in patients with ND-CKD and PD-CKD. One extremely small trial in which 2 children on HD therapy were administered IV L-carnitine 3.3.1 L-carnitine: (FULLY APPLICABLE showed an increase in their Hct by 34% with no TO CHILDREN) change in erythropoietin dose delivered.371 In the opinion of the Work Group, A relatively small study of 16 pediatric dialy- there is insufficient evidence to recom- sis patients, 11 on HD therapy, administered oral mend the use of L-carnitine in the carnitine divided twice daily at a dose of 20 mg/ management of anemia in patients kg/d for 26 weeks and did not show a benefit with CKD. with respect to increased Hb or Hct values or 3.3.2 Vitamin C: (FULLY APPLICABLE decreased dose requirement for erythropoi- TO CHILDREN) etin.372 There also is indirect evidence that be- In the opinion of the Work Group, cause the study was not designed to look at Hb there is insufficient evidence to recom- or Hct levels, achieving much higher serum mend the use of vitamin C (ascorbate) L-carnitine levels using a 5 times greater oral in the management of anemia in pa- dose in children on continuous ambulatory PD tients with CKD. therapy would not change the Hb levels.370 3.3.3 Androgens: (FULLY APPLICABLE Finally, a word of caution is warranted: there TO CHILDREN) also has been concern expressed over the use of Androgens should not be used as an oral carnitine because it may produce toxic me- adjuvant to ESA treatment in anemic tabolites.202 patients with CKD. (STRONG RECOMMENDATION) Vitamin C This guideline is considered applicable to chil- RATIONALE dren because there are no special data in the pediatric population and there is no reason for a L-carnitine different recommendation. This guideline is considered applicable to children because there are no special data in the Androgens pediatric population and there is no reason for a This guideline is considered applicable to chil- different recommendation. dren because there are no special data in the There are only a few small studies published pediatric population and there is no reason for a addressing the topic of L-carnitine in children. different recommendation.

American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: p S105 S105 CPR FOR PEDIATRICS 3.4: TRANSFUSION THERAPY

Red blood cell transfusions should be used older than the neonate who do not have concur- judiciously in patients with CKD, especially be- rent hematologic symptoms or CVD.271,373,374 cause of the potential development of sensitivity, It also should be remembered that red blood affecting future kidney transplantation. How- cell transfusions, at least in the critically ill ever, despite the use of ESA and iron therapy, pediatric patient, may not be “benign” therapy. transfusion with red blood cells occasionally is This was highlighted by a retrospective cohort required, in particular in the setting of acute study of 240 children in 5 pediatric intensive care bleeding. units, 130 of whom received red blood cell 3.4.1 (FULLY APPLICABLE TO CHIL- transfusions. The study showed that even after DREN) In the opinion of the Work controlling for a number of factors, such transfu- Group, no single Hb concentration sions were associated with increased use of oxy- justifies or requires transfusion. In gen, days of mechanical ventilation, vasoactive particular, the target Hb recom- agent infusions, length of intensive care unit stay, and total length of hospital stay (all P Ͻ mended for chronic anemia manage- 373 ment (see Guideline 2.1) should not 0.05). serve as a transfusion trigger. Finally, with respect to transfusions in pediatric patients awaiting renal transplantation, it is RATIONALE possible that the need for or use of packed red blood cell transfusions may benefit the pediatric There is no good evidence to support any patient if chosen carefully and provided under different recommendations in children compared immunosuppressive coverage in a planned man- with adults with respect to this guideline because ner. The most current and largest study involved the physiological principles underlying this guide- 193 children, 91 of whom received 2 sequential line hold true in children with anemia and CKD. blood transfusions, either random or donor spe- The pediatric practitioner also may wish to con- cific if a living related donation was expected, sider the following evidence. done 1 month apart under cover of oral cyclospor- Currently, there are no specific guidelines re- ine.375 Significant improvements in both 1-year lated to transfusion of red blood cells in children and 5-year graft survival were reported: 96% with any level of CKD, on or off dialysis therapy. versus 78% and 90% versus 64% (P Ͻ 0.001). However, it should be recognized that any of Whereas not minimizing the risk for transmis- these children, as with an adult, will benefit most sion of infectious agents from blood transfu- from red blood cell transfusions only in the face sions, especially if that transfusion is unneces- of impaired oxygen delivery, ie, not based on an sary, data suggest further prospective trials of red Hb “number.” Poor oxygenation is a universally blood cell transfusion before transplantation followed indication for transfusions in children might be of value in children.

S106 American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: p S106 CPR FOR PEDIATRICS 3.5: EVALUATING AND CORRECTING PERSISTENT FAILURE TO REACH OR MAINTAIN INTENDED HB LEVEL

Although relative resistance to the effect of ESAs pediatric population and there is no reason for a is a common problem in managing the anemia of different recommendation. CKD and is the subject of intense interest, the bulk of available information suggests that—in Definition of Hyporesponsiveness the absence of iron deficiency—there are few The principles outlined in the adult guideline readily reversible factors that contribute to ESA should be kept in mind when assessing a child hyporesponsiveness. who appears to have hyporesponsiveness to ESA 3.5.1 Hyporesponse to ESA and iron therapy; however, the indications as outlined in therapy: (FULLY APPLICABLE TO) Guideline 3.5.1 may not be directly applicable to CHILDREN children because there are no published data In the opinion of the Work Group, the from which we can confidently define an exces- patient with anemia and CKD should sive epoetin dose to maintain a given Hb target. undergo evaluation for specific causes In terms of the mean ESA dose required, it is of hyporesponse whenever the Hb clear from the NAPRTCS 2004 annual report level is inappropriately low for the that in all children, the average dose never ap- ESA dose administered. Such condi- pears to exceed approximately 350 U/kg/wk, even in the youngest children and even during up tions include, but are not limited to: 325 ● A significant increase in the ESA to 30 months of follow-up. What is not clear is dose requirement to maintain a cer- what the upper limit of this dose is in children tain Hb level or a significant decrease who seem to respond with an increase in Hb in Hb levels at a constant ESA dose. level; hence, it is not currently feasible to set an ● A failure to increase the Hb level to upper limit dose that, if exceeded in a child greater than 11 g/dL despite an whose Hb level is persistently less than 11.0 ESA dose equivalent to epoetin g/dL, should trigger a consideration for the pres- greater than 500 IU/kg/wk. ence of hyporesponsiveness, although the target in children younger than 5 years would intu- 3.5.2 Evaluation for PRCA: (FULLY AP- itively be greater than that set in the adult guide- PLICABLE TO CHILDREN) lines. In the opinion of the Work Group, evaluation for antibody-mediated PRCA Potentially Treatable Disorders should be undertaken when a patient In terms of the factors outlined in the adult receiving ESA therapy for more than 4 guideline as being associated with persistent fail- weeks develops each of the following: ure to achieve target Hb levels, these are relevant ● Sudden rapid decline in Hb level at to children as well, and in a retrospective review the rate of 0.5 to 1.0 g/dL/wk, or of 23 patients on HD therapy for more than 6 requirement of red blood cell trans- months, it was shown that there appeared to be a fusions at the rate of approximately relationship between the need for high doses of 1 to 2 per week, AND rHuEPO (defined as Ͼ 450 U/kg/wk) in patients ● Normal platelet and white blood who were younger, weighed less, had greater cell counts, AND parathyroid hormone levels, and had more epi- ● Absolute reticulocyte count less sodes of bacteremia (all P Ͻ 0.03).341 An unex- ␮ than 10,000/ L. pected finding in this study, relating a high serum RATIONALE ferritin level to an increased risk for hyporesponsiveness, likely (as suggested by the investiga- Hyporesponsiveness to ESA and Iron tors) was related to the presence of inflammation This guideline is considered applicable to chil- in patients with more frequent episodes of bacte- dren because there are no special data in the remia.

American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S107-S108 S107 S108 ANEMIA IN CHRONIC KIDNEY DISEASE IN CHILDREN

Evaluation for Diagnosing and Treating Anti- year old patients developed PRCA after 18 months body-Mediated PRCA of exposure to an ESA, whereas the 17-year-old This guideline is considered applicable to chil- developed it 7 months after ﬁrst exposure to an dren because there are no special data in the ESA. The latter subsequently died of cardiac pediatric population and there is no reason for a arrest 20 months after the original diagnosis; no different recommendation. data about any form of therapy are available. A brief review of pediatric PRCA cases is There are no outcome or therapy data available presented next. for the 19-year-old patient; however, the 11-year- A study reviewed all 191 conﬁrmed cases of old received a renal transplant and subsequently PRCA up until April 2004, including 3 pediatric “recovered” from PRCA. patients; an 11-year-old boy from the United Obviously, these numbers are very small and, States, a 17-year-old girl from Brazil, and a for practical purposes, suggest a very low risk for 19-year-old boy from France.278 Two of these PRCA in children compared with the risk inher- patients, the 11- and 17-year-olds, received only ent in being anemic or requiring repeated blood 1 ESA, epoetin alfa (Procrit®; Ortho Biotech transfusions. However, it should be recognized Products, LP, Bridgewater, NJ and Eprex®;Or- that this problem can occur in children and tho Biotech Products, LP, Bridgewater, NJ, re- should be part of the differential for a child with spectively) before developing PRCA. The 19-year- a low reticulocyte anemia in the presence of a old received both epoetin alfa and epoeten beta, decreasing Hb level and preserved white blood Eprex® and Neorecormon® (F. Hoffmann-La cell and platelet lines that fail to respond to Roche, Ltd, Basel, Switzerland). The 11- and 19- increased iron and/or ESA therapy. IV. CLINICAL PRACTICE RECOMMENDATIONS FOR ANEMIA IN CHRONIC KIDNEY DISEASE IN TRANSPLANT RECIPIENTS ANEMIA IN CKD IN TRANSPLANT RECIPIENTS

BACKGROUND functioning transplants include preemptive trans- Anemia is relatively common after transplan- plant recipients, recipients with variable expo- tation. Regular screening and careful evaluation sure to dialysis, and repeat and multiorgan trans- of the multiple factors that can contribute to plant recipients. Consequently, the burden of anemia after transplantation are recommended. comorbid disease varies among transplant recipi- Table 41 summarizes the key features of post- ents, and this may have important implications transplantation anemia (PTA). As in other forms for the severity and management of anemia after of CKD, the prevalence of anemia in transplant transplantation. CKD clearly is associated with the level of Most transplant recipients have decreased kid- allograft function. However, a number of other ney function. Among 40,963 transplant recipi- factors unique to transplant recipients may con- ents studied in the United States between 1987 tribute to the development of PTA. There is little and 1996, mean eGFR achieved at 6 months after information regarding the association of anemia the time of transplantation was 49.6 Ϯ 15.4 (SD) with either graft or patient outcomes. Similarly, mL/min/1.73 m2.376 In a study of transplant there is limited information to suggest that the recipients in the United States between 1987 and response to treatment with ESAs differs between 1998, more than 70% of patients with allograft patients with CKD with and without a transplant. function at 1, 3, and 5 years after transplantation Although there are reports of increased delayed had CKD stage 3 or higher.377 Transplant func- graft function and hypertension with the use of tion may change over time. Although the mean ESAs, there is insufficient evidence to suggest rate of GFR decline was only 1.66 Ϯ 6.51 that these agents should not be used in either the mL/min/1.73 m2 per year in a large cohort of immediate posttransplantation or late posttrans- long-term kidney transplant recipients, kidney plantation period. Conversely, with the existing allograft function may change rapidly; thus, fre- information, it is recommended that treatment quent evaluation of kidney function and Hb level guidelines for management of anemia in the is desirable.376 general CKD population be followed in the trans- Transplant recipients differ from most other plant population. patients with CKD because they are chronically exposed to immunosuppressive medications that INTRODUCTION may suppress erythropoiesis. The current KDOQI Transplant recipients are a subgroup of pa- classification of CKD includes transplant recipi- tients with CKD with unique considerations with ents.4 A recent worldwide consensus statement regard to anemia management. of KDIGO also confirmed that the CKD classifi- Transplant recipients have variable exposure cation system should be applied to transplant to CKD before transplantation. Patients with recipients and suggested to add the index letter

Table 41. Literature Review of Anemia In Transplant CKD: Key Conclusions Prevalence of Pathophysiology Clinical outcomes Response to Side effects of Anemia treatment treatment Prevalence varies Level of transplant Evidence linking Transplant patients Evidence is insufficient to by time after kidney function is anemia and anemia may be relatively determine the relationship, transplantation and important, but other treatment with graft resistant to ESA. if any, between ESA use by level of graft factors, including or patient outcomes, The precise and potential adverse function. medications, may including survival, mechanisms have events in the transplant Relationship decrease erythropoietin cardiovascular not been CKD patient, including between Hb and production or increase disease, & elucidated. delayed graft function, eGFR may not be resistance to hospitalization is decline in kidney function, the same as that erythropoietin in limited. or hypertension. seen in other CKD patients with stable disorders. allograft function.

S110 American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S110-S116 ANEMIA IN CHRONIC KIDNEY DISEASE IN TRANSPLANT RECIPIENTS S111

“T” to the CKD stage category for these patients presenting for kidney transplantation. A patients.377A Although many features distinguish study found that the prevalence of pretransplan- transplant recipients from other patients with tation anemia (defined as Hct Ͻ 33%) was 41% CKD, transplant recipients are clearly at risk for in adults.393 In a study of pediatric transplant anemia because of decreased kidney function. recipients, the prevalence of pretransplantation This section reviews the available information anemia (defined as Hct Ͼ 2 SDs less than age- regarding the prevalence, pathophysiology, clini- specific means) was 67%.394 In the Transplant cal correlates, and treatment considerations of European Survey on Anemia Management anemia in kidney transplant recipients. The rel- (TRESAM), 4 cohorts of prevalent transplant evant considerations for anemia management recipients were defined by the duration of trans- before transplantation and after transplant failure plantation (6 months and 1, 3, and 5 years). also are discussed. Mean pretransplantation Hb levels in these cohorts were 11.9 Ϯ 1.7 (SD), 11.7 Ϯ 1.8, 11.2 Ϯ PREVALENCE 1.8, and 10.8 Ϯ 1.8 g/dL, respectively. The There is no accepted definition of anemia in higher mean Hb levels in cohorts that underwent transplant recipients, and variable definitions are transplantation more recently suggested an im- used in the literature. A second important consid- provement in pretransplantation anemia manage- eration is that the prevalence of anemia is depen- ment over time.395 In a study of adult first dent on the time of observation after transplanta- deceased-donor transplant recipients in the United tion. States between 1995 and 2000, the prevalence of During the early posttransplantation period, anemia before transplantation was described arbitrarily defined as the first 6 months after among the subset of patients who had Medicare transplantation, anemia of varying degrees is as the primary payer for the 12 months before very common. The prevalence and degree of transplantation and who received ESAs before anemia during this period are dependent on the transplantation. Mean pretransplantation Hb lev- pretransplantation Hb level, amount of perioper- els recorded in the 6 months before transplanta- ative blood loss, frequency of blood draws, iron tion by transplantation year were: 1995 (n ϭ depletion,378-383 persistence of uremia,382 endog- 2296), 10.5 Ϯ 0.02 (SE) g/dL; 1996 (n ϭ 2394), enous erythropoietin levels,382,384,385 erythropoi- 10.7 Ϯ 0.02 g/dL; 1997 (n ϭ 2694), 10.9 Ϯ 0.02 etin responsiveness,382,384-387 and exposure to g/dL; 1998 (n ϭ 2838), 11.0 Ϯ 0.02 g/dL; 1999 immunosuppressive agents.388,389 (n ϭ 2723), 11.4 Ϯ 0.02 g/dL; and 2000 (n ϭ The time course of erythropoiesis after trans- 2857), 11.6 Ϯ 0.02 g/dL. The higher Hb levels plantation has been studied by a number of among patients who underwent transplantation investigators,378-382,385,390,391 and reviews on more recently suggested an improvement in pre- this subject are available.389,391,392 A transient transplantation anemia management over time. early peak of erythropoietin is detectable within Nonetheless, in 20% of patients, Hb levels de- the first 24 hours after transplantation, particu- creased to less than the KDOQI target of 11 larly in patients with delayed graft function, g/dL.395A and is not associated with a measurable in- Table 42 summarizes the published studies crease in Hb level. Within the first number of describing the prevalence of PTA. Together, the days after successful transplantation, a smaller, existing literature indicates that PTA is highly more sustained erythropoietin peak is detect- prevalent. The results from the few longitudinal able. This peak is associated with the subse- studies showed a very high prevalence of anemia quent onset of erythropoiesis and recovery of in the early posttransplantation period; anemia renal anemia during the next several months appears to be least prevalent 1 year after trans- after transplantation. plantation and then increases in prevalence with The prevalence and degree of anemia in the time after transplantation. This increase possibly immediate posttransplantation period also will is related to declining allograft function. The be dependent on the pretransplantation Hb level. available information suggests that the preva- Despite the widespread availability of ESAs, lence of anemia is greater in pediatric compared anemia continues to be a significant issue in with adult patients. S112 ANEMIA IN CHRONIC KIDNEY DISEASE IN TRANSPLANT RECIPIENTS

Table 42. Prevalence of PTA by Duration of Posttransplantation Period Adult/ Timing of Anemia Pediatric or Determination after Author, Year Location Mean Age N Definition of Anemia Transplantation Prevalence Saito, 1998 396, 397 Japan Adult 60 Male Hb <12.8 g/dL Cross-sectional 23% Female Hb <11.5 g/dL Yorgin, 2002 393 US 1> 8 yr 21 8 cH t < %33 ylraeY rof 5 ry emiT tnalpsnart-tsop )ry( 0 = 43% 1 = 12% 2 = 12% 3 = 14% 4 = 18% 5 = 26% Mix, 2003 398 SU dA ult 42 0 cH t <36% Lo gn itudi an l evo r 5 ry Time p so art-t alpsn nt ( ry ) 0 = 76% 1 = 21% 4 = 36% Lorenz, 2002 399 Austria Adult 438 Male Hb <13 g/dL Cross-sectional mean 39.7% Female Hb <12 g/dL 4.9 yr

Winklelmayer, 2004 400 US Adult 374 Hct <33% Cross-sectional 28.6% Mean 7.7 ± 6.7 yr posttransplantation Vanrenterghem, 2004 395 Europe 48 ±13 yr 4,263 Male Hb ≤13 g/dL 4 cohorts 38.6% overall included children Female Hb ≤12 g/dL >10 yr 6 mo (n = 1003) No significant difference in 12 mo (n = 960) prevalence among cohorts 3 yr (n = 1254) 5 yr (n = 1046) Yorgin, 2002 394 SU eP id atr ci 61 2 2 DS b elow -ega s ep fic ic means Longitudinal 67% at transplant 84% 1 mo 64%-82% between 0-60 mo Shibagaki, 2004 401 US Adult 192 Male Hb ≤13 g/dL 6 mo 6 mo = 41% Female Hb ≤12 g/dL 12 mo 12 mo = 45%

Hb ≤11 g/dL in women, ≤12 g/dL in men was 19% and 20% at 6 mo and 12 mo Kausman, 2004 402 Australia Pediatric 50 Hb <110 g/L Cross-sectional 60% Hb <110 g/dL

Severe anemia Hb <100 g/L 30% Hb <100 g/dL

PATHOPHYSIOLOGY CKD stages 1, 2, 3, 4, and 5 were 0%, 2.9%, 403 A number of factors may cause PTA; some are 6.6%, 27%, and 33%, respectively. shared with other patients with CKD, whereas The association of level of allograft function others are unique to transplant recipients.398 with Hb level appears to vary with the time of observation after transplantation. One study found Factors Shared With Other Patients With CKD that, among patients with an eGFR greater than 2 In general, the evaluation of PTA should paral- 90 mL/min/1.73 m , 11% and 7% of patients had lel that among nontransplantation patients with anemia at 6 and 12 months after transplantation, CKD. The discussion regarding factors contribut- whereas among patients with an eGFR less than 2 ing to PTA that are shared with other patients 30 mL/min/1.73 m , at 6 and 12 months after with CKD is limited to the most common consid- transplantation, 60% and 76% were anemic, re- erations and to unique considerations in the trans- spectively.398 These findings suggest that factors plantation setting. in addition to level of allograft function may be important determinants of anemia, particularly Kidney Function during the early posttransplantation period. Level of allograft function is clearly an impor- Whether the association between anemia and tant determinant of PTA. In the TRESAM, there level of kidney function differs in patients with was a strong association of anemia with kidney CKD who did and did not undergo transplanta- transplant function. Of 904 patients with an SCr tion is uncertain. In a study of 23 renal transplant level greater than 2 mg/dL, 60.1% were anemic recipients with stable kidney function, 12 anemic compared with 29% of those with an SCr less patients were compared with the 11 nonanemic than 2 mg/dL, P Ͻ 0.01. In a single-center study control patients.386 Of the 12 anemic patients, 10 of 459 patients at least 6 months after transplan- had low erythropoietin levels suggestive of eryth- tation, prevalences of anemia, defined as an Hb ropoietin deficiency, 2 patients had higher than level less than 11 g/dL, among patients with anticipated erythropoietin levels suggestive of ANEMIA IN CHRONIC KIDNEY DISEASE IN TRANSPLANT RECIPIENTS S113 erythropoietin resistance, whereas 5 of 11 nonane- tion and anemia, a cluster of 11 genes involved in mic control patients had higher than expected Hb transcription and synthesis, iron and folate erythropoietin levels. Thus, there appears to be binding, and transport were found to be down- significant variation in erythropoietin production regulated. An additional mechanism for the devel- and responsiveness in transplant recipients, which opment of anemia during rejection is thrombotic may alter the association between kidney func- microangiopathy, which may develop during epi- tion and anemia in transplant recipients. Other sodes of severe vascular rejection. clinically evident examples of the dissociation Medications between Hb level and kidney function in trans- Immunosuppressive medications. The use plant recipients include posttransplantation eryth- of myelosuppressive medications for immunosup- rocytosis and, as in nontransplantation patients pression and antiviral prophylaxis or treatment with CKD, the lower incidence of anemia among transplant recipients with polycystic kidney dis- may be important factors in the development of ease.395 anemia after transplantation. Azathioprine and mycophenolate mofetil are myelosuppressive; Iron Deficiency therefore, anemia caused by these drugs often is Iron deficiency may be an important factor in associated with leukopenia and/or thrombocytopenia. Very rarely, PRCA may occur with the use the development of anemia after transplanta- 388,408,409 tion.404 There is limited information regarding of these drugs. the prevalence of iron deficiency after transplan- Calcineurin inhibitors infrequently are associ- tation. In a cross-sectional study of 438 prevalent ated with anemia. The most common mechanism for PTA associated with the use of calcineurin transplant recipients, the prevalence of iron defi- 410- ciency, defined as a percentage of hypochromic inhibitors is microangiopathy and hemolysis. red blood cells of 2.5% or greater, was 20.1%. In 414 The immunosuppressant OKT3 also has been associated with hemolytic uremic syndrome another study of 439 prevalent transplant recipi- 415,416 ents, 41% of patients had a TSAT less than 20%, (HUS) and microangiopathy. whereas 44% had a ferritin level less than 100 Anemia was a significant AE in a phase III ng/mL.403 trial in which sirolimus was administered with 417 The prevalence of iron deficiency may be cyclosporine and corticosteroids. A group greater in the early transplantation period be- reviewed the 10-year experience with siroli- cause of low pretransplantation iron stores in mus and reported a dose-dependent associa- dialysis patients and increased iron utilization tion of anemia with the drug in phase I and II 418 with the onset of erythropoiesis after successful trials. The association of sirolimus with transplantation.383 In 1 study, 24 of 51 patients anemia also recently was shown in single- 419 were found to be iron deficient in the early center analyses. Sirolimus may inhibit eryth- posttransplantation period.381 In a prospective ropoiesis by interfering with intracellular sig- study of 112 transplant recipients, serum ferritin naling pathways normally activated after the levels decreased from 109.6 ␮g/L (range, 21 to binding of erythropoietin to its receptor, and 4,420 ␮g/L) at transplantation to 54.9 ␮g/L sirolimus also may be associated with throm- (range, 2 to 1,516 ␮g/L) at 6 months after trans- botic microangiopathy.419,420 plantation.405 Antiviral and antimicrobial medications. A number of commonly used antivirals and antibi- Transplant-Specific Factors otics may cause anemia, including ganciclovir Acute Rejection and trimethoprim-sulfamethoxazole. Early acute rejection is reported to cause a Angiotensin-converting enzyme inhibitors sharp decrease in erythropoietin production and and angiotensin II receptor antagonists. An- anemia.406 Insights into the molecular mecha- giotensin-converting enzyme (ACE) inhibitors nisms involved in the development of anemia and angiotensin receptor blockers (ARBs) may during allograft rejection have been elucidated be associated with PTA. Anemic patients in the from gene expression studies.407 Among 4 pedi- TRESAM had higher odds of receiving ACE atric renal allograft recipients with acute rejec- inhibitors or ARBs (odds ratio, 1.55; 95% CI, S114 ANEMIA IN CHRONIC KIDNEY DISEASE IN TRANSPLANT RECIPIENTS

1.34 to 1.80; P Ͻ 0.001).395 In a single-center sis in cultured microvascular endothelial cells.429 retrospective study, a significant curvilinear dose- The mechanism appears to be linked to induction response relationship was identified between of Fas (CD95) on cultured endothelial cells be- ACE-inhibitor dose and Hct.400 The underlying cause the erythropoietin receptor is expressed on mechanisms are complex and include inhibition vascular endothelial cells.430, 431 Erythropoietin of endogenous erythropoietin production, inhibi- prevents lipopolysaccharide-induced apoptosis in tion of angiotensin IIÐmediated stimulation of cultured endothelial cells, suggesting that eryth- red blood cell precursors,421 and the generation ropoietin may have a protective effect and may of an erythropoiesis-inhibiting protein by ACE be of therapeutic benefit in TTP/HUS in the inhibitors.422 nontransplantation setting.432 The use of ESAs Infections and Malignancy after transplantation in patients with HUS war- Anemia may be a feature of cytomegalovirus rants further study. (CMV) infection. Parvovirus B19 infection has Hemolytic Anemia Associated With Minor Blood been reported in transplant recipients with ane- Group A, Group B, Group O Incompatibility mia and may cause PRCA.423,424 Blood group A recipients receiving transplants Hemophagocytic syndrome (HPS) is a rare from blood group O donors or blood group AB cause of PTA. The syndrome often is caused by recipients receiving transplants from either group infectious or neoplastic disease and is defined by A or B donors may develop evidence of hemoly- bone marrow and organ infiltration by activated sis caused by anti-A or anti-B antibodies of the nonmalignant macrophages that phagocytose red donor or from autoantibodies produced by passen- blood cells. A retrospective analysis of 17 cases ger lymphocytes.433-435 among deceased donor transplant recipients showed that the syndrome developed after a CLINICAL OUTCOMES median duration of 52 days after transplantation. There is only limited information regarding Fever was present in all patients and hepato- the association of anemia with clinical outcomes splenomegaly was present in 9 of 17 patients. in transplant recipients. Eleven patients had received antilymphocyte Transplant recipients are known to be at in- globulins in the 3 months before presentation. In creased risk for cardiovascular events, particu- 9 patients, HPS was related to viral infections larly during the perioperative period. In a single- (CMV, Epstein-Barr virus, and human heprus center study of 404 transplant recipients with virus 6 and 8); other infections included tubercu- diabetes between 1997 and 2000, patients with at losis, toxoplasmosis, and Pneumocystis carinii least 1 monthly Hct level less than 30% during pneumoniae. Posttransplantation lymphoprolif- the first 6 months after transplantation had a erative disease was present in 2 patients. The significantly greater incidence of cardiovascular syndrome has a poor prognosis—8 of 17 patients events compared with patients with monthly Hct died despite the use of anti-infectious therapy 425 values greater than 30%. In a multivariate analy- and tapering of immunosuppression. sis that also included patient age and history of Hemolytic Uremic Syndrome ischemic heart disease (IHD), an increase in HUS may recur after transplantation and can monthly Hct to greater than 30% was associated result in allograft loss.426 De novo HUS may with a significantly lower risk for cardiovascular occur associated with the use of cyclosporine, events (RR, 0.65; 95% CI, 0.33 to 0.91; P ϭ tacrolimus, or OKT3. The syndrome also has 0.02).436 been associated with CMV and influenza A infec- Among long-term transplant recipients, there tion in transplant recipients.427,428 The possibil- is limited information regarding the association ity that erythropoietic agents may be beneficial of anemia with adverse cardiovascular events. In in patients with HUS after transplantation is a retrospective study of 638 transplant recipients suggested by observations from the nontransplan- between 1969 and 1999 who were alive and free tation population. Plasma from approximately of cardiac disease 1 year after transplantation, 75% of patients with sporadic thrombotic throm- lower Hb levels were associated with an in- bocytopenic purpura (TTP)/HUS induces apopto- creased risk for de novo CHF (RR, 1.24/10-g/dL ANEMIA IN CHRONIC KIDNEY DISEASE IN TRANSPLANT RECIPIENTS S115 decrease in Hb; 95% CI, 1.10 to 1.39; P ϭ immediate posttransplantation period have been 0.001).437 In a follow-up study, anemia was performed. associated with increase in left ventricular mass One study randomized 14 transplant recipients (as measured by Cornell voltage on electrocardio- to receive and 15 patients not to receive an gram) during the first 5 years after transplanta- ESA.387 The ESA (150 U/kg/wk) was started at tion.438 an Hct less than 30% and was increased at In a recent study, 438 prevalent transplant weekly intervals by 30 U/kg/wk as long as Hct recipients were followed up for a median of 7.8 remained at less than 25%. Hct increased from a years. Laboratory parameters obtained during a nadir of 22% Ϯ 4% 2 weeks after transplantation 4-week enrollment period in 1995 were tested to 30% Ϯ 4% at week 4 and 36% Ϯ 4% at week for their association with all-cause mortality and 6(P Ͻ 0.001 and P Ͻ 0.0001 versus week 2, allograft failure.439 The investigators did not respectively). Corresponding values in the non- identify an association between anemia (Hb Ͻ ESA group were 25% Ϯ 6%, 28% Ϯ 6% (P ϭ 10 g/dL) and all-cause mortality or graft sur- NS), and 32% Ϯ 6% (P Ͻ 0.05 versus week 2; vival. Compared with patients with hypochromic overall ESA versus non-ESA, P ϭ 0.038 by red blood cells less than 5%, patients with hypo- analysis of variance). The maximum ESA dose chromic red blood cells greater than 10% had an after transplantation was more than 2 times higher RR of 2.06 for all-cause mortality (95% CI, 1.12 than that required before transplantation (197.1 to 3.79; P ϭ 0.02). Ϯ 45.1 versus 85.0 Ϯ 76.0 U/kg/wk; P Ͻ 0.05). In a retrospective single-center study of re- The investigators concluded that ESAs could source utilization among 220 kidney transplant safely and effectively correct anemia during the recipients, patients with a higher Hct had a first weeks after transplantation despite relative decreased risk for hospitalization (RR, 0.95/1% erythropoietin resistance. increase in Hct; 95% CI, 0.92 to 0.98; P Ͻ In a recent study, patients were randomized to 0.001).440 receive (n ϭ 22) or not receive (n ϭ 18) ESA, 100 U/kg 3 times per week, if Hb level was less 444 RESPONSE TO TREATMENT WITH than 12.5 g/dL. Time to reach an Hb level Ϯ ESAs AND SAFETY OF TREATMENT greater than 12.5 g/dL was 66.5 14.5 versus 52.6 Ϯ 23.7 days in the non-ESA and ESA Use of ESAs Before Transplantation groups, respectively (P ϭ 0.05). After 3 months, Hb levels were not different between the non- The issue of whether ESA use before transplan- Ϯ Ϯ tation is associated with delayed graft function ESA and ESA groups (12.6 1.5 versus 12.0 1.5 g/dL, respectively). In a Cox regression anal- after transplantation has largely been dispelled ϭ by the decreased incidence of delayed graft func- ysis, ESA use (RR, 7.2; P 0.004) and dose P ϭ tion in registry data over time despite the in- (RR, 0.63; 0.04) were retained as independent variables predicting the time to reach an Hb creased use of ESAs. A few small retrospective level greater than 12.5 g/dL. In the ESA group, studies had suggested that ESA use may be 14 of 22 patients reached the target Hb level of associated with delayed graft function, presum- greater than 12.5 g/dL compared with 12 of 18 ably because of altered intrarenal blood flow patients in the non-ESA group (P ϭ NS). S during rewarming in patients with a higher Cr levels were not different between groups. The Hct.441,442 Previous treatment with ESAs does investigators concluded that the use of ESAs in not blunt the production of endogenous erythro- the immediate posttransplantation period had no poietin after transplantation or the ability to relevant clinical impact on the correction of respond to endogenous erythropoietin.391,443 anemia after transplantation. Together, these studies suggest that ESAs are Early Posttransplantation Period effective in correcting anemia after renal trans- Erythropoietic agents are effective in correct- plantation. The dose of ESA required may be ing anemia during the early posttransplantation higher than that before transplantation. The stud- period. Two small prospective randomized studies were not designed to determine whether the ies to study the efficacy of ESAs during the correction of anemia was associated with im- S116 ANEMIA IN CHRONIC KIDNEY DISEASE IN TRANSPLANT RECIPIENTS provement in clinical outcomes, resource utiliza- study.452 In a single-center study, correction of tion, or QOL. Further, few patients with delayed anemia was associated with a decreased rate of or impaired graft function were studied. Thus, decline in allograft function.456 whether there are clinically relevant benefits to Together, these studies suggest that ESAs are the early correction of anemia after transplanta- efficacious and likely do not accelerate renal tion remains uncertain. decline, but may aggravate hypertension. Whether These studies did not report significant AEs patients in the late posttransplantation period associated with the use of ESAs, such as delayed require higher doses of ESAs to correct anemia graft function or hypertension. A recent study compared with nontransplantation patients with reported a significantly increased incidence of CKD is unclear.457 A decreased response to ESAs transplant renal artery stenosis among pediatric among transplant recipients may be anticipated transplant recipients administered ESAs during because of the use of myelosuppressive medica- the first week after transplantation. All patients tions, chronic inflammation, or other factors. The were enrolled in a steroid-free immunosuppres- propensity of specific immunosuppressive agents sion protocol. Results of DNA microarray analy- to affect the action of ESAs in kidney transplant sis showed a series of 12 genes that differentiated recipients is unclear. There is an inverse correla- the patients who developed renal artery stenosis tion between ESA dose and CCr in transplant in this setting.444A recipients, and a sudden change in Hct in ESA- Erythropoietin modulates the cellular re- treated patients may indicate a change in graft sponse to stress and may attenuate apoptosis and function.453 The response to ESAs is impaired necrosis in various organs, including the kid- during episodes of allograft rejection.407,458 In ney.445-451 Whether use of erythropoietic agents the TRESAM, the median dose among ESA- can minimize ischemic reperfusion injury, as treated patients was 4,000 IU/wk (mean, 5,831 Ϯ shown experimentally, warrants investigation. 4,217 IU/wk), but ESA-treated patients had lower Hb values than nonÐESA-treated patients, sug- Late Posttransplantation Period gesting that the ESA was underdosed and pa- There are only a few uncontrolled studies describ- tients were ESA resistant.395 ing the use of ESAs to treat late PTA.452-455 In the largest of these studies, in 40 patients with failing Posttransplantation Failure renal allografts, mean Hb levels increased from Patients with allograft failure have a high 78.9 Ϯ 10.4 to 102.6 Ϯ 18.4 g/L after 24 weeks of mortality that is related primarily to CVD. The treatment with ESA, 50 U/kg 3 times per week.453 management of anemia among patients with fail- The increase in Hb level was associated with a ing allografts is suboptimal. In a study of patients significant improvement in QOL measures. Twelve initiating dialysis therapy after transplant failure patients returned to dialysis therapy. These patients in the United States between 1995 and 1998, all had poor allograft function, and the study could mean Hct was 27.5% Ϯ 5.9% and 67% had an not exclude the possibility that treatment with ESA Hct less than 30%. The use of ESAs at the time accelerated renal allograft decline. Although no of dialysis therapy initiation was infrequent change in systolic or diastolic blood pressure was (35%).459 Anemia management may be more noted, the need for antihypertensive medications difficult in patients with transplant failure be- was significantly increased in 18 patients. An in- cause of the presence of chronic inflammation crease in hypertension also was noted in another and relative resistance to ESAs.460 V. APPENDIX 1: METHODS OF EVIDENCE REVIEW AND SYNTHESIS

Aim ● Write guideline recommendations and sup- The overall aim of the project is to update the porting rationale statements and grade the 2000 KDOQI CPGs for Anemia of CKD.2 The strength of the recommendations Work Group sought to update the guidelines by ● Write CPRs based on consensus of the using an evidence-based approach. After topics expert Work Group in the absence of and relevant clinical questions were identified, sufficient evidence. the available scientific literature on those topics was systematically searched and summarized. Creation of Groups High-quality or moderately high-quality evidence formed the basis for the development of The KDOQI Co-Chairs appointed the Co- evidence-based CPGs. When evidence was of Chairs of the Work Group, who then as- low or very low quality or was entirely lacking, sembled groups to be responsible for the devel- the Work Group could develop CPRs based on opment of the guidelines. The Work Group consensus of expert opinion. consisted of domain experts, including individuals with expertise in adult and pediatric Overview of Process nephrology, hematology, nursing and nutri- Update of the guidelines required many concur- tion, cognitive function, QOL, and CVD out- rent steps to: comes in patients with CKD. Support in evi- ● Form the Work Group and ERT that were dence review and methods expertise was to be responsible for different aspects of provided by an ERT contracted by the NKF at the process the NKF Center for CPG Development and ● Confer to discuss process, methods, and Implementation. The Work Group and the ERT results collaborated closely throughout the project. ● Develop and refine topics The first task of the Work Group members was ● Create draft guideline statements and ra- to define the overall topics and goals for the tionales update. Smaller groups of 2 to 4 individuals were ● Define exact populations, interventions, formed and assigned to each topic. The Work predictors, comparisons groups and out- Group and ERT then further developed and re- comes of interest and study design and fined each topic and specified screening criteria minimum follow-up time criteria (PICOD) for PICOD, literature search strategies, and data ● Create and standardize quality assessment extraction forms (described next). Work Group and applicability metrics members were the principal reviewers of the ● Create data extraction forms ● literature, and from their reviews and detailed Develop literature search strategies and data extractions, they summarized the available run searches evidence and took the primary roles of writing ● Screen abstracts and retrieve full articles the guidelines and rationale statements. ● Review articles The ERT consisted of individuals (staff, fel- ● Extract data and perform critical appraisal of the literature lows, and research assistants) from TuftsÐNew ● Tabulate data from articles into summary England Medical Center with expertise in ne- tables phrology and development of evidence-based ● Grade quality and applicability of each CPGs. It instructed the Work Group members in study all steps of systematic review and critical litera- ● Grade the quality of evidence for each ture appraisal. The ERT also coordinated the outcome and assess the overall quality of methodological and analytical process of the re- the evidence across all outcomes with the port; it defined and standardized the method of aid of evidence profiles performing literature searches, data extraction, and

American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S117-S125 S117 S118 APPENDIX 1: METHODS OF EVIDENCE REVIEW AND SYNTHESIS

* First Look being available. When experts were uncertain about the current evidence basis of a topic, a “first look” of the topic was undertaken to inform this process (Fig 19). After literature review of potentially relevant abstracts and studies, members of the Work Group focused the specific † Narrative Review ‡ Systematic Review questions deemed clinically relevant and amenable to systematic review or decided to produce Fig 19. Process of triaging a topic to a systematic a narrative summary of the literature. review or a narrative review. *First Look Topic: Top- The Work Groups and ERT developed: (1) ics for which the substance of the evidence base draft guideline statements, (2) draft rationale was unclear were first explored to determine their suitability for systematic review. A sensitive Ovid statements that summarized the expected perti- search was performed for each first look topic by the nent evidence, and (3) data extraction forms ERT. Abstracts were reviewed by the Work Group requesting the data elements to be retrieved from members to determine whether there was an adequately defined and sufficient base of scientific the primary articles. The topic refinement pro- information from which to answer the clinically rel- cess began before literature retrieval and contin- evant question or resolve controversies. Topics that ued through the process of reviewing individual qualified were submitted to systematic review, while topics lacking a sufficient evidence base for system- articles. atic review were summarized by Work Group members in narrative reviews. †Narrative Review: Work Literature Search Group members had wide latitude in summarizing reviews and original articles for topics that were A master reference list was compiled from determined not to be amenable to a systemic review references used in previous evidence-based guide- of the literature. Under special circumstances, fo- lines on Anemia and CKD: cused literature search was performed by ERT for a specific subtopic. ‡Systematic Review: A system- 1. EBPG II, 2004 atic review entailed systematic screening, data abstraction, appraisal, and synthesis of studies in sum- 2. EBPG I, 2000 mary tables and evidence profiles. 3. KDOQI Anemia Guideline Update, 2000 4. DOQI Anemia Guideline, 1997 5. Caring for Australasians with Renal Impair- ment (CARI) Anemia Guideline, 2003 summarizing the evidence in summary tables and For the topics addressed in EBPG II, update evidence profiles. It performed literature searches, searches of MEDLINE were performed. For organized abstract and article screening, created Hb Targets, a module for (Anemia and ESA forms to extract relevant data from articles, orga- and Kidney) was run on articles from January nized Work Group member data extraction, checked 2003 through March 2004. Selective updates data, and tabulated results. Throughout the project, of literature searches were performed through the ERT conducted seminars and provided instruc- November 2004. A pre-MEDLINE search also tion on systematic review, literature searches, data was performed to capture more recent trials extraction, assessment of quality and applicability not yet indexed in MEDLINE. For the topic of of articles, evidence synthesis, and grading of the Iron Targets, the (Anemia and ESA and Kid- quality of evidence and strength of guideline recom- ney) module was modified by adding addi- mendations. tional iron terms and was run to include publications between January 2003 and November Refinement of Topics and Development 2004. For the topics of adjuvants to ESA of Materials treatment, a MEDLINE search was conducted The Work Group reviewed the 2000 KDOQI for [(Anemia and Kidney) and (Androgens, CPGs for Anemia of CKD2 and determined which Statins, Carnitine, Vitamin E, or Ascorbic of the guideline recommendations required up- Acid)] for all articles published from Janu- dates and which could remain unchanged. These ary1989 through September 2004. A separate assessments were based primarily on expert opin- search for studies on prevalence of anemia by ion regarding the likelihood of new evidence eGFR was conducted from January 1999 APPENDIX 1: METHODS OF EVIDENCE REVIEW AND SYNTHESIS S119

Table 43. Example of a Summary Table

Arm 1 Clinical Outcomes (Arm 1 vs. Arm 2 vs. Arm 3) a Mean Hb Author, Year N Arm 2 (g/dL) Target CVD Events Mortality Hospitali- Dialysis Transfusion b (mo) (mo) LVH QOL

(Achieved) Quality

Hb (g/dL) (%) (%) zation Adequacy (%) Follow-up Baseline CKD Stage

Applicability Arm 3

ESA vs. ESA 14.0 *Nonfatal MI *29.6 vs. ∆Kt/V: ESA High 21 vs. 31 See QOL Besarab, 1998 108 1,233 HD-CKD 10.2 14 (12.7-13.3) 3.1 vs. 2.3 — 24.4 NS –0.03 vs. +0.06 P = 0.001 Table ESA Low 10.0 (10.0) NS c NS c P <0.001 13.5-14.5 CVA: ESA High (13.3) 4 vs. 1 ∆URR: See QOL Parfrey, 2005 109 596 HD-CKD 11.0 24 P = 0.045 *NS NS — 0 vs. +2 % — 9.5-11.5 Table ESA Low Other CVD: P <0.05 (10.9) NS ESA High 13-14 (13) Kt/V: See QOL Foley, 2000 110 146 HD-CKD 10.4 11 NS *NS NS — LVD: 1.41 vs. 1.50 — Table ESA Low 9.5-10.5 (10.5) P = 0.025 through February 2005. The searches also were Generation of Evidence Tables supplemented by articles identified by Work The ERT condensed the information from the Group members through September 2005. data extraction forms into evidence tables, which Only full journal articles of original data summarized individual studies. These tables were were included. Editorials, letters, abstracts, created for the Work Group members to assist and unpublished reports were not included. them with review of the evidence and are not Selected review articles identified in the included in this document. All extracted articles searches were provided to the Work Group for and all evidence tables were made available to all background material. Work Group members. During the development MEDLINE search results were screened by of the evidence tables, the ERT rescreened the members of the ERT for relevance by using accepted articles to verify that each of them met predefined eligibility criteria, described in Table the initial screening criteria and checked the data 44. Retrieved articles were screened by the ERT. extraction for accuracy. If the criteria were not Potentially relevant studies were sent to Work met, the article was rejected, in consultation with Group members for rescreening and data extrac- the Work Group. tion. Domain experts, along with the ERT, made the final decision for inclusion or exclusion of all Format for Summary Tables articles. Summary tables describe the studies according to the following dimensions: study size and Generation of Data Extraction Forms follow-up duration, applicability or generalizability, results, and methodological quality (see Table Data extraction forms were designed to cap- 43). The ERT generated summary tables by us- ture information on various aspects of the pri- ing data from extraction forms, evidence tables, mary studies. Data fields for all topics included and/or the articles. All summary tables were study setting, demographics, eligibility criteria, reviewed by the Work Group members. causes of kidney disease, numbers of subjects, In the summary tables, studies were ordered study design, study funding source, dialysis char- first by method quality (best to worst), then by acteristics, comorbid conditions, descriptions of applicability (most to least), and then by study relevant risk factors or interventions, description size (largest to smallest). Results are presented in of outcomes, statistical methods, results, study their appropriate metric or in summary symbols, quality (discussed later), study applicability (dis- as defined in the table footnotes. cussed later), and free text field for comments To provide consistency throughout summary and assessment of biases. Training of the Work tables, data sometimes were converted or esti- Group members to extract data from primary mated. Follow-up times were converted to months articles occurred during Work Group meetings by estimating 1 month as 4 weeks. In general, and by E-mail and teleconference calls. Work data provided as percent Hct was converted into Group members then were assigned the task of grams per deciliter of Hb by dividing by 3. data extraction of articles. Additionally, results sometimes were estimated S120 APPENDIX 1: METHODS OF EVIDENCE REVIEW AND SYNTHESIS APPENDIX 1: METHODS OF EVIDENCE REVIEW AND SYNTHESIS S121

Table 45. Literature Search Yield of Primary Articles for Systematic Review Topics Articles Full Articles Articles Included in Abstracts Articles Added by Data Summary Guideline Topic Search Strategy Dates a Screened Retrieved experts Extracted Tables b 1.1: Identifying patients and Anemia AND kidney AND 1999- 125 13 3 6 6 initiating evaluation GFR 2005 Anemia AND kidney AND 2003- 2.1: Setting Hb targets ESA search as update to 2,013 38 2 28 24 c 2004 master reference list Anemia AND kidney AND ESA AND iron terms 2003- 3.2: Using iron agents 1,848 50 6 34 9 search as update to 2004 master reference list 3.3: Using pharmacological Anemia AND kidney AND and nonpharmacological androgens, statins, 1989- 370 36 8 15 12 adjuvants to ESA treatment carnitine, vitamin E, 2004 in HD-CKD ascorbic acid a. Additional articles were identified by work group members through September 2005. b. Does not include articles included in tables other than summary tables. c. Includes articles that were post hoc analysis of larger studies. from graphs. All estimated values have been adjuvants (carnitine, ascorbic acid, and andro- annotated as such. gens) proceeded to systematic review.

Systematic Review Topics, Study Eligibility Assessment of Individual Studies Criteria Study Size and Duration The topics covered by systematic review are The study (sample) size is used as a measure listed in Table 44. The screening criteria were of the weight of a study. In general, large studies deﬁned by the Work Group members in conjunc- provide more precise estimates of prevalence tion with the ERT. and associations. In addition, large studies are Literature Yield more likely to be generalizable; however, large size alone does not guarantee applicability. A For systematic review topics, the literature study that enrolled a large number of selected searches yielded 2,756 citations. Of these, 137 patients may be less generalizable than several articles were reviewed in full. An additional 19 smaller studies that included a broad spectrum of were added by Work Group members. A total of patient populations. Similarly, longer duration 83 were extracted and of these, 51 studies are studies may be of better quality and more appli- included in Summary Tables. Details of the yield cable, depending on other factors. by topic can be found in Table 45. The literature search yields for ﬁrst-look top- Applicability ics can be found in Table 46. Upon reviewing the Applicability (also known as generalizability resultant abstracts, only the topics of noniron or external validity) addresses the issue of whether S122 APPENDIX 1: METHODS OF EVIDENCE REVIEW AND SYNTHESIS

Table 46. Details of First-Look Topics, Ovid Literature Searches, and Yield by Topic Relevant Abstracts Submitted to No. of Experts for cipoT hcraeS ygetartS setaD Citations Review Pediatric patients Anemia AND kidney AND children 1989-2004 2,101 192 Special populations: Anemia AND kidney AND transplant 1989-2004 680 107 transplant rP egnan yc aimenA A DN yendik A DN erp gnancy 02-9891 04 602 01 Anemia AND kidney AND Sickle cell/ hemoglobinopathies 1989-2004 582 20 hemoglobinopathies HIV enA aim AND k di en y NA D VIH A/ IDS 91 8 -9 002 4 22 8 12 Cancer: MM, MDS, etc. Anemia AND kidney AND cancer 1989-2004 1,557 15 opseropyH isn ssenev aimenA A DN k di yen 2 -100 002 4 95,1 8 46 PRCA ruP e der cell lpa sa ai RO P CR A 002-3002 4 881 641 To lat ,7 41 0 45 8 the study population is sufficiently broad so that study. Because studies with a variety of types of the results can be generalized to the population design were evaluated, a 3-level classification of of interest at large. The study population typi- study quality was devised (Table 46A). cally is defined primarily by the inclusion and Quality of studies of interventions. The evalu- exclusion criteria. The target population was ation of questions of interventions was limited to defined to include patients with anemia and kid- RCTs. The grading of these studies included consid- ney disease and subdivided into those with CKD eration of the methods (ie, duration, degree of stages 3 to 5 not on dialysis therapy and those blinding, number and reasons for dropouts, and so with CKD stage 5 on HD or PD therapy. Further- on), population (ie, does the population studied more, topic 3.6 includes such special patient introduce bias?), outcomes (ie, are the outcomes populations as kidney transplant recipients and clearly defined and properly measured?), thorough- patients with nonrenal anemias. Applicability ness/precision of reporting, statistical methods (ie, was specified for each study according to a was the study sufficiently powered and were the 3-level scale (Table 45A). In making this as- statistical methods valid?), and the funding source. sessment, sociodemographic characteristics Quality of studies of prevalence. The ideal were considered, as well as comorbid condi- study design to assess prevalence of anemia and its tions and prior treatments. Applicability is association with eGFR is a cross-sectional study of graded in reference to the population of inter- a population representative of the general popula- est as defined in the clinical question. Target tion. Criteria for evaluation of cross-sectional stud- populations are specified in the titles of each ies to assess prevalence are listed in Table 47. summary table (discussed later). Results Study Quality The type of results used from a study was Method quality (or internal validity) refers to determined by the study design, the purpose of the design, conduct, and reporting of the clinical the study, and the question(s) being asked for APPENDIX 1: METHODS OF EVIDENCE REVIEW AND SYNTHESIS S123

Table 47. Evaluation of Studies of Prevalence Evaluation of Validity Cross-sectional study design Was there a representative and well-defined sample of the population of interest? • Minimize nonresponse • Define sampling strategy • Subgroups defined in advance Were objective and unbiased criteria used to define cases and controls? Were methods for data collection applied equally to all study participants? Was there adjustment for important prognostic factors? Evaluation of Results How large is the prevalence of cases? How precise are the estimates of prevalence? Are there important differences among subgroups? Evaluation of Clinical Applicability Is the population, or subgroups of the population, under study similar to the population from which my patients are drawn? Were the definitions and measures useful in practice? Are the results useful for estimating probability of disease? which the results were used. Decisions were target audience. Implementation issues and re- based on the screening criteria and prespecifed search recommendations were formulated after this outcomes of interest (Table 47). guideline document had been completed and will be presented in another supplement. Summarizing Reviews and Selected Original Articles Rating the Quality of Evidence and the Work Group members had wide latitude in Strength of Guideline Recommendations summarizing reviews and citing original articles. A structured approach, facilitated by the use of evidence proﬁles and modeled after the Grades Guideline Format of Recommendation, Assessment, Development The format for each section containing an evi- and Evaluation (GRADE) approach,461 was used dence-based guideline or a CPR is outlined in Table to grade the quality of the overall evidence and 48. Each guideline contains 1 or more speciﬁc the strength of recommendations. For each topic, “guideline or statements,” which are presented as the discussion on grading of the quality of the “bullets” that represent recommendations to the overall evidence and the strength of the recom-

Table 48. Format for Guidelines

Introductory Statement Guideline or CPR Statement 1 (strength of recommendation) Guideline or CPR Statement 2 (strength of recommendation) BACKGROUND RATIONALE Definitions (if appropriate) Strength of Evidence Rationale statement 1 Supporting text, summary tables and evidence profiles (where applicable) and description of quality of evidence Rationale statement 2 Supporting text, summary tables and evidence profiles (where applicable) and description of quality of evidence LIMITATIONS S124 APPENDIX 1: METHODS OF EVIDENCE REVIEW AND SYNTHESIS

Table 49. Balance of Benefit and Harm

When there was enough evidence to determine the balance of medical benefits and harm of an intervention to a patient, one of th following conclusions were drawn: • There is net benefit • There is no net benefit • There is net harm When there is not enough or not enough good quality evidence to weigh benefit and harm (for example, inconsistency, unclear harm), then the balance of benefit and harm was classified as: • It is uncertain that there is benefit mendations was led by the primary expert review- of each of the outcomes (eg, death and access ers of each topic, with participation by the Work thrombosis having greater weight than change in Group chairs, all other Work Group members, ESA dose or Hb level). The actual results were and the ERT members. reviewed for each outcome to judge the balance between benefits and harm (Table 49). Four final Grading the Quality of Evidence categories for the quality of overall evidence The quality of a body of evidence pertaining to a were used, as shown in Table 50. particular outcome of interest initially was catego- Evidence profiles were constructed by the ERT rized based on study design. For questions of inter- to record decisions about grades and interpreta- ventions, the initial quality grade was high if the tion of summary effects by the Work Group evidence consisted of RCTs, low if it consisted of members. These profiles serve to make transpar- observational studies, or very low if it consisted of ent to the reader the thinking process of the Work studies of other study designs. Work Group mem- Group in systematically combining evidence and bers decided a priori to include only RCTs for judgments. Each evidence profile was filled in by questions of interventions other than harm. The Work Group experts with ERT guidance. Deci- quality rating for each intervention/outcome pair sions were based on facts and findings from the then was decreased if there were some or serious primary studies listed in corresponding summary limitations to the quality of the aggregate of stud- tables; additional information related to AEs in ies, there were important inconsistencies in the nontarget populations, when applicable; and judg- results across studies, the applicability of the stud- ments of the Work Group. Judgments about the ies to the population of interest was limited or there quality, consistency, and directness of evidence was uncertainty about the directness of evidence, often were complex, as were judgments about the data were imprecise or sparse, or there was a the importance of an outcome or the net medical high likelihood of bias. The final grade for the benefit across all outcomes. quality of the evidence for an intervention/outcome The evidence profiles provided a structured pair could be 1 of the following 4 grades: high, approach to grading, rather than a rigorous moderately high, low, or very low. method of quantitatively summing up grades. In The quality of the overall body of evidence an effort to balance simplicity with full and then was determined based on the quality grades transparent consideration of the important issues, for all outcomes of interest, taking into account footnotes were placed to provide the rationale for explicit judgments about the relative importance grading (Table 51).

Table 50. Definitions of Grades for Quality of Overall Evidence Grade Definition High Further research is unlikely to change our confidence in the estimate of effect. Moderately High Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Low Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Very Low Any estimate of effect is very uncertain APPENDIX 1: METHODS OF EVIDENCE REVIEW AND SYNTHESIS S125

Table 51. Example of an Evidence Profile

Directness of Summary of Findings Total N Methodological Evidence, Quality of No. of Studies & of Quality of Consistency including Other Evidence for Importance of Outcome Study Design Patients Studies across Studies Applicability Considerations Outcome Qualitative Description of Effect Size Outcome No consistent benefit. Only 1 of 3 papers Hb/Hct Serious Important showing significant between-arm Level / 3 RCTs 63 Direct c,d Sparse data Very low Moderate limitations a inconsistencies b comparison in Hb/Hct (difference in Hb of ESA dose 1.6 g/dL). c CKD population: 3 RCTs Mild to severe AEs noted in RCTs included severe acne, elevated AST e, discomfort at injection site. This is consistent with reported AEs AEs 63+ from androgens in non-CKD populations which include virilization, priapism, peliosis hepatis, liver enzyme abnormalities, and hepatocellular High General population: carcinoma. Mechanism of action and profile of AEs of androgens are believed to be similar in non-CKD and CKD populations. trials, case reports, narrative reviews Balance of Benefit and Harm: Quality of Overall Evidence: No Net Benefit Very Low for Benefit

Grading the Strength of the Recommendations of the recommendation than would be support- The quality of evidence for each outcome and able based on the quality of evidence alone. across all outcomes was graded in the evidence After grading the quality of the overall evi- profile. The guideline recommendation was dence for a topic, the Work Group triaged the graded based on the quality of the overall evi- recommendations to either an evidence-based dence, as well as additional considerations. Addi- guideline recommendation when the quality of tional considerations, such as feasibility, availabil- evidence was high or moderately high or other- ity of a service, and regional and population wise to an opinion-based CPR. differences were implicitly considered. Costs also In the absence of strong or moderately strong were considered implicitly, but, in most cases, it quality evidence or when additional consider- was believed that the grading of the evidence and ations did not support strong or moderately strong formulation of a guideline and its strength should evidence-based guideline recommendations, the rest primarily on the evidence for medical benefit Work Group could elect to issue CPRs based on to a patient. consensus of expert opinions. These recommen- The strength of each guideline recommenda- dations are prefaced by “In the opinion of the tion was rated as either “strong” or “moderately Work Group,” and are based on the consensus of strong.” A “strong” rating indicates “it is strongly the Work Group that following the recommenda- recommended that clinicians routinely follow tions might improve health outcomes. As such, the guideline for eligible patients. There is high- the Work Group recommends that clinicians con- quality evidence that the practice results in net sider following the recommendation for eligible medical benefit to the patient.” The “moderately patients. Issues considered in the grading of the strong” rating indicates “it is recommended that quality of the evidence and the strength of the clinicians routinely follow the guideline for eli- recommendation are discussed in the Rationale gible patients. There is at least moderately high- section of each recommendation. quality evidence that the practice results in net medical benefit to the patient.” Overall, the Limitations of Approach strength of the guideline recommendation was While the literature searches were intended to be based on the extent to which the Work Group comprehensive, they were not exhaustive. MED- could be confident that adherence will do more LINE was the only database searched, and searches good than harm. Strong guidelines require sup- were limited to English-language publications. Hand port by evidence of high quality. Moderately searches of journals were not performed, and re- strong guidelines require support by evidence of view articles and textbook chapters were not system- at least moderately high quality. Incorporation of atically searched. However, important studies additional considerations modified the linkage known to the domain experts that were missed by between quality of evidence and strength of the electronic literature were added for consider- guidelines, usually resulting in a lower strength ation. WORK GROUP BIOGRAPHIES

John W. Adamson, MD, has served as Execu- KDOQI. He was a member of the KDOQI Work tive Vice President for Research and Director of Group that developed the guidelines for Blood the Blood Research Institute of the Blood Center Pressure and Antihypertensive Agents in CKD. of Southeastern Wisconsin in Milwaukee since He is a member of the International Society of 1998. He holds the position of Professor of Peritoneal Dialysis’ Work Group for the Manage- Medicine (Hematology) at the Medical College ment of Peritonitis. He was designated Fellow of of Wisconsin. Before moving to Milwaukee, he the Royal Pharmaceutical Society of Great Brit- was Director of the Lindsley F. Kimball Re- ain for his sustained contributions to the litera- search Institute of the New York Blood Center ture of pharmacy and is a Fellow of the American since 1989 and President of the Center from Society of Nephrology. He is a director of Ne- 1989 to 1997. Dr Adamson received his MD phrology Pharmacy Associates, Inc, a private from the University of California, Los Angeles, consulting company. Dr Bailie has received re- after which he trained at the University of Wash- search funds, grants or contracts from Renal ington in Seattle and at the National Institutes of Research Institute, American Regent, Genzyme Health in Bethesda, MD, in the fields of internal and Omnicare. medicine and hematology. Before assuming Jeffrey S. Berns, MD, earned his MD at Case his position in New York, Dr Adamson was Western Reserve University, then went on to professor of medicine and head of the Division complete his internship and residency in Internal of Hematology at the University of Washing- Medicine at University Hospitals of Cleveland. ton. Dr Adamson is a past-President of the He did a fellowship in Nephrology and was an American Society of Hematology and past chair- Associate Research Scientist in the Department man of its committees on scientific affairs and of Physiology at Yale University. Dr Berns re- transfusion medicine. Dr Adamson served as a cently was promoted to Professor of Medicine at member of the Advisory Council of the National the University of Pennsylvania School of Medi- Institute of Diabetes, Digestive and Kidney Dis- cine, where he is Director of Clinical Nephrol- eases of the National Institutes of Health. In ogy and Director of the Renal Fellowship Pro- 1988, he was designated clinical research professor gram for the Renal, Electrolyte and Hypertension by the American Cancer Society and elected a Division. Dr Berns was a member of the NKF- Fellow of the American Association for the Ad- DOQI and KDOQI Anemia Workgroup. He has vancement of Science. Dr Adamson is past editor- published and lectured on topics related to chronic in-chief of Blood, past editor of the Journal of kidney disease, anemia management in patients Cellular Physiology, and founding editor of Cur- with CKD, and other areas in clinical nephrol- rent Opinion in Hematology. Altogether, he has ogy. He is co-editor of Drug Prescribing in authored or co-authored more than 400 scientific Renal Failure-Dosing Guidelines for Adults.He publications. Dr Adamson has received research also serves on the editorial board of Seminars in funds, grants, or contracts from Watson Pharma- Dialysis, American Journal of Kidney Diseases, ceuticals and Navigant Biotechnologies. and Clinical Journal of the American Society of George R. Bailie, MSc, PharmD, PhD, is Nephrology. He is an active investigator in clini- Professor of Pharmacy Practice at Albany Col- cal trials related to anemia treatment in patients lege of Pharmacy. He has 20 years of experience with CKD. Dr Berns has received research funds, in teaching, practice, and research in pharmaco- grants or contracts from Hoffman La Roche, therapeutics and pharmacokinetics in CKD and Ortho Biotech, Advanced Magnetics, Inc., and dialysis. He has published more than 100 origi- Amgen. nal research papers, plus numerous chapters, Kai-Uwe Eckardt, MD (Work Group Co- abstracts, reviews, and cases in the medical and Chair), is Professor of Medicine and Chief of pharmacy literature. He has served on the edito- Nephrology and Hypertension at the University rial board and is a regular reviewer for many of ErlangenÐNuremberg, Germany. He received pharmacy and nephrology journals. Dr Bailie is a his MD from the Westfälische Wilhelms-Univer- member of the Advisory Board for the NKF- sität Münster, Germany. In 1993, following post-

S126 American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S126-S130 WORK GROUP BIOGRAPHIES S127 graduate training in internal medicine, pathol- 2002, Dr Foley worked at Hope Hospital, Sal- ogy, and physiology, he was appointed Assistant ford, UK, and has been Director, Nephrology Professor of Physiology at the University of Analytical Services, Minneapolis Medical Re- Regensburg, Germany. Subsequently, he contin- search Foundation since September of 2002. Dr ued his training in internal medicine and nephrol- Foley also is a Co-Editor of the American Jour- ogy at the Charité, Humboldt University in Ber- nal of Kidney Diseases. His major interest is in lin, where he was appointed Associate Professor outcomes research, especially the interplay of of Nephrology in 2000. His major scientific cardiovascular and renal disease. Dr Foley is interests are in the molecular mechanisms and active in anemia correction trials, as well as in physiological/pathophysiological relevance of the USRDS Cardiovascular Special Study Cen- oxygen sensing and the management of anemia. ter. Dr Foley has received research funds, grants Professor Eckardt is Subject Editor of Nephrol- or contracts from Ortho Biotech, Amgen and ogy, Dialysis and Transplantation and serves on Roche. the editorial board of several other journals. He Sana Ghaddar, PhD, RD, is an Assistant contributed to the development of the EBPGs for Professor at the American University of Beirut, Anemia Management and is a member of the Lebanon. She has over 10 years experience in the executive committee of KDIGO. Dr Eckardt has renal and clinical dietetics field. She was one of received research funds, grants or contracts from the renal dietitians at the Peninsula Nephrology the German Research Organization. Addition- Inc., in San Mateo, CA, currently a division of ally, Dr Eckardt has received lecture and consul- Satellite Healthcare, and has served as a Prin- tant fees from Amgen, Roche, Johnson & John- ciple Investigator for two research studies that son, and Affymax. He is also associated with examined the ability of heme-iron-polypeptide CREATE and TREAT studies. to sustain response to Recombinant Erythropoi- Steven Fishbane, MD, currently is Chief of etin in both hemo and peritoneal dialysis pa- Nephrology and Associate Chair of the Depart- tients. She has presented these and other studies ment of Medicine at Winthrop-University Hospi- she has been involved in at many national confer- tal (WUH) in Mineola, NY, as well as Professor ences, including National Kidney Foundation, of Medicine at SUNY Stony Brook School of American Dietetics Association and Gerontologi- Medicine. He is the Medical Director of WUH cal Society of America. Dr Ghaddar has received Dialysis Network, which includes 4 outpatient research funds, grants or contracts from the dialysis units and 3 hospital units. Dr Fishbane American University of Beirut Research Board. serves as the Chairman of the Long Island Health Additionally, Dr Ghaddar is associated with the Network Quality Council; Chairman of the De- Kidney Nutrition Education & Life Improve- partment of Medicine Quality Improvement Pro- ment (K/NELI) study funded by the American gram, WUH; Chairman of Clinical Guidelines Univesity of Beirut. Committee, WUH; Co-Chairman of WUH Patient John S. Gill, MD, MS, obtained his MD from Care Committee; and Associate Chairman of the the University of British Columbia (UBC) in Department of Medicine, WUH. Dr Fishbane is a 1995. He completed his internal medicine train- member of the Network 2 Medical Review Board. ing at the University of Western Ontario in 1998 Dr Fishbane has received research funds, grants and his nephrology training in 2000 at UBC. He or contracts from Shire Inc., Amgen Inc., Abbott then completed his transplantation training at Labs Inc., Roche Inc., and Watson Inc. Addition- TuftsÐNew England Medical Center in Boston ally, Dr Fishbane is associated with Ortho Bio- and obtained in Masters in Clinical Care Re- tech and Watson Pharmaceuticals. search from Tufts in 2002. Dr Gill currently is Robert N. Foley, MD, was born in Ireland assistant professor of medicine in the division of and received his undergraduate MD from Univer- Nephrology at UBC and has a cross appointment sity College Cork. He completed Internal Medi- at TuftÐNew England Medical Center. Dr Gill’s cine training in Cork, later moving to Saint research interests focus on clinical outcomes in John’s, Newfoundland, Canada, where he com- kidney transplant recipients. He is the principal pleted a residency in nephrology, as well as a investigator and co-investigator on current Cana- Masters in Clinical Epidemiology. From 1999 to dian Institutes of Health Research (CIHR), Kid- S128 WORK GROUP BIOGRAPHIES ney Foundation, and Michael Smith Funded stud- lines Working Group, the KDIGO Board of Di- ies. Dr Gill is chair of the Canadian Society of rectors, the Global Scientific Advisory Board for Transplantation Work Group for Pan-Canadian PRCA, and the Council of the European Renal database development, member of the Cana- Association. He is frequently invited to lecture dian Organ Replacement Register Advisory both nationally and internationally on this topic, Board, and member of a number of NKF and he has co-authored the section on renal Committees. Dr Gill received research funds, anaemia for the last 2 editions of the Oxford grants or contracts from the Michael Smith Textbook of Clinical Nephrology and the current Foundation for Health Research, the Canadian edition of Comprehensive Clinical Nephrology.He Institute for Health Research, Wyeth Pharma- also is Subject Editor for Nephrology Dialysis ceuticals, and Roche Pharmaceuticals. Transplantation. Dr Macdougall has received Kathy Jabs, MD, is a Pediatric Nephrologist research funds, grants or contracts from Amgen, who trained at Babies Hospital, NY, and Chil- Hoffman La Roche, Affymax, Vifor, and John- dren’s Hospital, Boston. She has been a faculty son & Johnson. Additionally, Dr Macdougall is member at Children’s in Boston (1988 to 1996) associated with the CERA (Phase III), OPUS, and was the Director of Dialysis and Renal Hematide (Phase II), TREAT (Phase IV) studies Transplantation at Children’s Hospital of Phila- sponsored by Roche, Amgen, Affymax, and Am- delphia (1996 to 2000). She currently is the gen, respectively. Director of Pediatric Nephrology at Vanderbilt Patricia Bargo McCarley, RN, MSN, NP, is Children’s Hospital and an Associate Professor a nephrology nurse practitioner at Diablo Ne- of Pediatrics at Vanderbilt University School of phrology Medical Group in Walnut Creek, Cali- Medicine, Nashville, TN. Dr Jabs has had a fornia. Ms. McCarley received her BSN and long-standing interest in the care of children with MSN from Vanderbilt University. She is active in CKD. Dr Jabs has received research funds, grants ANNA having served on local, regional and or contracts from King Pharmaceuticals and national committees. She is currently a member Watson Pharmaceuticals. Additionally, Dr Jabs is of the Nephrology Nursing Journal Board. Ms. associated with the CKid and FSGS studies spon- McCarley has authored many publications includ- sored by the NIH. ing most recently chapters in the 2005 ANNA Francesco Locatelli, MD, is Scientific Direc- Nephrology Nursing Standards of Practice and tor and Head of the Department of Nephrology Guidelines for Care and the Contemporary Ne- and Dialysis of A. Manzoni Hospital, Lecco, and phrology Nursing: Principles and Practice (2nd Postgraduate Professor of Nephrology at Brescia Ed.). Dr McCarley has received research funds, and Milan Universities in Italy. He is past-President grants or contracts from Amgen, Ortho Biotech of the European Renal AssociationÐEuropean Di- and Renal Care Group. alysis and Transplantation Association, Italian Soci- Hans H. Messner, MD, is the Director of the ety of Nephrology, and International Society of Blood and Marrow Program at the Princess Mar- Blood Purification and was Chairman of the garet Hospital. He has a long-standing interest in EBPGs. exploring the role of transplantation in the man- Iain C. Macdougall, MD, is a combined agement of patients with various hematopoietic medical and science graduate of Glasgow Univer- diseases. In the laboratory, he has used culture sity, Scotland, from which he was awarded a First techniques to examine normal and malignant Class Honours BSc in Pharmacology in 1980. For hematopoietic progenitor cell populations. In the the last 10 years, Dr Macdougall has been Consul- recent past, he has been involved in the establish- tant Nephrologist and Honorary Senior Lecturer at ment of a cell-processing facility at the Princess King’s College Hospital in London. He has devel- Margaret Hospital that meets the standards of oped both a clinical and a basic science research good manufacturing practices (GMP). This labo- interest in factors affecting responsiveness to ratory will facilitate clinical skill of cell separa- erythropoietic agents. He has served on the Work- tion procedures, cell expansions, and vector trans- ing Parties responsible for both (1999 and 2004) actions into relevant cell populations. This EBPGs on Renal Anaemia Management, and he technology also will be instrumental in faciliting currently is a member of the US Anemia Guide- explorations of regenerative capability of stem WORK GROUP BIOGRAPHIES S129 cells for other organ systems. Dr Messner has more than 25 invited lectures each year and received research funds, grants or contracts from developed and chairs an annual second-year ne- Ortho Biotech. phrology fellows preceptorship program, serving Allen R. Nissenson, MD, FACP, is Professor more than 65 Fellows from throughout the United of Medicine and Director of the Dialysis Pro- States. Dr Nissenson has received research funds, gram at the David Geffen School of Medicine at grants or contracts from Amgen, Ortho Biotech, UCLA, where he has developed a comprehen- Roche, Watson and ARL. sive dialysis program. He is President of the Gregorio T. Obrador, MD, MPH, is Profes- National Anemia Action Council (NAAC) and sor of Medicine and Dean at the Universidad recently chaired a Chancellor’s committee at Panamericana School of Medicine in Mexico UCLA on Financial Conflicts of Interest in Clini- City. He also serves as Adjunct Staff at the cal Research. He currently is serving on a Univer- Division of Nephrology of the TuftsÐNew En- sity of California Task Force on Institutional gland Medical Center and Assistant Professor of Conflicts of Interest in Research. Dr Nissenson is Medicine at the Tufts University School of Medi- Chair of the Faculty Executive Council for the cine in Boston. While doing a clinical research David Geffen School of Medicine at UCLA. He fellowship at the TuftsÐNew England Medical has served as Chair of the Southern California Center and a Master of Public Health at Harvard ESRD Network during its organizational years in University, he began a line of investigation in the the early 1980s and recently was elected as area of CKD. Through several publications, he President-elect. He is Chair of the Medical Re- and others showed that the preÐESRD manage- view Board. Dr Nissenson currently is consult- ment of patients with CKD is suboptimal, and ing for RMS Disease Management, Inc and for that this is an important factor for the high Philtre, Ltd, an organization developing new morbidity and mortality observed in these pa- renal replacement therapies based on the applica- tients. A particular area of interest has been tion of nanotechnology to this field. Dr Nissen- anemia management before the initiation of dialy- son served as a Robert Wood Johnson Health sis therapy. By using population registry data, he Policy Fellow of the Institute of Medicine in and his colleagues have reported trends in ane- 1994 to 1995. He is Immediate Past President of mia and iron management. Dr Obrador has served the Renal Physicians Association and has served as reviewer for several journals, including Kid- as a member of the Advisory Group overseeing ney International, the Journal of the American the entire NKF-DOQI, as well as serving as a Society of Nephrology, and the American Jour- member of the anemia Work Group. Dr Nissen- nal of Kidney Diseases. He also has been a son’s major research interests focus on the qual- member of the Advisory Board of the NKF- ity of care for patients with CKD. His research KDOQI. Dr Obrador has received research funds, has included extensive clinical trials of new grants or contracts from Amgen, and is associ- devices and drugs related to renal disease. Dr ated with the TREAT study. Nissenson is co-principal investigator on a John C. Stivelman, MD, is Chief Medical recently obtained National Institutes of Health Officer of the Northwest Kidney Centers and Center Grant looking at issues of disparities in Associate Professor of Medicine in the Division health care delivery for patients with CKD. He of Nephrology, Department of Medicine, at the is the author of 2 dialysis textbooks, both in University of Washington School of Medicine in their fourth editions, and was the founding Seattle. Dr Stivelman obtained his MD degree Editor-in-Chief of Advances in Renal Replace- from the University of Pennsylvania, completed ment Therapy, an official journal of the NKF. He his residency in Internal Medicine at Harbor- currently is Editor-in-Chief of Hemodialysis In- UCLA Medical Center, and nephrology training ternational, the official journal of the Interna- at Brigham and Women’s Hospital. Dr Stivelman tional Society for Hemodialysis. He has more has been involved in investigative efforts to than 340 publications in the field of nephrology, optimize hematopoietic therapy for dialysis pa- dialysis, anemia management, and health care tients since the phase III recombinant erythropoi- delivery and policy. Among his numerous honors etin trials in 1986. His major interests and litera- is the President’s Award of the NKF. He delivers ture contributions center on iron utilization, S130 WORK GROUP BIOGRAPHIES mechanisms of resistance of erythropoietin the major nephrology journals. Dr Van Wyck therapy, improved dialytic survival in disadvan- served on the original KDOQI Anemia Work taged populations, and the interaction of regula- Group. He assumed Chair responsibilities in 2002. tory issues with optimization of care. Dr Stivel- Frequently invited to speak, Dr Van Wyck has man has served as the Chair of the Network 6 lectured on the molecular and cellular control Medical Review Board and a member of the of erythropoiesis and iron homeostasis, diag- Forum of ESRD Networks Board of Directors. nostic and treatment issues in anemia and iron He currently serves as medical director of one of management, protocol development in the treat- Northwest Kidney Centers’ free-standing facili- ment of dialysis-associated anemia, and new ties and as a member of the Boards of Directors approaches to iron and erythropoietin replace- of the Renal Physicians’ Association and the ment therapy. Dr Van Wyck has received re- Northwest Renal Network (Network 16). Dr search funds, grants or contracts from Amgen Stivelman has received research funds, grants or Inc., American Regent Inc., Gambro Health- contracts from Watson, Amgen and Shire. care and Shire Pharmaceuticals. David B. Van Wyck, MD (Work Group Co- Colin T. White, MD, is a pediatric nephrolo- Chair), is Professor of Medicine and Surgery at gist at BC Children’s Hospital in Vancouver and the University of Arizona College of Medicine in clinical assistant professor at the University of Tucson. After completing his undergraduate stud- British Columbia in Canada. He completed medi- ies at Washington University, St Louis, Dr Van cal school in Ottawa and Pediatrics in London, Wyck earned his MD at the University of Ari- Ontario. There, he did 3 years of pediatric ne- zona College of Medicine. There, he undertook a phrology training before moving to Vancouver to research fellowship in Surgical Biology and com- complete 3 more years. He has been on staff as a pleted his residency in Internal Medicine and Pediatric Nephrologist since 2003 and currently fellowship in Nephrology. Dr Van Wyck has is the Director of Dialysis at BC Children’s written or contributed to books, book chapters, Hospital. He has a number of research interests, articles, and abstracts on basic iron metabolism including medical education, optimizing dialysis and reticuloendothelial function and on clinical care in children, estimation of GFR, and CKD aspects of iron and anemia in patients with CKD. and its complications. Dr White’s interest in On the subject of anemia and kidney disease, he anemia management is geared toward children. pursues research, provides consultation to indus- He is presently completing a Masters degree in try including American Regent, Amgen, and Medical Education. Dr White is associated with Gambro Healthcare, and reviews manuscripts for the CKid study and various NAPRTC protocols. ACKNOWLEDGMENTS

The Work Group appreciates the careful review Donna Mapes, DNSc, MS, Luz Maria Munoz, of the draft guidelines and suggestions for im- Aletha Matsis, BSN, RN, CNN, Seiichi Matsuo, provement by external reviewers. Each comment MD, Marva M. Moxey-Mims, MD, Andrew S. was carefully considered and, whenever pos- Narva, MD, Suzanne Norby, MD, Nancy J. Pel- sible, suggestions for change were incorporated frey, MSN, NP, Allen G. Peplinski, Suanna into the ﬁnal report. As a result, the KDOQI Petroff, Luana Pillon, MD, Francesco Pizzarelli, Anemia in Chronic Kidney Disease guidelines is MD, Karthik Ranganna, MD, Pawan Rao, MD, the product of the Work Group, the Evidence Anton C. Schoolwerth, MD, FAHA, Douglas Review Team, the NKF, and all those who con- Schram, MD, Stephen Seliger, MD, Surendralal tributed their effort to improve the Guidelines. Shah, MD, Amita Sharma, MD, Bhupinder Singh, The following individuals provided written MD, Leah Smith, Stephanie J. Stewart, MSW, review of the draft guidelines: Nihal Younis Rachel L. Sturdivant, MD, David Tovbin, MD, Abosaif, MD, Harith Aljebory, MD, Beth Ann Edgar V. Lerma, MD, FACP, FASN, Lynne S. Avanzado, BSN, RN, CNN, George R. Bailie, Weiss, Cheryl Weller, LCSW, Nanette Wenger, PharmD, PhD, Beth Bandor McCarthy, RN, Vi- MD, Joseph W. Eschbach, MD, David C. nod K. Bansal, MD, Sally Burrows-Hudson, MS, RN, Carlos A. Caramelo, Dennis Cotter, MSE, Wheeler, MD, FRCP, Holly Whitcombe, MBA, Daniel W. Coyne, MD, Robert E. Cronin, Veta CFRE,Yu Yang, Karen Yeates, MD, MPH. Cumbaa, RN, BA, CCA, Mike Cunningham, Participation in the review does not necessar- Neval Duman, MD, Earl J. Dunnigan, MD, FACP, ily constitute endorsement of the content of the Briggett C. Ford, Cheryl Gilmartin, PharmD, report by the individuals or the organization or Marc Giovannini, Takanobu Imada, Atul V. In- institution they represent. gale, MD, Mark Kasselik, MD, Pam Kimball, The National Kidney Foundation, as well as MD, Janice Knouff, RN, BS, Chn, Craig B. the Work Group, would like to recognize the Langman, MD, David J. Leehey, MD, Nathan W. support of Amgen for the development of the Levin, MD, FACP,Yao-Lung Liu, MD, Francesco Guidelines. The National Kidney Foundation is Locatelli, MD, Gerard M. London, MD, Victor proud to partner with Amgen on this important Lorenzo, MD, Samir G. Mallat, MD, FASN, initiative.

American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: p S131 S131 REFERENCES

1. National Kidney Foundation: DOQI Clinical Practice 19. McGonigle RJ, Wallin JD, Shadduck RK, Fisher JW: Guidelines for the Treatment of Anemia of Chronic Renal Erythropoietin deficiency and inhibition of erythropoiesis in Failure. Am J Kidney Dis 30:S192-S240, 1997 (suppl 3) renal insufficiency. Kidney Int 25:437-444, 1984 2. National Kidney Foundation: K/DOQI Clinical Practice 20. Fisher JW: Mechanism of the anemia of chronic renal Guidelines for Anemia of Chronic Kidney Disease: Update failure. Nephron 25:106-111, 1980 2000. Am J Kidney Dis 37:S182-S238, 2001 (suppl 1) 21. Erslev AJ, Besarab A: The rate and control of baseline 3. National Kidney Foundation: K/DOQI Clinical Prac- red cell production in hematologically stable patients with tice Guidelines on Hypertension and Antihypertensive Agents uremia. J Lab Clin Med 126:283-286, 1995 in Chronic Kidney Disease. Am J Kidney Dis 43:S1-S290, 22. Besarab A, Levin A: Defining a renal anemia manage- 2004 (suppl 1) ment period. Am J Kidney Dis 36:S13-S23, 2000. (suppl 1) 4. National Kidney Foundation: K/DOQI Clinical Prac- 23. Chandra M, Miller ME, Garcia JF, Mossey RT, tice Guidelines for Chronic Kidney Disease: Evaluation, McVicar M: Serum immunoreactive erythropoietin levels in classification, and stratification. Am J Kidney Dis 39:S1- patients with polycystic kidney disease as compared with S266, 2002 (suppl 1) other hemodialysis patients. Nephron 39:26-29, 1985 5. US Renal Data System: USRDS 2000 Annual Data 24. Eschbach JW Jr, Funk D, Adamson J, Kuhn I, Scribner Report. Bethesda, MD, National Institutes of Health, National BH, Finch CA: Erythropoiesis in patients with renal failure Institute of Diabetes and Digestive and Kidney Diseases, undergoing chronic dialysis. N Engl J Med 276:653-658, 2000 1967 6. National Kidney Foundation: K/DOQI Clinical Prac- 25. Hsu CY, Bates DW, Kuperman GJ, Curhan GC: tice Guidelines for Management of Dyslipidemias in Pa- Relationship between hematocrit and renal function in men tients With Kidney Disease. Am J Kidney Dis 41:S1-S91, and women. Kidney Int 59:725-731, 2001 2003 (suppl 3) 26. Hsu CY, McCulloch CE, Curhan GC: Epidemiology 7. National Kidney Foundation: K/DOQI Clinical Practice of anemia associated with chronic renal insufficiency among Guidelines for Bone Metabolism and Disease in Chronic adults in the United States: Results from the Third National Kidney Disease. Am J Kidney Dis 42:S1-S201, 2003 (suppl 3) Health and Nutrition Examination Survey. J Am Soc Neph- 8. National Kidney Foundation: K/DOQI Clinical Prac- rol 13:504-510, 2002 tice Guidelines for Nutrition in Chronic Renal Failure. Am J 27. Astor BC, Muntner P, Levin A, Eustace JA, Coresh J: Kidney Dis 35:S1-S140, 2000 (suppl 1) Association of kidney function with anemia: The Third 9. National Kidney Foundation: K/DOQI Clinical Prac- National Health and Nutrition Examination Survey (1988- tice Guidelines for Cardiovascular Disease in Dialysis 1994). Arch Intern Med 162:1401-1408, 2002 Patients. Am J Kidney Dis 45:S1-S153, 2005 (suppl 3) 28. US Renal Data System: USRDS 2004 Annual Data 10. Gilbertson DT, Liu J, Xue JL, et al: Projecting the Report. Bethesda, MD, The National Institutes of Health, number of patients with end-stage renal disease in the National Institute of Diabetes and Digestive and Kidney United States to the year 2015. J Am Soc Nephrol 16:3736- 3741, 2005 Diseases, 2004 11. Donnelly S: Why is erythropoietin made in the 29. McClellan W, Aronoff SL, Bolton WK, et al: The kidney? The kidney functions as a critmeter. Am J Kidney prevalence of anemia in patients with chronic kidney Dis 38:415-425, 2001 disease. Curr Med Res Opin 20:1501-1510, 2004 12. Metzen E, Ratcliffe PJ: HIF hydroxylation and 30. Levin A, Thompson CR, Ethier J, et al: Left ventricu- cellular oxygen sensing. Biol Chem 385:223-230, 2004 lar mass index increase in early renal disease: Impact of 13. Koury MJ, Ponka P: New insights into erythropoiesis: decline in hemoglobin. Am J Kidney Dis 34:125-134, 1999 31. Obrador GT, Ruthazer R, Arora P, Kausz AT, Pereira The roles of folate, vitamin B12, and iron. Annu Rev Nutr 24:105-131, 2004 BJ: Prevalence of and factors associated with suboptimal 14. De Maria R, Testa U, Luchetti L, et al: Apoptotic role care before initiation of dialysis in the United States. J Am of Fas/Fas ligand system in the regulation of erythropoiesis. Soc Nephrol 10:1793-1800, 1999 Blood 93:796-803, 1999 32. Guralnik JM, Eisenstaedt RS, Ferrucci L, Klein HG, 15. Andrews NC: Anemia of inflammation: The cytokine- Woodman RC: Prevalence of anemia in persons 65 years and hepcidin link. J Clin Invest 113:1251-1253, 2004 older in the United States: Evidence for a high rate of 16. Locatelli F, Aljama P, Barany P, et al: Revised unexplained anemia. Blood 104:2263-2268, 2004 European Best Practice Guidelines for the management of 33. Ishimura E, Nishizawa Y, Okuno S, et al: Diabetes anaemia in patients with chronic renal failure. Nephrol Dial mellitus increases the severity of anemia in non-dialyzed Transplant 19:Sii1-Sii47, 2004 (suppl 2) patients with renal failure. J Nephrol 11:83-86, 1998 17. Eschbach JW, Adamson JW: Anemia of end-stage 34. Bosman DR, Winkler AS, Marsden JT, Macdougall renal disease (ESRD). Kidney Int 28:1-5, 1985 IC, Watkins PJ: Anemia with erythropoietin deficiency 18. Radtke HW, Claussner A, Erbes PM, Scheuermann occurs early in diabetic nephropathy. Diabetes Care 24:495- EH, Schoeppe W, Koch KM: Serum erythropoietin concen- 499, 2001 tration in chronic renal failure: Relationship to degree of 35. Thomas MC, MacIsaac RJ, Tsalamandris C, Power D, anemia and excretory renal function. Blood 54:877-884, Jerums G: Unrecognized anemia in patients with diabetes: A 1979 cross-sectional survey. Diabetes Care 26:1164-1169, 2003

S132 American Journal of Kidney Diseases, Vol 47, No 5, Suppl 3 (May), 2006: pp S132-S145 REFERENCES S133

36. Thomas MC, MacIsaac RJ, Tsalamandris C, et al: The 52. Beall CM, Goldstein MC: Hemoglobin concentration burden of anaemia in type 2 diabetes and the role of of pastoral nomads permanently resident at 4,850-5,450 nephropathy: A cross-sectional audit. Nephrol Dial Trans- meters in Tibet. Am J Phys Anthropol 73:433-438, 1987 plant 19:1792-1797, 2004 53. Richalet JP, Souberbielle JC, Antezana AM, et al: 37. El-Achkar TM, Ohmit SE, McCullough PA, et al: Control of erythropoiesis in humans during prolonged Higher prevalence of anemia with diabetes mellitus in exposure to the altitude of 6,542 m. Am J Physiol 266:R756- moderate kidney insufficiency: The Kidney Early Evalua- R764, 1994 tion Program. Kidney Int 67:1483-1488, 2005 54. Whitehead TP, Robinson D, Hale AC, Bailey AR: 38. Roger SD, McMahon LP, Clarkson A, et al: Effects of Clinical chemistry and haematology: Adult reference values. early and late intervention with epoetin alpha on left ventricular London, UK, BUPA Medical Research, 1994 mass among patients with chronic kidney disease (stage 3 or 4): 55. National Kidney Foundation: Kidney Early Evalua- Results of a randomized clinical trial. J Am Soc Nephrol tion Program. Am J Kidney Dis 45:S1-S135, 2005 (suppl 2) 15:148-156, 2004 56. Berkahn L, Keating A: Hematopoiesis in the elderly. 39. Levin A, Djurdjev O, Thompson C, et al: Canadian Hematology 9:159-163, 2004 randomized trial of hemoglobin maintenance to prevent or 57. Svanberg B, Arvidsson B, Norrby A, Rybo G, Solvell delay left ventricular mass growth in patients with CKD. L: Absorption of supplemental iron during pregnancy—A Am J Kidney Dis 46:799-811, 2005 longitudinal study with repeated bone-marrow studies and 40. Hollowell JG, van Assendelft OW, Gunter EW, Lewis absorption measurements. Acta Obstet Gynecol Scand Suppl BG, Najjar M, Pfeiffer C: Hematological and iron-related 48:S87-S108, 1975 analytes—Reference data for persons aged 1 year and over: 58. Puolakka J, Janne O, Pakarinen A, Jarvinen PA, Vihko United States, 1988-94. Vital Health Stat 11:1-156, 2005 R: Serum ferritin as a measure of iron stores during and after 41. Centers for Disease Control and Prevention: Recom- normal pregnancy with and without iron supplements. Acta mendations to prevent and control iron deficiency in the Obstet Gynecol Scand Suppl 95:S43-S51, 1980 United States. MMWR Recomm Rep 47:1-29, 1998 59. Taylor DJ, Mallen C, McDougall N, Lind T: Effect of 42. World Health Organization: Iron Deficiency Anaemia: iron supplementation on serum ferritin levels during and Assessment, Prevention and Control. A Guide for Pro- after pregnancy. Br J Obstet Gynaecol 89:1011-1017, 1982 gramme Managers. Geneva, Switzerland, World Health 60. Meyers LD, Habicht JP, Johnson CL: Components of Organization, 2001 the difference in hemoglobin concentrations in blood be- 43. Levin A: How should anaemia be managed in tween black and white women in the United States. Am J pre-dialysis patients? Nephrol Dial Transplant 14:S66-S74, Epidemiol 109:539-549, 1979 1999 (suppl 2) 61. Cresanta JL, Croft JB, Webber LS, Nicklas TA, 44. Xue JL, St Peter WL, Ebben JP, Everson SE, Collins Berenson GS: Racial difference in hemoglobin concentra- AJ: Anemia treatment in the pre-ESRD period and associ- tion of young adults. Prev Med 16:659-669, 1987 ated mortality in elderly patients. Am J Kidney Dis 40:1153- 62. Pan WH, Habicht JP: The non-iron-deficiency-related 1161, 2002 difference in hemoglobin concentration distribution between 45. Parfrey PS, Foley RN, Harnett JD, Kent GM, Murray blacks and whites and between men and women. Am J DC, Barre PE: Outcome and risk factors for left ventricular Epidemiol 134:1410-1416, 1991 disorders in chronic uraemia. Nephrol Dial Transplant 63. Jackson RT: Separate hemoglobin standards for blacks 11:1277-1285, 1996 and whites: A critical review of the case for separate and 46. Levin A, Foley RN: Cardiovascular disease in chronic unequal hemoglobin standards. Med Hypotheses 32:181- renal insufficiency. Am J Kidney Dis 36:S24-S30, 2000 189, 1990 (suppl 1) 64. Johnson-Spear MA, Yip R: Hemoglobin difference 47. Kausz AT, Obrador GT, Pereira BJ: Anemia manage- between black and white women with comparable iron ment in patients with chronic renal insufficiency. Am J status: Justification for race-specific anemia criteria. Am J Kidney Dis 36:S39-S51, 2000 (suppl 1) Clin Nutr 60:117-121, 1994 48. Locatelli F, Conte F, Marcelli D: The impact of 65. Tirlapur VG, Gicheru K, Charalambous BM, Evans haematocrit levels and erythropoietin treatment on overall PJ, Mir MA: Packed cell volume, haemoglobin, and oxygen and cardiovascular mortality and morbidity—The experi- saturation changes in healthy smokers and non-smokers. ence of the Lombardy Dialysis Registry. Nephrol Dial Thorax 38:785-787, 1983 Transplant 13:1642-1644, 1998 66. Taralov Z, Koumtchev E, Lyutakova Z: Erythrocyte 49. Canadian Erythropoietin Study Group: Association ferritin levels in chronic renal failure patients. Folia Med between recombinant human erythropoietin and quality of (Plovdiv) 40:65-70, 1998 life and exercise capacity of patients receiving haemodialy- 67. Clyne N, Jogestrand T, Lins LE, Pehrsson SK: sis. BMJ 300:573-578, 1990 Progressive decline in renal function induces a gradual 50. Ritz E, Eisenhardt A: Early epoetin treatment in decrease in total hemoglobin and exercise capacity. Nephron patients with renal insufficiency. Nephrol Dial Transplant 67:322-326, 1994 15:S40-S44, 2000 (suppl 3) 68. Nankivell BJ, Allen RD, O’Connell PJ, Chapman JR: 51. Pugh LG: Muscular exercise on Mount Everest. Erythrocytosis after renal transplantation: Risk factors and J Physiol 141:233-261, 1958 relationship with GFR. Clin Transplant 9:375-382, 1995 S134 REFERENCES

69. de Klerk G, Wilmink JM, Rosengarten PC, Vet RJ, 85. Fehr T, Ammann P, Garzoni D, et al: Interpretation of Goudsmit R: Serum erythropoietin (ESF) titers in anemia of erythropoietin levels in patients with various degrees of chronic renal failure. J Lab Clin Med 100:720-734, 1982 renal insufficiency and anemia. Kidney Int 66:1206-1211, 70. Urabe A, Saito T, Fukamachi H, Kubota M, Takaku F: 2004 Serum erythropoietin titers in the anemia of chronic renal 86. Morris MW, Davey FR: Basic examination of blood, failure and other hematological states. Int J Cell Cloning in Henry JB (ed): Clinical Diagnosis and Management by 5:202-208, 1987 Laboratory Methods. Philadelphia, PA, WB Saunders, 1996, 71. Silverberg DS, Iaina A, Peer G, et al: Intravenous iron pp 549-593 supplementation for the treatment of the anemia of moderate 87. Brittin GM, Brecher G, Johnson CA, Elashoff RM: to severe chronic renal failure patients not receiving dialysis. Stability of blood in commonly used anticoagulants. Use of Am J Kidney Dis 27:234-238, 1996 refrigerated blood for quality control of the Coulter Counter 72. Lin JL, Kou MT, Leu ML: Effect of long-term Model S. Am J Clin Pathol 52:690-694, 1969 low-dose aluminum-containing agents on hemoglobin syn- 88. Holt JT, DeWandler MJ, Arvan DA: Spurious eleva- thesis in patients with chronic renal insufficiency. Nephron tion of the electronically determined mean corpuscular 74:33-38, 1996 volume and hematocrit caused by hyperglycemia. Am J Clin 73. Clyne N, Jogestrand T: Effect of erythropoietin Pathol 77:561-567, 1982 treatment on physical exercise capacity and on renal func- 89. van Duijnhoven HL, Treskes M: Marked interference tion in predialytic uremic patients. Nephron 60:390-396, of hyperglycemia in measurements of mean (red) cell 1992 volume by Technicon H analyzers. Clin Chem 42:76-80, 74. Portoles J, Torralbo A, Martin P, Rodrigo J, Herrero 1996 JA, Barrientos A: Cardiovascular effects of recombinant 90. Paterakis GS, Laoutaris NP, Alexia SV, et al: The human erythropoietin in predialysis patients. Am J Kidney effect of red cell shape on the measurement of red cell Dis 29:541-548, 1997 volume. A proposed method for the comparative assessment 75. Dimitrakov D, Kumchev E, Tllkian E: Study of the of this effect among various haematology analysers. Clin effect of recombinant erythropoietin on renal anaemia in Lab Haematol 16:235-245, 1994 predialysis patients with chronic renal failure. Folia Med 91. Fraser CG, Wilkinson SP, Neville RG, Knox JD, King JF, MacWalter RS: Biologic variation of common hemato- (Plovdiv) 36:31-36, 1994 logic laboratory quantities in the elderly. Am J Clin Pathol 76. Klahr S, Breyer JA, Beck GJ, et al: Dietary protein 92:465-470, 1989 restriction, blood pressure control, and the progression of 92. Movilli E, Pertica N, Camerini C, et al: Predialysis polycystic kidney disease. Modification of Diet in Renal versus postdialysis hematocrit evaluation during erythropoi- Disease Study Group. J Am Soc Nephrol 5:2037-2047, 1995 etin therapy. Am J Kidney Dis 39:850-853, 2002 77. Howard AD, Moore J Jr, Welch PG, Gouge SF: 93. Bellizzi V, Minutolo R, Terracciano V, et al: Influence Analysis of the quantitative relationship between anemia of the cyclic variation of hydration status on hemoglobin and chronic renal failure. Am J Med Sci 297:309-313, 1989 levels in hemodialysis patients. Am J Kidney Dis 40:549- 78. Lim VS, Fangman J, Flanigan MJ, DeGowin RL, 555, 2002 Abels RT: Effect of recombinant human erythropoietin on 94. Hillman RS, Finch CA: Red Cell Manual (ed 7). renal function in humans. Kidney Int 37:131-136, 1990 Philadelphia, PA, Davis, 1996 79. Besarab A, Caro J, Jarrell B, et al: Effect of cyclosporine 95. Naets JP, Garcia JF, Tousaaint C, Buset M, Waks D: and delayed graft function on posttransplantation erythropoi- Radioimmunoassay of erythropoietin in chronic uraemia or esis. Transplantation 40:624-631, 1985 anephric patients. Scand J Haematol 37:390-394, 1986 80. Brod J, Hornych A: Effect of correction of anemia on 96. Ross RP, McCrea JB, Besarab A: Erythropoietin the glomerular filtration rate in chronic renal failure. Isr response to blood loss in hemodialysis patients in blunted J Med Sci 3:53-59, 1967 but preserved. ASAIO J 40:M880-M885, 1994 81. Roth D, Smith RD, Schulman G, et al: Effects of 97. Fishbane S, Lynn RI: The efficacy of iron dextran for recombinant human erythropoietin on renal function in the treatment of iron deficiency in hemodialysis patients. chronic renal failure predialysis patients. Am J Kidney Dis Clin Nephrol 44:238-240, 1995 24:777-784, 1994 98. Fernandez-Rodriguez AM, Guindeo-Casasus MC, 82. Kuriyama S, Tomonari H, Yoshida H, Hashimoto T, Molero-Labarta T, et al: Diagnosis of iron deficiency in Kawaguchi Y, Sakai O: Reversal of anemia by erythropoi- chronic renal failure. Am J Kidney Dis 34:508-513, 1999 etin therapy retards the progression of chronic renal failure, 99. Kalantar-Zadeh K, Hoffken B, Wunsch H, Fink H, especially in nondiabetic patients. Nephron 77:176-185, Kleiner M, Luft FC: Diagnosis of iron deficiency anemia in 1997 renal failure patients during the post-erythropoietin era. 83. Hayashi T, Suzuki A, Shoji T, et al: Cardiovascular Am J Kidney Dis 26:292-299, 1995 effect of normalizing the hematocrit level during erythropoi- 100. Annibale B, Lahner E, Chistolini A, et al: Endo- etin therapy in predialysis patients with chronic renal failure. scopic evaluation of the upper gastrointestinal tract is Am J Kidney Dis 35:250-256, 2000 worthwhile in premenopausal women with iron-deficiency 84. Schwartz AB, Prior JE, Mintz GS, Kim KE, Kahn SB: anaemia irrespective of menstrual flow. Scand J Gastroen- Cardiovascular hemodynamic effects of correction of ane- terol 38:239-245, 2003 mia of chronic renal failure with recombinant-human eryth- 101. Coban E, Timuragaoglu A, Meric M: Iron deficiency ropoietin. Transplant Proc 23:1827-1830, 1991 anemia in the elderly: Prevalence and endoscopic evaluation REFERENCES S135 of the gastrointestinal tract in outpatients. Acta Haematol 118. Furuland H, Linde T, Ahlmen J, Christensson A, 110:25-28, 2003 Strombom U, Danielson BG: A randomized controlled trial 102. Ioannou GN, Rockey DC, Bryson CL, Weiss NS: of haemoglobin normalization with epoetin alfa in pre- Iron deficiency and gastrointestinal malignancy: A popula- dialysis and dialysis patients. Nephrol Dial Transplant tion-based cohort study. Am J Med 113:276-280, 2002 18:353-361, 2003 103. Ioannou GN, Spector J, Scott K, Rockey DC: 119. Suzuki M, Hirasawa Y, Hirashima K, et al: Dose- Prospective evaluation of a clinical guideline for the diagno- finding, double-blind, clinical trial of recombinant human sis and management of iron deficiency anemia. Am J Med erythropoietin (Chugai) in Japanese patients with end-stage 113:281-287, 2002 renal disease. Research Group for Clinical Assessment of 104. Goddard AF, McIntyre AS, Scott BB: Guidelines for rhEPO. Contrib Nephrol 76:179-192; discussion 212-178, the management of iron deficiency anaemia. British Society 1989 of Gastroenterology. Gut 46:SIV1-SIV5, 2000 (suppl 3-4) 120. Furuland H, Linde T, Wikstrom B, Danielson BG: 105. Joosten E, Ghesquiere B, Linthoudt H, et al: Upper Reduced hemodialysis adequacy after hemoglobin normal- and lower gastrointestinal evaluation of elderly inpatients ization with epoetin. J Nephrol 18:80-85, 2005 who are iron deficient. Am J Med 107:24-29, 1999 121. McMahon LP, Mason K, Skinner SL, Burge CM, 106. Ho CH, Chau WK, Hsu HC, Gau JP, You JY, Chen Grigg LE, Becker GJ: Effects of haemoglobin normalization CC: Predictive risk factors and prevalence of malignancy in on quality of life and cardiovascular parameters in end-stage patients with iron deficiency anemia in Taiwan. Am J renal failure. Nephrol Dial Transplant 15:1425-1430, 2000 Hematol 78:108-112, 2005 122. McMahon LP, McKenna MJ, Sangkabutra T, et al: 107. Akmal M, Sawelson S, Karubian F, Gadallah M: The Physical performance and associated electrolyte changes prevalence and significance of occult blood loss in patients after haemoglobin normalization: A comparative study in with predialysis advanced chronic renal failure (CRF), or haemodialysis patients. Nephrol Dial Transplant 14:1182- receiving dialytic therapy. Clin Nephrol 42:198-202, 1994 1187, 1999 108. Besarab A, Bolton WK, Browne JK, et al: The effects 123. Nissenson AR, Korbet S, Faber M, et al: Multicenter of normal as compared with low hematocrit values in trial of erythropoietin in patients on peritoneal dialysis. J Am patients with cardiac disease who are receiving hemodialysis Soc Nephrol 5:1517-1529, 1995 and epoetin. N Engl J Med 339:584-590, 1998 124. Bahlmann J, Schoter KH, Scigalla P, et al: Morbidity 109. Parfrey PS, Foley RN, Wittreich BH, Sullivan DJ, and mortality in hemodialysis patients with and without Zagari MJ, Frei D: Double-blind comparison of full and erythropoietin treatment: A controlled study. Contrib Neph- partial anemia correction in incident hemodialysis patients rol 88:90-106, 1991 without symptomatic heart disease. J Am Soc Nephrol 125. Sikole A, Polenakovic M, Spirovska V, Polenakovic B, 16:2180-2189, 2005 Masin G: Analysis of heart morphology and function following 110. Foley RN, Parfrey PS, Morgan J, et al: Effect of erythropoietin treatment of anemic dialysis patients. Artif hemoglobin levels in hemodialysis patients with asymptom- Organs 17:977-984, 1993 atic cardiomyopathy. Kidney Int 58:1325-1335, 2000 126. Morris KP, Sharp J, Watson S, Coulthard MG: 111. Lacson E Jr, Ofsthun N, Lazarus JM: Effect of Non-cardiac benefits of human recombinant erythropoietin variability in anemia management on hemoglobin outcomes in end stage renal failure and anaemia. Arch Dis Child in ESRD. Am J Kidney Dis 41:111-124, 2003 69:580-586, 1993 112. Collins AJ, Brenner RM, Ofman JJ, et al: Epoetin 127. Abraham PA, Macres MG: Blood pressure in hemo- alfa use in patients with ESRD: An analysis of recent US dialysis patients during amelioration of anemia with erythro- prescribing patterns and hemoglobin outcomes. Am J Kid- poietin. J Am Soc Nephrol 2:927-936, 1991 ney Dis 46:481-488, 2005 128. Berns JS, Rudnick MR, Cohen RM, Bower JD, 113. Berns JS, Elzein H, Lynn RI, Fishbane S, Meisels IS, Wood BC: Effects of normal hematocrit on ambulatory Deoreo PB: Hemoglobin variability in epoetin-treated hemo- blood pressure in epoetin-treated hemodialysis patients with dialysis patients. Kidney Int 64:1514-1521, 2003 cardiac disease. Kidney Int 56:253-260, 1999 114. Fishbane S, Berns JS: Hemoglobin cycling in 129. Conlon PJ, Kovalik E, Schumm D, Minda S, Schwab hemodialysis patients treated with recombinant human eryth- SJ: Normalization of hematocrit in hemodialysis patients ropoietin. Kidney Int 68:1337-1343, 2005 with cardiac disease does not increase blood pressure. Ren 115. Eckardt KU: The CREATE trial—Building the evi- Fail 22:435-444, 2000 dence. Nephrol Dial Transplant 16:S16-S18, 2001 (suppl 2) 130. Levin A: The relationship of haemoglobin level and 115A. US National Institutes of Health: CHOIR Trial. 2005. survival: Direct or indirect effects? Nephrol Dial Transplant Available at: www.clinicaltrials.gov/ct/show/NCT00211120? 17:S8-S13, 2002 (suppl 5) orderϭ1. Accessed March 31, 2006 130A. Levin A, Djurdjev O, Thompson C, et al: Canadian 116. Man-Son-Hing M, Gage BF, Montgomery AA, et al: randomized trial of hemoglobin maintenance to prevent or Preference-based antithrombotic therapy in atrial fibrilla- delay left ventricular mass growth in patients with CKD. tion: Implications for clinical decision making. Med Decis Am J Kidney Dis 46:799-811, 2005 Making 25:548-559, 2005 131. Lim VS, DeGowin RL, Zavala D, et al: Recombinant 117. Canadian Erythropoietin Study Group: Effect of human erythropoietin treatment in pre-dialysis patients. A recombinant human erythropoietin therapy on blood pres- double-blind placebo-controlled trial. Ann Intern Med 110: sure in hemodialysis patients. Am J Nephrol 11:23-26, 1991 108-114, 1989 S136 REFERENCES

132. Revicki DA, Brown RE, Feeny DH, et al: Health- 146. Locatelli F, Olivares J, Walker R, et al: Novel related quality of life associated with recombinant human erythropoiesis stimulating protein for treatment of anemia in erythropoietin therapy for predialysis chronic renal disease chronic renal insufficiency. Kidney Int 60:741-747, 2001 patients. Am J Kidney Dis 25:548-554, 1995 147. Toto RD, Pichette V, Navarro J, et al: Darbepoetin 133. Kleinman KS, Schweitzer SU, Perdue ST, Bleifer alfa effectively treats anemia in patients with chronic kidney KH, Abels RI: The use of recombinant human erythropoietin disease with de novo every-other-week administration. Am J in the correction of anemia in predialysis patients and its Nephrol 24:453-460, 2004 effect on renal function: A double-blind, placebo-controlled 148. Besarab A, Amin N, Ahsan M, et al: Optimization of trial. Am J Kidney Dis 14:486-495, 1989 epoetin therapy with intravenous iron therapy in hemodialy- 134. Watson AJ, Gimenez LF, Cotton S, Walser M, Spivak sis patients. J Am Soc Nephrol 11:530-538, 2000 JL: Treatment of the anemia of chronic renal failure with 149. DeVita MV, Frumkin D, Mittal S, Kamran A, subcutaneous recombinant human erythropoietin. Am J Med Fishbane S, Michelis MF: Targeting higher ferritin concen- 89:432-435, 1990 trations with intravenous iron dextran lowers erythropoietin 135. Abraham PA, Opsahl JA, Rachael KM, Asinger R, requirement in hemodialysis patients. Clin Nephrol 60:335- Halstenson CE: Renal function during erythropoietin therapy 340, 2003 for anemia in predialysis chronic renal failure patients. Am J 150. Chang CH, Chang CC, Chiang SS: Reduction in Nephrol 10:128-136, 1990 erythropoietin doses by the use of chronic intravenous iron 136. Locatelli F, Baldamus CA, Villa G, Ganea A, Martin supplementation in iron-replete hemodialysis patients. Clin de Francisco AL: Once-weekly compared with three-times- Nephrol 57:136-141, 2002 weekly subcutaneous epoetin beta: Results from a random- 151. Fishbane S, Galgano C, Langley RC Jr, Canfield W, ized, multicenter, therapeutic-equivalence study. Am J Kid- Maesaka JK: Reticulocyte hemoglobin content in the evalu- ney Dis 40:119-125, 2002 ation of iron status of hemodialysis patients. Kidney Int 137. Locatelli F, Canaud B, Giacardy F, Martin-Malo A, 52:217-222, 1997 Baker N, Wilson J: Treatment of anaemia in dialysis patients 152. Fishbane S, Kowalski EA, Imbriano LJ, Maesaka with unit dosing of darbepoetin alfa at a reduced dose JK: The evaluation of iron status in hemodialysis patients. frequency relative to recombinant human erythropoietin J Am Soc Nephrol 7:2654-2657, 1996 (rHuEpo). Nephrol Dial Transplant 18:362-369, 2003 153. Chuang CL, Liu RS, Wei YH, Huang TP, Tarng DC: 138. Nissenson AR, Swan SK, Lindberg JS, et al: Early prediction of response to intravenous iron supplemen- Randomized, controlled trial of darbepoetin alfa for the tation by reticulocyte haemoglobin content and high- treatment of anemia in hemodialysis patients. Am J Kidney fluorescence reticulocyte count in haemodialysis patients. Dis 40:110-118, 2002 Nephrol Dial Transplant 18:370-377, 2003 139. Tolman C, Richardson D, Bartlett C, Will E: 154. Tessitore N, Solero GP, Lippi G, et al: The role of Structured conversion from thrice weekly to weekly erythro- iron status markers in predicting response to intravenous poietic regimens using a computerized decision-support iron in haemodialysis patients on maintenance erythropoi- system: A randomized clinical study. J Am Soc Nephrol etin. Nephrol Dial Transplant 16:1416-1423, 2001 16:1463-1470, 2005 155. Mittman N, Sreedhara R, Mushnick R, et al: Reticulo- 140. Kaufman JS, Reda DJ, Fye CL, et al: Subcutaneous cyte hemoglobin content predicts functional iron deficiency in compared with intravenous epoetin in patients receiving hemodialysis patients receiving rHuEPO. Am J Kidney Dis hemodialysis. Department of Veterans Affairs Cooperative 30:912-922, 1997 Study Group on Erythropoietin in Hemodialysis Patients. 156. Fishbane S, Frei GL, Maesaka J: Reduction in N Engl J Med 339:578-583, 1998 recombinant human erythropoietin doses by the use of 141. Vanrenterghem Y, Barany P, Mann JF, et al: Random- chronic intravenous iron supplementation. Am J Kidney Dis ized trial of darbepoetin alfa for treatment of renal anemia 26:41-46, 1995 at a reduced dose frequency compared with rHuEPO in 157. Fishbane S, Shapiro W, Dutka P, Valenzuela OF, dialysis patients. Kidney Int 62:2167-2175, 2002 Faubert J: A randomized trial of iron deficiency testing 142. Ling B, Walczyk M, Agarwal A, Carroll W, Liu W, strategies in hemodialysis patients. Kidney Int 60:2406- Brenner R: Darbepoetin alfa administered once monthly 2411, 2001 maintains hemoglobin concentrations in patients with chronic 158. Kaneko Y, Miyazaki S, Hirasawa Y, Gejyo F, Suzuki kidney disease. Clin Nephrol 63:327-334, 2005 M: Transferrin saturation versus reticulocyte hemoglobin 143. Jadoul M, Vanrenterghem Y, Foret M, Walker R, content for iron deficiency in Japanese hemodialysis pa- Gray SJ: Darbepoetin alfa administered once monthly tients. Kidney Int 63:1086-1093, 2003 maintains haemoglobin levels in stable dialysis patients. 159. Mitsuiki K, Harada A, Miyata Y: Assessment of iron Nephrol Dial Transplant 19:898-903, 2004 deficiency in chronic hemodialysis patients: Investigation of 144. Eschbach JW, Abdulhadi MH, Browne JK, et al: cutoff values for reticulocyte hemoglobin content. Clin Exp Recombinant human erythropoietin in anemic patients with Nephrol 7:52-57, 2003 end-stage renal disease. Results of a phase III multicenter 160. Besarab A, Kaiser JW, Frinak S: A study of clinical trial. Ann Intern Med 111:992-1000, 1989 parenteral iron regimens in hemodialysis patients. Am J 145. Eschbach JW, Kelly MR, Haley NR, Abels RI, Kidney Dis 34:21-28, 1999 Adamson JW: Treatment of the anemia of progressive renal 161. Lin JL, Chang MY, Tan DT, Leu ML: Short-term failure with recombinant human erythropoietin. N Engl small-dose intravenous iron trial to detect functional iron J Med 321:158-163, 1989 deficiency in dialysis patients. Am J Nephrol 21:91-97, 2001 REFERENCES S137

162. Low CL, Bailie GR, Eisele G: Sensitivity and 177. Fletes R, Lazarus JM, Gage J, Chertow GM: specificity of transferrin saturation and serum ferritin as Suspected iron dextran-related adverse drug events in hemo- markers of iron status after intravenous iron dextran in dialysis patients. Am J Kidney Dis 37:743-749, 2001 hemodialysis patients. Ren Fail 19:781-788, 1997 178. Hamstra RD, Block MH, Schocket AL: Intravenous 163. Canavese C, Bergamo D, Ciccone G, et al: Validation iron dextran in clinical medicine. JAMA 243:1726-1731, of serum ferritin values by magnetic susceptometry in 1980 predicting iron overload in dialysis patients. Kidney Int 179. Fishbane S, Ungureanu VD, Maesaka JK, Kaupke 65:1091-1098, 2004 CJ, Lim V, Wish J: The safety of intravenous iron dextran in 164. Kalantar-Zadeh K, Regidor DL, McAllister CJ, Mi- hemodialysis patients. Am J Kidney Dis 28:529-534, 1996 chael B, Warnock DG: Time-dependent associations between 180. Faich G, Strobos J: Sodium ferric gluconate complex iron and mortality in hemodialysis patients. J Am Soc Nephrol in sucrose: Safer intravenous iron therapy than iron dex- 16:3070-3080, 2005 trans. Am J Kidney Dis 33:464-470, 1999 165. Feldman HI, Joffe M, Robinson B, Knauss J, et al: 181. Bailie GR, Clark JA, Lane CE, Lane PL: Hypersensi- Administration of parenteral iron and mortality among tivity reactions and deaths associated with intravenous iron hemodialysis patients. J Am Soc Nephrol 15:1623-1632, preparations. Nephrol Dial Transplant 20:1443-1449, 2005 2004 182. Stang PE, Fox JL: Adverse drug events and the 166. Macdougall IC, Tucker B, Thompson J, Tomson CR, Freedom of Information Act: An apple in Eden. Ann Baker LR, Raine AE: A randomized controlled study of iron Pharmacother 26:238-243, 1992 supplementation in patients treated with erythropoietin. 183. Brimble KS, Rabbat CG, McKenna P, Lambert K, Kidney Int 50:1694-1699, 1996 Carlisle EJ: Protocolized anemia management with erythro- 167. Fudin R, Jaichenko J, Shostak A, Bennett M, Gotloib poietin in hemodialysis patients: A randomized controlled L: Correction of uremic iron deficiency anemia in hemodia- trial. J Am Soc Nephrol 14:2654-2661, 2003 lyzed patients: A prospective study. Nephron 79:299-305, 184. Cervelli MJ, Gray N, McDonald S, Gentgall MG, 1998 Disney AP: Randomized cross-over comparison of intrave- 168. Nissenson AR, Berns JS, Sakiewicz P, et al: Clinical nous and subcutaneous darbepoetin dosing efficiency in evaluation of heme iron polypeptide: Sustaining a response haemodialysis patients. Nephrology (Carlton) 10:129-135, 2005 to rHuEPO in hemodialysis patients. Am J Kidney Dis 185. Markowitz GS, Kahn GA, Feingold RE, Coco M, 42:325-330, 2003 Lynn RI: An evaluation of the effectiveness of oral iron 169. Van Wyck DB, Roppolo M, Martinez CO, Mazey therapy in hemodialysis patients receiving recombinant RM, McMurray S: A randomized, controlled trial comparing human erythropoietin. Clin Nephrol 48:34-40, 1997 IV iron sucrose to oral iron in anemic patients with 186. McCarthy JT, Regnier CE, Loebertmann CL, Bergs- nondialysis-dependent CKD. Kidney Int 68:2846-2856, 2005 tralh EJ: Adverse events in chronic hemodialysis patients 170. Aggarwal HK, Nand N, Singh S, Singh M, Hemant, receiving intravenous iron dextran—A comparison of two Kaushik G: Comparison of oral versus intravenous iron therapy products. Am J Nephrol 20:455-462, 2000 in predialysis patients of chronic renal failure receiving recom- 187. Aronoff GR, Bennett WM, Blumenthal S, et al: Iron binant human erythropoietin. J Assoc Physicians India 51:170- sucrose in hemodialysis patients: Safety of replacement and 174, 2003 maintenance regimens. Kidney Int 66:1193-1198, 2004 171. Stoves J, Inglis H, Newstead CG: A randomized 188. Blaustein DA, Schwenk MH, Chattopadhyay J, et al: study of oral vs intravenous iron supplementation in patients The safety and efficacy of an accelerated iron sucrose dosing with progressive renal insufficiency treated with erythropoi- regimen in patients with chronic kidney disease. Kidney Int etin. Nephrol Dial Transplant 16:967-974, 2001 Suppl 64:S72-S77, 2003 172. Charytan C, Qunibi W, Bailie GR: Comparison of 189. Charytan C, Schwenk MH, Al-Saloum MM, Spinow- intravenous iron sucrose to oral iron in the treatment of itz BS: Safety of iron sucrose in hemodialysis patients anemic patients with chronic kidney disease not on dialysis. intolerant to other parenteral iron products. Nephron Clin Nephron Clin Pract 100:c55-c62, 2005 Pract 96:c63-c66, 2004 173. Novey HS, Pahl M, Haydik I, Vaziri ND: Immuno- 190. Aronoff GR: Safety of intravenous iron in clinical logic studies of anaphylaxis to iron dextran in patients on practice: Implications for anemia management protocols. renal dialysis. Ann Allergy 72:224-228, 1994 J Am Soc Nephrol 15:S99-S106, 2004 (suppl 2) 174. Agarwal R, Vasavada N, Sachs NG, Chase S: 191. Hoppel C: The role of carnitine in normal and altered Oxidative stress and renal injury with intravenous iron in fatty acid metabolism. Am J Kidney Dis 41:S4-S12, 2003 patients with chronic kidney disease. Kidney Int 65:2279- (suppl 4) 2289, 2004 192. Bohmer T, Bergrem H, Eiklid K: Carnitine defi- 175. Michael B, Coyne DW, Fishbane S, et al: Sodium ciency induced during intermittent haemodialysis for renal ferric gluconate complex in hemodialysis patients: Adverse failure. Lancet 1:126-128, 1978 reactions compared to placebo and iron dextran. Kidney Int 193. Golper TA, Goral S, Becker BN, Langman CB: 61:1830-1839, 2002 L-Carnitine treatment of anemia. Am J Kidney Dis 41:S27- 176. Walters BA, Van Wyck DB: Benchmarking iron S34, 2003 (suppl 4) dextran sensitivity: reactions requiring resuscitative medica- 194. Berns JS, Mosenkis A: Pharmacologic adjuvants to tion in incident and prevalent patients. Nephrol Dial Trans- epoetin in the treatment of anemia in patients on hemodialy- plant 20:1438-1442, 2005 sis. Hemodial Int 9:7-22, 2005 S138 REFERENCES

195. Levocarnitine for Use in the Treatment of Carnitine 210. Lin CL, Hsu PY, Yang HY, Huang CC: Low dose Deficiency in ESRD Patients (program memorandum). intravenous ascorbic acid for erythropoietin-hyporesponsive Baltimore, MD, Department of Health & Human Services anemia in diabetic hemodialysis patients with iron overload. (DHHS), Centers for Medicare & Medicaid Services (CMS), Ren Fail 25:445-453, 2003 2003 211. Sezer S, Ozdemir FN, Yakupoglu U, Arat Z, Turan 196. Caruso U, Leone L, Cravotto E, Nava D: Effects of M, Haberal M: Intravenous ascorbic acid administration for L-carnitine on anemia in aged hemodialysis patients treated erythropoietin-hyporesponsive anemia in iron loaded hemo- with recombinant human erythropoietin: A pilot study. Dial dialysis patients. Artif Organs 26:366-370, 2002 Transplant 27:498-506, 1998 212. Tarng DC, Huang TP: A parallel, comparative study 197. Kletzmayr J, Sunder-Plassmann G, Horl WH: High of intravenous iron versus intravenous ascorbic acid for dose intravenous iron: A note of caution. Nephrol Dial erythropoietin-hyporesponsive anaemia in haemodialysis Transplant 17:962-965, 2002 patients with iron overload. Nephrol Dial Transplant 13:2867- 198. Labonia WD: L-Carnitine effects on anemia in 2872, 1998 hemodialyzed patients treated with erythropoietin. Am J 213. Tarng DC, Wei YH, Huang TP, Kuo BI, Yang WC: Kidney Dis 26:757-764, 1995 Intravenous ascorbic acid as an adjuvant therapy for recom- 199. Semeniuk J, Shalansky KF, Taylor N, Jastrzebski J, binant erythropoietin in hemodialysis patients with hyperfer- Cameron EC: Evaluation of the effect of intravenous ritinemia. Kidney Int 55:2477-2486, 1999 L-carnitine on quality of life in chronic hemodialysis pa- 214. Giancaspro V, Nuzziello M, Pallotta G, Sacchetti A, tients. Clin Nephrol 54:470-477, 2000 Petrarulo F: Intravenous ascorbic acid in hemodialysis 200. Vaux EC, Taylor DJ, Altmann P, et al: Effects of patients with functional iron deficiency: A clinical trial. J carnitine supplementation on muscle metabolism by the use Nephrol 13:444-449, 2000 of magnetic resonance spectroscopy and near-infrared spec- 215. Deira J, Diego J, Martinez R, et al: Comparative troscopy in end-stage renal disease. Nephron Clin Pract study of intravenous ascorbic acid versus low-dose desferrox- 97:c41-c48, 2004 amine in patients on hemodialysis with hyperferritinemia. J 200a. Steiber AL, Davis AT, Spry L, Strong J, Buss AL, Nephrol 16:703-709, 2003 Ratkiewicz MM, Weatherspoon LJ: Carnitine treatment 216. Mydlik M, Derzsiova K, Boldizsar J, Hribikova M, Petrovicova J: Oral use of iron with vitamin C in hemodia- improved quality-of-life measure in a sample of Midwestern lyzed patients. J Ren Nutr 13:47-51, 2003 hemodialysis patients. JPEN J Parenter Enteral Nutr 2006 217. Nguyen TV: Oral ascorbic acid as adjuvant to Jan-Feb; 30(1):10-5. epoetin alfa in hemodialysis patients with hyperferritinemia. 201. Hurot JM, Cucherat M, Haugh M, Fouque D: Effects Am J Health Syst Pharm 61:2007-2008, 2004 of L-carnitine supplementation in maintenance hemodialysis 218. Tovbin D, Markowitz A, Mostoslavsky M, Landver patients: A systematic review. J Am Soc Nephrol 13:708- R: Effectiveness of erythropoiesis on supervised intradia- 714, 2002 lytic oral iron and vitamin C therapy is correlated with Kt/V 202. Eknoyan G, Latos DL, Lindberg J: Practice recom- and patient weight. Clin Nephrol 53:276-282, 2000 mendations for the use of L-carnitine in dialysis-related 219. Chan D, Irish A, Dogra G: Efficacy and safety of oral carnitine disorder. National Kidney Foundation Carnitine versus intravenous ascorbic acid for anaemia in haemodialy- Consensus Conference. Am J Kidney Dis 41:868-876, 2003 sis patients. Nephrology (Carlton) 10:336-340, 2005 203. National Kidney Foundation: K/DOQI Clinical Prac- 220. Balcke P, Schmidt P, Zazgornik J, Kopsa H, Hauben- tice Guidelines for Nutrition in Chronic Renal Failure. Am J stock A: Ascorbic acid aggravates secondary hyperoxalemia Kidney Dis 35:S66-S70, 2000 (suppl 2) in patients on chronic hemodialysis. Ann Intern Med 101:344- 204. Bridges KR, Hoffman KE: The effects of ascorbic 345, 1984 acid on the intracellular metabolism of iron and ferritin. 221. Pru C, Eaton J, Kjellstrand C: Vitamin C intoxication J Biol Chem 261:14273-14277, 1986 and hyperoxalemia in chronic hemodialysis patients. Nephron 205. Lipschitz DA, Bothwell TH, Seftel HC, Wapnick 39:112-116, 1985 AA, Charlton RW: The role of ascorbic acid in the metabo- 222. Canavese C, Petrarulo M, Massarenti P, et al: lism of storage iron. Br J Haematol 20:155-163, 1971 Long-term, low-dose, intravenous vitamin C leads to plasma 206. Descombes E, Hanck AB, Fellay G: Water soluble calcium oxalate supersaturation in hemodialysis patients. vitamins in chronic hemodialysis patients and need for Am J Kidney Dis 45:540-549, 2005 supplementation. Kidney Int 43:1319-1328, 1993 223. Berger TM, Polidori MC, Dabbagh A, et al: Antioxi- 207. Sullivan JF, Eisenstein AB, Mottola OM, Mittal AK: dant activity of vitamin C in iron-overloaded human plasma. The effect of dialysis on plasma and tissue levels of vitamin J Biol Chem 272:15656-15660, 1997 C. Trans Am Soc Artif Intern Organs 18:277-282, 1972 224. Chen WT, Lin YF, Yu FC, Kao WY, Huang WH, Yan 208. Costello JF, Sadovnic MJ, Cottington EM: Plasma HC: Effect of ascorbic acid administration in hemodialysis oxalate levels rise in hemodialysis patients despite increased patients on in vitro oxidative stress parameters: Influence of oxalate removal. J Am Soc Nephrol 1:1289-1298, 1991 serum ferritin levels. Am J Kidney Dis 42:158-166, 2003 209. Gastaldello K, Vereerstraeten A, Nzame-Nze T, 225. Sirken G, Kung SC, Raja R: Decreased erythropoi- Vanherweghem JL, Tielemans C: Resistance to erythropoietin requirements in maintenance hemodialysis patients with etin in iron-overloaded haemodialysis patients can be over- statin therapy. ASAIO J 49:422-425, 2003 come by ascorbic acid administration. Nephrol Dial Trans- 226. Cooper A, Mikhail A, Lethbridge MW, Kemeny DM, plant 10:S44-S47, 1995 (suppl 6) Macdougall IC: Pentoxifylline improves hemoglobin levels REFERENCES S139 in patients with erythropoietin-resistant anemia in renal 243. Schiffl H, Lang SM, Bergner A: Ultrapure dialysate failure. J Am Soc Nephrol 15:1877-1882, 2004 reduces dose of recombinant human erythropoietin. Nephron 227. Weissgarten J, Modai D, Oz D, et al: Vitamin B(6) 83:278-279, 1999 therapy does not improve hematocrit in hemodialysis pa- 244. Sitter T, Bergner A, Schiffl H: Dialysate related tients supplemented with iron and erythropoietin. Nephron cytokine induction and response to recombinant human 87:328-332, 2001 erythropoietin in haemodialysis patients. Nephrol Dial Trans- 228. Ayli D, Ayli M, Azak A, et al: The effect of high-flux plant 15:1207-1211, 2000 hemodialysis on renal anemia. J Nephrol 17:701-706, 2004 245. Eiselt J, Racek J, Opatrny K Jr: The effect of 229. Locatelli F, Andrulli S, Pecchini F, et al: Effect of hemodialysis and acetate-free biofiltration on anemia. Int J high-flux dialysis on the anaemia of haemodialysis patients. Artif Organs 23:173-180, 2000 Nephrol Dial Transplant 15:1399-1409, 2000 246. Ward RA, Schmidt B, Hullin J, Hillebrand GF, 230. Opatrny K Jr, Reischig T, Vienken J, et al: Does Samtleben W: A comparison of on-line hemodiafiltration treatment modality have an impact on anemia in patients and high-flux hemodialysis: A prospective clinical study. with chronic renal failure? Effect of low- and high-flux J Am Soc Nephrol 11:2344-2350, 2000 biocompatible dialysis. Artif Organs 26:181-188, 2002 247. Wizemann V, Lotz C, Techert F, Uthoff S: On-line 231. Richardson D, Lindley EJ, Bartlett C, Will EJ: A haemodiafiltration versus low-flux haemodialysis. A prospec- randomized, controlled study of the consequences of hemo- tive randomized study. Nephrol Dial Transplant 15:S43-S48, dialysis membrane composition on erythropoietic response. 2000 (suppl 1) Am J Kidney Dis 42:551-560, 2003 248. Maduell F, del Pozo C, Garcia H, et al: Change from 232. Ifudu O, Feldman J, Friedman EA: The intensity of conventional haemodiafiltration to on-line haemodiafiltra- hemodialysis and the response to erythropoietin in patients tion. Nephrol Dial Transplant 14:1202-1207, 1999 with end-stage renal disease. N Engl J Med 334:420-425, 249. Lindsay RM: The London, Ontario, Daily/Nocturnal 1996 Hemodialysis Study. Semin Dial 17:85-91, 2004 233. Ifudu O, Uribarri J, Rajwani I, et al: Adequacy of 250. Ting GO, Kjellstrand C, Freitas T, Carrie BJ, dialysis and differences in hematocrit among dialysis facili- Zarghamee S: Long-term study of high-comorbidity ESRD ties. Am J Kidney Dis 36:1166-1174, 2000 patients converted from conventional to short daily hemodialysis. Am J Kidney Dis 42:1020-1035, 2003 234. Madore F, Lowrie EG, Brugnara C, et al: Anemia in 251. Rao M, Muirhead N, Klarenbach S, Moist L, Lindsay hemodialysis patients: Variables affecting this outcome RM: Management of anemia with quotidian hemodialysis. predictor. J Am Soc Nephrol 8:1921-1929, 1997 Am J Kidney Dis 42:18-23, 2003 235. Movilli E, Cancarini GC, Vizzardi V, et al: Epoetin 252. Schwartz DI, Pierratos A, Richardson RM, Fenton requirement does not depend on dialysis dose when Kt/V Ͼ SS, Chan CT: Impact of nocturnal home hemodialysis on 1.33 in patients on regular dialysis treatment with cellulosic anemia management in patients with end-stage renal dis- membranes and adequate iron stores. J Nephrol 16:546-551, ease. Clin Nephrol 63:202-208, 2005 2003 253. Woods JD, Port FK, Orzol S, et al: Clinical and 236. Movilli E, Cancarini GC, Zani R, Camerini C, biochemical correlates of starting “daily” hemodialysis. Sandrini M, Maiorca R: Adequacy of dialysis reduces the Kidney Int 55:2467-2476, 1999 doses of recombinant erythropoietin independently from the 254. Yuen D, Richardson RM, Fenton SS, McGrath- use of biocompatible membranes in haemodialysis patients. Chong ME, Chan CT: Quotidian nocturnal hemodialysis Nephrol Dial Transplant 16:111-114, 2001 improves cytokine profile and enhances erythropoietin re- 237. Kobayashi S, Moriya H, Aso K, Ohtake T: Vitamin sponsiveness. ASAIO J 51:236-241, 2005 E-bonded hemodialyzer improves atherosclerosis associated 255. Coronel F, Herrero JA, Montenegro J, et al: Erythro- with a rheological improvement of circulating red blood poietin requirements: A comparative multicenter study be- cells. Kidney Int 63:1881-1887, 2003 tween peritoneal dialysis and hemodialysis. J Nephrol 238. Miller ER III, Pastor-Barriuso R, Dalal D, Riem- 16:697-702, 2003 ersma RA, Appel LJ, Guallar E: Meta-analysis: High-dosage 256. Snyder JJ, Foley RN, Gilbertson DT, Vonesh EF, vitamin E supplementation may increase all-cause mortality. Collins AJ: Hemoglobin levels and erythropoietin doses in Ann Intern Med 142:37-46, 2005 hemodialysis and peritoneal dialysis patients in the United 239. Ward RA: Ultrapure dialysate. Semin Dial 17:489- States. J Am Soc Nephrol 15:174-179, 2004 497, 2004 257. Taji Y, Morimoto T, Okada K, Fukuhara S, Fukui T, 240. Ward RA: Ultrapure dialysate: A desirable and Kuwahara T: Effects of intravenous ascorbic acid on erythro- achievable goal for routine hemodialysis. Semin Dial 13:378- poiesis and quality of life in unselected hemodialysis 380, 2000 patients. J Nephrol 17:537-543, 2004 241. Kleophas W, Haastert B, Backus G, Hilgers P, 258. Keven K, Kutlay S, Nergizoglu G, Erturk S: Westhoff A, van Endert G: Long-term experience with an Randomized, crossover study of the effect of vitamin C on ultrapure individual dialysis fluid with a batch type machine. EPO response in hemodialysis patients. Am J Kidney Dis Nephrol Dial Transplant 13:3118-3125, 1998 41:1233-1239, 2003 242. Hsu PY, Lin CL, Yu CC, et al: Ultrapure dialysate 259. Sheashaa H, Abdel-Razek W, El-Husseini A, et al: improves iron utilization and erythropoietin response in Use of nandrolone decanoate as an adjuvant for erythro- chronic hemodialysis patients—A prospective cross-over poietin dose reduction in treating anemia in patients on study. J Nephrol 17:693-700, 2004 hemodialysis. Nephron Clin Pract 99:c102-c106, 2005 S140 REFERENCES

260. Gaughan WJ, Liss KA, Dunn SR, et al: A 6-month epoetin-induced pure red cell aplasia. Eur J Haematol study of low-dose recombinant human erythropoietin alone 73:389-396, 2004 and in combination with androgens for the treatment of 278. Bennett CL, Luminari S, Nissenson AR, et al: Pure anemia in chronic hemodialysis patients. Am J Kidney Dis red-cell aplasia and epoetin therapy. N Engl J Med 351:1403- 30:495-500, 1997 1408, 2004 261. Berns JS, Rudnick MR, Cohen RM: A controlled trial 279. Cournoyer D, Toffelmire EB, Wells GA, et al: of recombinant human erythropoietin and nandrolone decano- Anti-erythropoietin antibody-mediated pure red cell aplasia ate in the treatment of anemia in patients on chronic after treatment with recombinant erythropoietin products: hemodialysis. Clin Nephrol 37:264-267, 1992 Recommendations for minimization of risk. J Am Soc 262. McCullough PA, Lepor NE: The deadly triangle of Nephrol 15:2728-2734, 2004 anemia, renal insufficiency, and cardiovascular disease: 280. Kausz AT, Solid C, Pereira BJ, Collins AJ, St Peter implications for prognosis and treatment. Rev Cardiovasc W: Intractable anemia among hemodialysis patients: A sign Med 6:1-10, 2005 of suboptimal management or a marker of disease? Am J 263. Linman JW: Physiologic and pathophysiologic ef- Kidney Dis 45:136-147, 2005 fects of anemia. N Engl J Med 279:812-818, 1968 281. Cheng IK, Lu HB, Wei DC, Cheng SW, Chan CY, 264. Rao SV, Jollis JG, Harrington RA, et al: Relationship Lee FC: Influence of thalassemia on the response to of blood transfusion and clinical outcomes in patients with recombinant human erythropoietin in dialysis patients. Am J acute coronary syndromes. JAMA 292:1555-1562, 2004 Nephrol 13:142-148, 1993 265. House AA, Pham B, Page DE: Transfusion and 282. Di Iorio B, Guastaferro P, Bellizzi V: Relationship recombinant human erythropoietin requirements differ be- between resistance to erythropoietin and high anomalous tween dialysis modalities. Nephrol Dial Transplant 13:1763- hemoglobin levels in hemodialysis patients with beta- 1769, 1998 thalassemia minor. Blood Purif 21:376-380, 2003 266. Popovsky MA, Ransil BJ: Long-term impact of 283. Lai KN, Wong KC, Li PK, Lui SF: Use of recombi- recombinant human erythropoietin on transfusion support in nant erythropoietin in thalassemic patients on dialysis. Am J patients with chronic renal failure. Immunohematology Kidney Dis 19:239-245, 1992 12:1-3, 1996 284. Di Iorio B, De Nicola L, Bellizzi V, et al: Efficacy of 267. Eschbach JW: The anemia of chronic renal failure: erythropoietin on dialysis in patients with beta thalassemia Pathophysiology and the effects of recombinant erythropoi- minor. Blood Purif 22:453-460, 2004 etin. Kidney Int 35:134-148, 1989 285. Paritpokee N, Wiwanitkit V, Bhokaisawan N, Boon- 268. European Best Practice Guidelines for Renal Trans- chalermvichian C, Preechakas P: Serum erythropoietin plantation (part 1). Nephrol Dial Transplant 15:S1-S85, levels in pediatric patients with beta-thalassemia/hemoglo- 2000 (suppl 7) bin E. Clin Lab 48:631-634, 2002 269. Welch HG, Meehan KR, Goodnough LT: Prudent 286. Katopodis KP, Elisaf MS, Pappas HA, et al: Renal strategies for elective red blood cell transfusion. Ann Intern abnormalities in patients with sickle cell-beta thalassemia. J Med 116:393-402, 1992 Nephrol 10:163-167, 1997 270. American College of Physicians: Practice strategies 287. Papassotiriou I, Traeger-Synodinos J, Kanavakis E, for elective red blood cell transfusion. Ann Intern Med Karagiorga M, Stamoulakatou A, Kattamis C: Erythroid 116:403-406, 1992 marrow activity and hemoglobin H levels in hemoglobin H 271. Crosby E, Ferguson D, Hume HA, Kronick JB, disease. J Pediatr Hematol Oncol 20:539-544, 1998 Larke B, LeBlond P, et al: Guidelines for red blood cell and 288. Powe NR, Griffiths RI, Greer JW, et al: Early dosing plasma transfusion for adults and children. CMAJ 156:S1- practices and effectiveness of recombinant human erythro- S24, 1997 (suppl 11) poietin. Kidney Int 43:1125-1133, 1993 272. Ifudu O, Macey LJ, Friedman EA: Resurgence of 289. Tomson CR, Edmunds ME, Chambers K, Bricknell S, blood transfusion therapy in erythropoietin treated hemodi- Feehally J, Walls J: Effect of recombinant human erythropoi- alysis patients. ASAIO J 41:M426-M430, 1995 etin on erythropoiesis in homozygous sickle-cell anaemia 273. Casadevall N, Nataf J, Viron B, et al: Pure red-cell and renal failure. Nephrol Dial Transplant 7:817-821, 1992 aplasia and antierythropoietin antibodies in patients treated 290. Papassotiriou I, Voskaridou E, Stamoulakatou A, with recombinant erythropoietin. N Engl J Med 346:469- Loukopoulos D: Increased erythropoietin level induced 475, 2002 by hydroxyurea treatment of sickle cell patients. Hematol 274. Rossert J, Casadevall N, Eckardt KU: Anti- J 1:295-300, 2000 erythropoietin antibodies and pure red cell aplasia. J Am Soc 291. Dammacco F, Luccarelli G, Prete M, Silvestris F: Nephrol 15:398-406, 2004 The role of recombinant human erythropoietin alpha in the 275. Boven K, Stryker S, Knight J, et al: The increased treatment of chronic anemia in multiple myeloma. Rev Clin incidence of pure red cell aplasia with an Eprex formulation Exp Hematol Suppl 1:S32-S38, 2002 in uncoated rubber stopper syringes. Kidney Int 67:2346- 292. Weber G, Gross J, Kromminga A, Loew HH, Eckardt 2353, 2005 KU: Allergic skin and systemic reactions in a patient with 276. Eckardt KU, Casadevall N: Pure red-cell aplasia due pure red cell aplasia and anti-erythropoietin antibodies to anti-erythropoietin antibodies. Nephrol Dial Transplant challenged with different epoetins. J Am Soc Nephrol 18:865-869, 2003 13:2381-2383, 2002 277. Casadevall N, Cournoyer D, Marsh J, et al: Recom- 293. Verhelst D, Rossert J, Casadevall N, Kruger A, mendations on haematological criteria for the diagnosis of Eckardt KU, Macdougall IC: Treatment of erythropoietin- REFERENCES S141 induced pure red cell aplasia: A retrospective study. Lancet 311. Bassi S, Montini G, Edefonti A, et al: Cardiovascular 363:1768-1771, 2004 function in a chronic peritoneal dialysis pediatric population 294. Rossert J, Macdougall I, Casadevall N: Antibody- on recombinant human erythropoietin treatment. Perit Dial mediated pure red cell aplasia (PRCA) treatment and Int 13:S267-S269, 1993 (suppl 2) re-treatment: Multiple options. Nephrol Dial Transplant 312. Gerson A, Hwang W, Fiorenza J, et al: Anemia and 20:Siv23-Siv26, 2005 (suppl 4) health-related quality of life in adolescents with chronic 295. Asari A, Gokal R: Pure red cell aplasia secondary to kidney disease. Am J Kidney Dis 44:1017-1023, 2004 epoetin alpha responding to darbepoetin alpha in a patient on 313. Van Damme-Lombaerts R, Broyer M, Businger J, peritoneal dialysis. J Am Soc Nephrol 15:2204-2207, 2004 Baldauf C, Stocker H: A study of recombinant human 296. Tarng DC, Chang JG, Huang TP: Resistance to erythropoietin in the treatment of anaemia of chronic renal recombinant erythropoietin in a hemodialysis patient with failure in children on haemodialysis. Pediatr Nephrol 8:338- heterozygous hemoglobinopathy J-Meinung. Am J Kidney 342, 1994 Dis 29:769-772, 1997 314. Kurtin PS, Landgraf JM, Abetz L: Patient-based 297. Drueke TB: Resistance to recombinant human eryth- health status measurements in pediatric dialysis: Expanding ropoietin in hemodialysis patients. Am J Nephrol 10:S34- the assessment of outcome. Am J Kidney Dis 24:376-382, S39, 1990 (suppl 2) 1994 298. Ataga KI, Orringer EP: Renal abnormalities in sickle 315. Halterman JS, Kaczorowski JM, Aligne CA, Auinger cell disease. Am J Hematol 63:205-211, 2000 P, Szilagyi PG: Iron deficiency and cognitive achievement 299. Drueke TB: Modulating factors in the hematopoietic among school-aged children and adolescents in the United response to erythropoietin. Am J Kidney Dis 18:87-92, 1991 States. Pediatrics 107:1381-1386, 2001 300. Yorgin PD, Belson A, Al-Uzri AY, Alexander SR: 316. Burke JR: Low-dose subcutaneous recombinant The clinical efficacy of higher hematocrit levels in erythropoietin in children with chronic renal failure. Austra- children with chronic renal insufficiency and those undergo- lian and New Zealand Paediatric Nephrology Association. ing dialysis. Semin Nephrol 21:451-462, 2001 Pediatr Nephrol 9:558-561, 1995 301. Greenbaum LA: Anemia management in children, in 317. Jabs K: The effects of recombinant human erythropoi- Nissenson AR, Fine RN (eds): Clinical Dialysis. New York, etin on growth and nutritional status. Pediatr Nephrol 10:324-327, 1996 NY, McGraw Hill, 2005, pp 1111-1134 318. Scigalla P, Bonzel KE, Bulla M, et al: Therapy of 302. McGonigle RJ, Boineau FG, Beckman B, et al: renal anemia with recombinant human erythropoietin in Erythropoietin and inhibitors of in vitro erythropoiesis in the children with end-stage renal disease. Contrib Nephrol development of anemia in children with renal disease. J Lab 76:227-240; discussion 240-221, 1989 Clin Med 105:449-458, 1985 319. Brandt JR, Avner ED, Hickman RO, Watkins SL: 303. Nathan DG, Orkin SH: Appendix 11. Normal Safety and efficacy of erythropoietin in children with hematologic values in children, in Nathan DG, Orkin SH, chronic renal failure. Pediatr Nephrol 13:143-147, 1999 Ginsburg D, Look AT, Oski FA (eds): Nathan and Oski’s 320. Burghard R, Gordjani N, Leititis JU: [Treatment of Hematology of Infancy and Childhood (ed 6). Philadelphia, renal anemia with recombinant human erythropoietin]. PA, Saunders, 2003, p 1841 Monatsschr Kinderheilkd 137:174-177, 1989 304. Warady BA, Kausz A, Lerner G, et al: Iron therapy in 321. Campos A, Garin EH: Therapy of renal anemia in the pediatric hemodialysis population. Pediatr Nephrol 19: children and adolescents with recombinant human erythro- 655-661, 2004 poietin (rHuEPO). Clin Pediatr (Phila) 31:94-99, 1992 305. Warady BA, Zobrist RH, Wu J, Finan E: Sodium 322. Offner G, Hoyer PF, Latta K, Winkler L, Brodehl J, ferric gluconate complex therapy in anemic children on Scigalla P: One year’s experience with recombinant erythro- hemodialysis. Pediatr Nephrol 20:1320-1327, 2005 poietin in children undergoing continuous ambulatory or 306. Warady BA, Ho M: Morbidity and mortality in cycling peritoneal dialysis. Pediatr Nephrol 4:498-500, 1990 children with anemia at initiation of dialysis. Pediatr Neph- 323. Sinai-Trieman L, Salusky IB, Fine RN: Use of rol 18:1055-1062, 2003 subcutaneous recombinant human erythropoietin in children 307. Mitsnefes MM, Daniels SR, Schwartz SM, Meyer undergoing continuous cycling peritoneal dialysis. J Pediatr RA, Khoury P, Strife CF: Severe left ventricular hypertrophy 114:550-554, 1989 in pediatric dialysis: Prevalence and predictors. Pediatr 324. Eschbach JW,Adamson JW: Guidelines for recombi- Nephrol 14:898-902, 2000 nant human erythropoietin therapy. Am J Kidney Dis 14:2-8, 308. Mitsnefes MM, Kimball TR, Border WL, et al: 1989 Impaired left ventricular diastolic function in children with 325. North American Pediatric Renal Transplant Coopera- chronic renal failure. Kidney Int 65:1461-1466, 2004 tive Study (NAPRTCS) 2004 Annual Report. Boston, MA, 309. Martin GR, Ongkingo JR, Turner ME, Skurow ES, NAPRTCS Administrative Office, 2004 Ruley EJ: Recombinant erythropoietin (Epogen) improves 326. Sieniawska M, Roszkowska-Blaim M: Recombinant cardiac exercise performance in children with end-stage human erythropoietin dosage in children undergoing hemo- renal disease. Pediatr Nephrol 7:276-280, 1993 dialysis and continuous ambulatory peritoneal dialysis. 310. Baraldi E, Montini G, Zanconato S, Zacchello G, Pediatr Nephrol 11:628-630, 1997 Zacchello F: Exercise tolerance after anaemia correction 327. Port RE, Kiepe D, Van Guilder M, Jelliffe RW, with recombinant human erythropoietin in end-stage renal Mehls O: Recombinant human erythropoietin for the treat- disease. Pediatr Nephrol 4:623-626, 1990 ment of renal anaemia in children: no justification for S142 REFERENCES bodyweight-adjusted dosage. Clin Pharmacokinet 43:57-70, 345. Cook JD, Carriaga M, Kahn SG, Schalch W, Skikne 2004 BS: Gastric delivery system for iron supplementation. 328. Juul SE: Nonerythropoietic roles of erythropoietin in Lancet 335:1136-1139, 1990 the fetus and neonate. Clin Perinatol 27:527-541, 2000 346. Goch J, Birgegard G, Danielson BG, Wikstrom B: 329. De Palo T, Giordano M, Palumbo F, et al: Clinical Iron absorption in patients with chronic uremia on mainte- experience with darbepoietin alfa (NESP) in children under- nance hemodialysis and in healthy volunteers measured with going hemodialysis. Pediatr Nephrol 19:337-340, 2004 a simple oral iron load test. Nephron 73:403-406, 1996 330. Lerner G, Kale AS, Warady BA, et al: Pharmacoki- 347. Ruiz-Jaramillo Mde L, Guizar-Mendoza JM, Gutier- netics of darbepoetin alfa in pediatric patients with chronic rez-Navarro Mde J, Dubey-Ortega LA, Amador-Licona N: kidney disease. Pediatr Nephrol 17:933-937, 2002 Intermittent versus maintenance iron therapy in children on 331. Geary DF, Keating LE, Vigneux A, Stephens D, hemodialysis: A randomized study. Pediatr Nephrol 19:77- Hebert D, Harvey EA: Darbepoetin alfa (Aranesp) in 81, 2004 children with chronic renal failure. Kidney Int 68:1759- 348. Greenbaum LA, Pan CG, Caley C, Nelson T, Sheth 1765, 2005 KJ: Intravenous iron dextran and erythropoietin use in 332. Bald M, Holder M, Leichter HE: Improved therapy pediatric hemodialysis patients. Pediatr Nephrol 14:908- of renal anemia after switching to darbepoietin (NESP) in 911, 2000 peritoneal dialysis patients. Pediatr Nephrol 9:249A, 2002 349. Frankenfield DL, Neu AM, Warady BA, Fivush BA, (abstr) Johnson CA, Brem AS: Anemia in pediatric hemodialysis 333. Scharer K, Klare B, Braun A, Dressel P, Gretz N: patients: Results from the 2001 ESRD Clinical Performance Treatment of renal anemia by subcutaneous erythropoietin Measures Project. Kidney Int 64:1120-1124, 2003 in children with preterminal chronic renal failure. Acta 350. Tenbrock K, Muller-Berghaus J, Michalk D, Quer- Paediatr 82:953-958, 1993 feld U: Intravenous iron treatment of renal anemia in 334. Yalcinkaya F, Tumer N, Cakar N, Ozkaya N: children on hemodialysis. Pediatr Nephrol 13:580-582, 1999 Low-dose erythropoietin is effective and safe in children on 351. Morris KP, Watson S, Reid MM, Hamilton PJ, continuous ambulatory peritoneal dialysis. Pediatr Nephrol Coulthard MG: Assessing iron status in children with 11:350-352, 1997 chronic renal failure on erythropoietin: Which measurements should we use? Pediatr Nephrol 8:51-56, 1994 335. Schroder CH: The management of anemia in pediat- 352. Brugnara C, Zurakowski D, DiCanzio J, Boyd T, ric peritoneal dialysis patients. Guidelines by an ad hoc Platt O: Reticulocyte hemoglobin content to diagnose iron European committee. Pediatr Nephrol 18:805-809, 2003 deficiency in children. JAMA 281:2225-2230, 1999 336. Van Damme-Lombaerts R, Herman J: Erythropoietin 353. Bothwell TH, Pirzio-Biroli G, Finch CA: Iron absorp- treatment in children with renal failure. Pediatr Nephrol tion. I. Factors influencing absorption. J Lab Clin Med 51:24- 13:148-152, 1999 36, 1958 337. Gershanik J, Boecler B, Ensley H, McCloskey S, 354. Yorgin PD, Belson A, Sarwal M, Alexander SR: George W: The gasping syndrome and benzyl alcohol Sodium ferric gluconate therapy in renal transplant and renal poisoning. N Engl J Med 307:1384-1388, 1982 failure patients. Pediatr Nephrol 15:171-175, 2000 338. Hiller JL, Benda GI, Rahatzad M, et al: Benzyl 355. Danielson BG: Structure, chemistry, and pharmaco- alcohol toxicity: impact on mortality and intraventricular kinetics of intravenous iron agents. J Am Soc Nephrol hemorrhage among very low birth weight infants. Pediatrics 15:S93-S98, 2004 (suppl 2) 77:500-506, 1986 356. Ahsan N: Intravenous infusion of total dose iron is 339. Evans JH, Brocklebank JT, Bowmer CJ, Ng PC: superior to oral iron in treatment of anemia in peritoneal Pharmacokinetics of recombinant human erythropoietin in dialysis patients: A single center comparative study. J Am children with renal failure. Nephrol Dial Transplant 6:709- Soc Nephrol 9:664-668, 1998 714, 1991 357. Anuradha S, Singh NP, Agarwal SK: Total dose 340. Reddingius RE, Schroder CH, Koster AM, Monnens infusion iron dextran therapy in predialysis chronic renal LA: Pharmacokinetics of recombinant human erythropoietin failure patients. Ren Fail 24:307-313, 2002 in children treated with continuous ambulatory peritoneal 358. Dahdah K, Patrie JT, Bolton WK: Intravenous iron dialysis. Eur J Pediatr 153:850-854, 1994 dextran treatment in predialysis patients with chronic renal 341. Seeherunvong W, Rubio L, Abitbol CL, et al: failure. Am J Kidney Dis 36:775-782, 2000 Identification of poor responders to erythropoietin among 359. Domrongkitchaiporn S, Jirakranont B, Atamasrikul children undergoing hemodialysis. J Pediatr 138:710-714, K, Ungkanont A, Bunyaratvej A: Indices of iron status in 2001 continuous ambulatory peritoneal dialysis patients. Am J 342. Muller-Wiefel DE, Sinn H, Gilli G, Scharer K: Kidney Dis 34:29-35, 1999 Hemolysis and blood loss in children with chronic renal 360. Prakash S, Walele A, Dimkovic N, Bargman J, Vas S, failure. Clin Nephrol 8:481-486, 1977 Oreopoulos D: Experience with a large dose (500 mg) of 343. Gillespie RS, Wolf FM: Intravenous iron therapy in intravenous iron dextran and iron saccharate in peritoneal pediatric hemodialysis patients: A meta-analysis. Pediatr dialysis patients. Perit Dial Int 21:290-295, 2001 Nephrol 19:662-666, 2004 361. Folkert VW, Michael B, Agarwal R, et al: Chronic 344. Morgan HE, Gautam M, Geary DF: Maintenance use of sodium ferric gluconate complex in hemodialysis intravenous iron therapy in pediatric hemodialysis patients. patients: safety of higher-dose (Ͼ or ϭ250 mg) administra- Pediatr Nephrol 16:779-783, 2001 tion. Am J Kidney Dis 41:651-657, 2003 REFERENCES S143

362. Leijn E, Monnens LA, Cornelissen EA: Intravenous 379. Beshara S, Birgegard G, Goch J, Wahlberg J, iron supplementation in children on hemodialysis. J Nephrol Wikstrom B, Danielson BG: Assessment of erythropoiesis 17:423-426, 2004 following renal transplantation. Eur J Haematol 58:167-173, 363. Puntarulo S: Iron, oxidative stress and human health. 1997 Mol Aspects Med 26:299-312, 2005 380. Brown JH, Lappin TR, Elder GE, Taylor TN, Bridges 364. Erdemoglu AK, Ozbakir S: Serum ferritin levels and JM, McGeown MG: The initiation of erythropoiesis follow- early prognosis of stroke. Eur J Neurol 9:633-637, 2002 ing renal transplantation. Nephrol Dial Transplant 4:1076- 365. Gulcelik NE, Kayatas M: Importance of serum 1079, 1989 ferritin levels in patients with renal failure. Nephron 92:230- 381. Moore LW, Smith SO, Winsett RP, Acchiardo SR, 231, 2002 Gaber AO: Factors affecting erythropoietin production and 366. Gloggler A, Bulla M, Furst P: Effect of low dose correction of anemia in kidney transplant recipients. Clin supplementation of L-carnitine on lipid metabolism in Transplant 8:358-364, 1994 hemodialyzed children. Kidney Int Suppl 27:S256-S258, 382. Sun CH, Ward HJ, Paul WL, Koyle MA, Yanagawa 1989 N, Lee DB: Serum erythropoietin levels after renal transplan- 367. Gloggler A, Bulla M, Puchstein C, Gulotta F, Furst P: tation. N Engl J Med 321:151-157, 1989 Plasma and muscle carnitine in healthy and hemodialyzed 383. Tornero F, Prats D, Alvarez-Sala JL, Coronel F, Sanchez children. Child Nephrol Urol 9:277-282, 1988 A, Barrientos A: Iron deficiency anemia after successful renal 368. Zachwieja J, Duran M, Joles JA, et al: Amino acid transplantation. J Urol 149:1398-1400, 1993 and carnitine supplementation in haemodialysed children. 384. Goch J, Birgegard G, Wikstrom B, Tufveson G, Pediatr Nephrol 8:739-743, 1994 Danielson BG: Serum erythropoietin levels in the immediate 369. Murakami R, Momota T, Yoshiya K, et al: Serum kidney-posttransplant period. Nephron 60:30-34, 1992 carnitine and nutritional status in children treated with 385. Wolff M, Jelkmann W: Erythropoiesis and erythropoi- continuous ambulatory peritoneal dialysis. J Pediatr Gastro- etin levels in renal transplant recipients. Klin Wochenschr enterol Nutr 11:371-374, 1990 69:53-58, 1991 370. Warady BA, Borum P, Stall C, Millspaugh J, Taggart 386. Nampoory MR, Johny KV, al-Hilali N, Seshadri MS, E, Lum G: Carnitine status of pediatric patients on continu- Kanagasabhapathy AS: Erythropoietin deficiency and relative resistance cause anaemia in post-renal transplant recipi- ous ambulatory peritoneal dialysis. Am J Nephrol 10:109- ents with normal renal function. Nephrol Dial Transplant 114, 1990 11:177-181, 1996 371. Berard E, Iordache A: Effect of low doses of 387. Van Loo A, Vanholder R, Bernaert P, De Roose J, L-carnitine on the response to recombinant human erythro- Lameire N: Recombinant human erythropoietin corrects poietin in hemodialyzed children: About two cases. Nephron anaemia during the first weeks after renal transplantation: A 62:368-369, 1992 randomized prospective study. Nephrol Dial Transplant 372. Lilien MR, Duran M, Quak JM, Frankhuisen JJ, 11:1815-1821, 1996 Schroder CH: Oral L-carnitine does not decrease erythropoi- 388. Arbeiter K, Greenbaum L, Balzar E, et al: Reproduc- etin requirement in pediatric dialysis. Pediatr Nephrol ible erythroid aplasia caused by mycophenolate mofetil. 15:17-20, 2000 Pediatr Nephrol 14:195-197, 2000 373. Goodman AM, Pollack MM, Patel KM, Luban NL: 389. Kessler M: Erythropoietin and erythropoiesis in renal Pediatric red blood cell transfusions increase resource use. transplantation. Nephrol Dial Transplant 10:S114-S116, 1995 J Pediatr 142:123-127, 2003 (suppl 6) 374. Hume HA, Limoges P: Perioperative blood transfu- 390. Eckardt KU, Frei U, Kliem V, Bauer C, Koch KM, sion therapy in pediatric patients. Am J Ther 9:396-405, Kurtz A: Role of excretory graft function for erythropoietin 2002 formation after renal transplantation. Eur J Clin Invest 375. Niaudet P, Dudley J, Charbit M, Gagnadoux MF, 20:563-572, 1990 Macleay K, Broyer M: Pretransplant blood transfusions with 391. Lee DB: Interrelationship between erythropoietin cyclosporine in pediatric renal transplantation. Pediatr Neph- and erythropoiesis: Insights from renal transplantation. Am J rol 14:451-456, 2000 Kidney Dis 18:54-56, 1991 376. Gill JS, Tonelli M, Mix CH, Pereira BJ: The change 392. Muirhead N: Erythropoietin and renal transplanta- in allograft function among long-term kidney transplant tion. Kidney Int Suppl 69:S86-S92, 1999 recipients. J Am Soc Nephrol 14:1636-1642, 2003 393. Yorgin PD, Scandling JD, Belson A, Sanchez J, 377. Gill JS, Mix CH, Pereira BJ: Glomerular filtration Alexander SR, Andreoni KA: Late post-transplant anemia in rate and the prevalence of chronic kidney disease among adult renal transplant recipients. An under-recognized prob- transplant recipients in the United States. Am J Transplant lem? Am J Transplant 2:429-435, 2002 2:358A, 2002 (abstr) 394. Yorgin PD, Belson A, Sanchez J, et al: Unexpectedly 377A. Levey AS, Eckardt KU, Tsukamoto Y, et al: high prevalence of posttransplant anemia in pediatric and Definition and classification of chronic kidney disease: A young adult renal transplant recipients. Am J Kidney Dis position statement from Kidney Disease: Improving Global 40:1306-1318, 2002 Outcomes (KDIGO). Kidney Int 67:2089-2100, 2005 395. Vanrenterghem Y, Ponticelli C, Morales JM, et al: 378. Besarab A, Caro J, Jarrell BE, Francos G, Erslev AJ: Prevalence and management of anemia in renal transplant Dynamics of erythropoiesis following renal transplantation. recipients: A European survey. Am J Transplant 3:835-845, Kidney Int 32:526-536, 1987 2003 S144 REFERENCES

395A. Gill JS, Solid C, Pereira BJG, Collins AJ: Anemia thy: Report of two cases and review of the literature. management prior to kidney transplantation and the risk of Transplantation 67:539-544, 1999 post-transplant blood transfusion. Am J Transplant 4:S484, 414. Zarifian A, Meleg-Smith S, O’Donovan R, Tesi RJ, 2004 (suppl 8) (abstr 1191) Batuman V: Cyclosporine-associated thrombotic microangi- 396. Saito S, Fujiwara T, Sakagami K, Matsuno T, Tanaka opathy in renal allografts. Kidney Int 55:2457-2466, 1999 N: Anemia following renal transplantation. Transplant Proc 415. Dussol B, Brunet P, Vacher-Coponat H, Saingra Y, 30:3025-3026, 1998 Casanova P, Berland Y: Haemolytic uraemic syndrome in a 397. Saito S, Fujiwara T, Sakagami K, Matsuno T, Tanaka renal transplant recipient during OKT3 therapy. Nephrol N: Anemia following renal transplantation. Transplant Proc Dial Transplant 9:1191-1193, 1994 30:3025-3026., 1998 416. Morris-Stiff G, Evans M, Baboolal K, et al: Haemo- 398. Mix TC, Kazmi W, Khan S, et al: Anemia: A lytic uraemic syndrome associated with OKT3. Transpl Int continuing problem following kidney transplantation. Am J 9:522-523, 1996 Transplant 3:1426-1433, 2003 417. MacDonald AS: A worldwide, phase III, random- 399. Lorenz M, Kletzmayr J, Perschl A, Furrer A, Horl ized, controlled, safety and efficacy study of a sirolimus/ WH, Sunder-Plassmann G: Anemia and iron deficiencies cyclosporine regimen for prevention of acute rejection in among long-term renal transplant recipients. J Am Soc recipients of primary mismatched renal allografts. Transplan- Nephrol 13:794-797, 2002 tation 71:271-280, 2001 400. Winkelmayer WC, Kewalramani R, Rutstein M, 418. Kahan BD, Knight R, Schoenberg L, et al: Ten years Gabardi S, VonvisgerT, Chandraker A: Pharmacoepidemiol- of sirolimus therapy for human renal transplantation: The ogy of anemia in kidney transplant recipients. J Am Soc University of Texas at Houston experience. Transplant Proc Nephrol 15:1347-1352, 2004 35:S25-S34, 2003 (suppl 3) 401. Shibagaki Y, Shetty A: Anaemia is common after 419. Augustine JJ, Knauss TC, Schulak JA, Bodziak KA, kidney transplantation, especially among African Ameri- Siegel C, Hricik DE: Comparative effects of sirolimus and cans. Nephrol Dial Transplant 19:2368-2373, 2004 mycophenolate mofetil on erythropoiesis in kidney trans- 402. Kausman JY, Powell HR, Jones CL: Anemia in plant patients. Am J Transplant 4:2001-2006, 2004 pediatric renal transplant recipients. Pediatr Nephrol 19:526- 420. Crew RJ, Radhakrishnan J, Cohen DJ, et al: De novo 530, 2004 thrombotic microangiopathy following treatment with siroli- 403. Karthikeyan V, Karpinski J, Nair RC, Knoll G: The mus: Report of two cases. Nephrol Dial Transplant 20:203- burden of chronic kidney disease in renal transplant recipi- 209, 2005 ents. Am J Transplant 4:262-269, 2004 421. Naito M, Kawashima A, Akiba T, Takanashi M, 404. Miles AM, Markell MS, Daskalakis P, et al: Anemia Nihei H: Effects of an angiotensin II receptor antagonist and following renal transplantation: Erythropoietin response and angiotensin-converting enzyme inhibitors on burst forming iron deficiency. Clin Transplant 11:313-315, 1997 units-erythroid in chronic hemodialysis patients. Am J 405. Teruel JL, Lamas S, Vila T, et al: Serum ferritin levels Nephrol 23:287-293, 2003 after renal transplantation: A prospective study. Nephron 51: 422. Le Meur Y, Lorgeot V, Comte L, et al: Plasma levels 462-465, 1989 406. Moulin B, Ollier J, George F, et al: Serum erythropoi- and metabolism of AcSDKP in patients with chronic renal etin and reticulocyte maturity index after renal transplantation: failure: Relationship with erythropoietin requirements. Am J A prospective longitudinal study. Nephron 69:259-266, 1995 Kidney Dis 38:510-517, 2001 407. Chua MS, Barry C, Chen X, Salvatierra O, Sarwal 423. So BJ, Chae KM, Lee KK, Lee YJ, Jeong BH: Pure MM: Molecular profiling of anemia in acute renal allograft red cell aplasia due to parvovirus B19 infection in a renal rejection using DNA microarrays. Am J Transplant 3:17-22, transplant patient: A case report. Transplant Proc 32:1954- 2003 1956, 2000 408. Creemers GJ, van Boven WP, Lowenberg B, van der 424. Vales-Albertos LJ, Garcia-Cardenas M, Chavez- Heul C: Azathioprine-associated pure red cell aplasia. Becerra S, Gomez-Navarro B, Monteon-Ramos F, Cueto- J Intern Med 233:85-87, 1993 Manzano AM: Pure red cell aplasia associated with parvovi- 409. Pruijt JF, Haanen JB, Hollander AA, den Ottolander rus B19 infection in renal transplantation: The first case GJ: Azathioprine-induced pure red-cell aplasia. Nephrol report in Mexico. Transplantation 79:739, 2005 (abstr) Dial Transplant 11:1371-1373, 1996 425. Karras A, Thervet E, Legendre C: Hemophagocytic 410. Abraham KA, Little MA, Dorman AM, Walshe JJ: syndrome in renal transplant recipients: Report of 17 cases Hemolytic-uremic syndrome in association with both cyclo- and review of literature. Transplantation 77:238-243, 2004 sporine and tacrolimus. Transpl Int 13:443-447, 2000 426. Artz MA, Steenbergen EJ, Hoitsma AJ, Monnens 411. Epstein M, Landsberg D: Cyclosporine-induced LA, Wetzels JF: Renal transplantation in patients with thrombotic microangiopathy resulting in renal allograft loss hemolytic uremic syndrome: High rate of recurrence and and its successful reuse: A report of two cases. Am J Kidney increased incidence of acute rejections. Transplantation Dis 17:346-348, 1991 76:821-826, 2003 412. Pham PT, Peng A, Wilkinson AH, et al: Cyclosporine 427. Asaka M, Ishikawa I, Nakazawa T, Tomosugi N, Yuri and tacrolimus-associated thrombotic microangiopathy. Am J T, Suzuki K: Hemolytic uremic syndrome associated with Kidney Dis 36:844-850, 2000 influenza A virus infection in an adult renal allograft 413. Trimarchi HM, Truong LD, Brennan S, Gonzalez recipient: Case report and review of the literature. Nephron JM, Suki WN: FK506-associated thrombotic microangiopa- 84:258-266, 2000 REFERENCES S145

428. Waiser J, Budde K, Rudolph B, Ortner MA, Neu- treatment of anemia immediately after renal transplantation. mayer HH: De novo hemolytic uremic syndrome postrenal Transplantation 79:367-368, 2005 transplant after cytomegalovirus infection. Am J Kidney Dis 444A. Nagarajan S, Vidhun J, Mansfield E, et al: Molecular 34:556-559, 1999 mechanisms of transplant renal artery stenosis from early use of 429. Laurence J, Mitra D, Steiner M, Staiano-Coico L, erythropoietin. Am J Transplant 4:S568-S569, 2004 (suppl 8) Jaffe E: Plasma from patients with idiopathic and human 445. Bagnis C, Beaufils H, Jacquiaud C, et al: Erythropoi- immunodeficiency virus-associated thrombotic thrombocy- etin enhances recovery after cisplatin-induced acute renal topenic purpura induces apoptosis in microvascular endothe- failure in the rat. Nephrol Dial Transplant 16:932-938, 2001 lial cells. Blood 87:3245-3254, 1996 446. Cerami A, Brines M, Ghezzi P, Cerami C, Itri LM: 430. Anagnostou A, Lee ES, Kessimian N, Levinson R, Neuroprotective properties of epoetin alfa. Nephrol Dial Steiner M: Erythropoietin has a mitogenic and positive Transplant 17:S8-S12, 2002 (suppl 1) chemotactic effect on endothelial cells. Proc Natl Acad Sci 447. Fishbane S, Ragolia L, Palaia T, Johnson B, Elzein USA 87:5978-5982, 1990 H, Maesaka JK: Cytoprotection by darbepoetin/epoetin alfa 431. Anagnostou A, Liu Z, Steiner M, et al: Erythropoi- in pig tubular and mouse mesangial cells. Kidney Int etin receptor mRNA expression in human endothelial cells. 65:452-458, 2004 Proc Natl Acad Sci U S A 91:3974-3978, 1994 448. Grimm C, Wenzel A, Groszer M, et al: HIF-1-induced 432. Carlini RG, Alonzo EJ, Dominguez J, et al: Effect of erythropoietin in the hypoxic retina protects against light- recombinant human erythropoietin on endothelial cell apo- induced retinal degeneration. Nat Med 8:718-724, 2002 ptosis. Kidney Int 55:546-553, 1999 449. Parsa CJ, Matsumoto A, Kim J, et al: A novel 433. Debska-Slizien A, Chamienia A, Krol E, et al: protective effect of erythropoietin in the infarcted heart. Hemolytic anemia after renal transplantation: Analysis of J Clin Invest 112:999-1007, 2003 case reports. Transplant Proc 35:2233-2237, 2003 450. Westenfelder C: Unexpected renal actions of erythro- 434. Li FK, Chan TM, Lai KN: Alloimmune hemolysis poietin. Exp Nephrol 10:294-298, 2002 after renal transplantation. Am J Nephrol 20:473-475, 2000 451. Yang CW, Li C, Jung JY, et al: Preconditioning with 435. Peces R, Diaz Corte C, Navascues RA: [Hemolytic erythropoietin protects against subsequent ischemia-reperfu- anemia caused by graft-versus-host reaction in ABO- sion injury in rat kidney. FASEB J 17:1754-1755, 2003 nonidentical renal transplants from blood group O donors]. 452. Jindal KK, Hirsch DJ, Belitsky P, Whalen MA: Nefrologia 21:395-401, 2001 Low-dose subcutaneous erythropoietin corrects the anaemia 436. Djamali A, Becker YT, Simmons WD, Johnson CA, of renal transplant failure. Nephrol Dial Transplant 7:143- Premasathian N, Becker BN: Increasing hematocrit reduces 146, 1992 early posttransplant cardiovascular risk in diabetic trans- 453. Muirhead N, Cattran DC, Zaltzman J, et al: Safety plant recipients. Transplantation 76:816-820, 2003 and efficacy of recombinant human erythropoietin in correct- 437. Rigatto C, Parfrey P, Foley R, Negrijn C, Tribula C, ing the anemia of patients with chronic renal allograft Jeffery J: Congestive heart failure in renal transplant recipi- dysfunction. J Am Soc Nephrol 5:1216-1222, 1994 ents: Risk factors, outcomes, and relationship with ischemic 454. Traindl O, Barnas U, Franz M, et al: Recombinant heart disease. J Am Soc Nephrol 13:1084-1090, 2002 human erythropoietin in renal transplant recipients with 438. Rigatto C, Foley R, Jeffery J, Negrijn C, Tribula C, renal anemia. Clin Transplant 8:45-48, 1994 Parfrey P: Electrocardiographic left ventricular hypertrophy in 455. Yoshimura N, Oka T, Ohmori Y,Aikawa I: Effects of renal transplant recipients: Prognostic value and impact of recombinant human erythropoietin on the anemia of renal blood pressure and anemia. J Am Soc Nephrol 14:462-468, transplant recipients with chronic rejection. Transplantation 2003 48:527-529, 1989 439. Winkelmayer WC, Lorenz M, Kramar R, Horl WH, 456. Becker BN, Becker YT, Leverson GE, Heisey DM: Sunder-Plassmann G: Percentage of hypochromic red blood Erythropoietin therapy may retard progression in chronic cells is an independent risk factor for mortality in kidney renal transplant dysfunction. Nephrol Dial Transplant 17: transplant recipients. Am J Transplant 4:2075-2081, 2004 1667-1673, 2002 440. Khan S, Tighiouart H, Kalra A, Raman G, Rohrer RJ, 457. Lezaic V, Djukanovic L, Pavlovic-Kentera V: Recom- Pereira BJ: Resource utilization among kidney transplant binant human erythropoietin treatment of anemia in renal recipients. Kidney Int 64:657-664, 2003 transplant patients. Ren Fail 17:705-714, 1995 441. Schmidt R, Kupin W, Dumler F, Venkat KK, Mozes 458. Page B, Zingraff J: Resistance to rHuEpo and kidney M: Influence of the pretransplant hematocrit level on early graft rejection. Nephrol Dial Transplant 9:1696-1697, 1994 graft function in primary cadaveric renal transplantation. 459. Gill JS, Abichandani R, Khan S, Kausz AT, Pereira Transplantation 55:1034-1040, 1993 BJ: Opportunities to improve the care of patients with 442. Vasquez EM, Pollak R: Effect of pretransplant kidney transplant failure. Kidney Int 61:2193-2200, 2002 erythropoietin therapy on renal allograft outcome. Transplan- 460. Lopez-Gomez JM, Perez-Flores I, Jofre R, et al: tation 62:1026-1028, 1996 Presence of a failed kidney transplant in patients who are on 443. Paganini EP, Braun WE, Latham D, Abdulhadi MH: hemodialysis is associated with chronic inflammatory state Renal transplantation: results in hemodialysis patients previ- and erythropoietin resistance. J Am Soc Nephrol 15:2494- ously treated with recombinant human erythropoietin. ASAIO 2501, 2004 Trans 35:535-538, 1989 461. Atkins D, Best D, Briss PA, et al: Grading quality of 444. Van Biesen W, Vanholder R, Veys N, Verbeke F, evidence and strength of recommendations. BMJ 328:1490- Lameire N: Efficacy of erythropoietin administration in the 1494, 2004 zkd1050600spc1 4/28/06 2:17 PM Page 1

Supplement to VOL 47, NO 5, SUPPL 3, MAY 2006 American Journal of Kidney Diseases Vol 47, No 5, Suppl 3, May 2006 pp S1-S146 American Journal of Kidney Diseases KDOQI Clinical Practice Guidelines and Clinical Practice Recommendations for Anemia in Chronic Kidney Disease

W. B. Saunders, an Imprint of Elsevier