Hemoglobin Targets, ESA Use, and FDA Upcoming Hearings: Avoiding Unintended Consequences

Adeera Levin, MD, FRCPC; Allen R. Nissenson, MD, FACP

September 10, 2007

The practice of evidence-based medicine requires an understanding and appropriate interpretation of evidence, contextualized within the knowledge of the complexity of human disease and conditions.

The current ongoing discussion about the appropriate hemoglobin (Hgb) for patients with chronic kidney disease (CKD) misses the point in many aspects. The studies done to date, excessively quoted in multiple editorials, position pieces, and commentaries, in both printed and Web-based milieus, are now well known to the nephrology public: they include both observational and randomized controlled trials.[1-10]

The information that is available to the clinician from these studies is as follows:

  1. In the real world, in both hemodialysis (HD) and CKD patients, those patients that achieve Hgb values > 11 have improved outcomes compared with those who do not.[11-13] This information is from observational studies performed in large cohorts of patients in various locations both in the United States and in other countries. These observational studies do not prove that targeting Hgb values of < 11 to a higher level will improve outcomes, only that in those who can and do achieve Hgb > 11, there is a relative benefit. Of interest, in those observational studies of achieved Hgb, those with the highest Hgb values required the lowest doses of erythropoiesis-stimulating agent (ESA).

  2. In the randomized controlled trials, it is clear that targeting a higher Hgb concentration (> 13) is associated with risk.[6,8,14] This appears to be most true in those populations with the greatest burden of illness. In both the Normal Hct Cardiac Trial (NHCT) and the CHOIR study, the proportion of patients with severe cardiovascular disease (CVD) and diabetes mellitus (DM) exceeded that of the other studies performed in Europe, Canada, the United Kingdom, and Australia.[6,7] It is correct that targeting higher Hgb to 13.5 in these cohorts appears to result in harm. However, in both US studies, the dose of ESA was substantially higher than the doses used in all other CKD studies (median 11,000 units per week vs 3-5000 units per week). Of note, the rates of events are very different in CHOIR compared with all other studies; albeit in the same direction in all. Thus, one might have to ask more complex questions about the relationship between targeting higher Hgb in patients with significant comorbidity and high doses of a hormone with vasoactive, thrombogenic, and hematopoietic properties.

  3. All trials have limitations: the UK, Canadian, and Australian studies were relatively small (< 200 patients each) and used surrogate outcomes of left ventricular mass index (LVMI); CREATE was underpowered; and CHOIR randomized a high-risk group of patients using unusual doses of ESA in CKD relative to the rest of the world and had significant loss of patients enrolled. It is the largest study and answers a specific question: in the cohort enrolled in the study, using high doses of ESA to target Hgb > 13.5 does not result in any demonstrable benefit and appears to potentiate harm in those patients. Like all studies in medicine, generalizing from this cohort to all CKD patients without proper critical evaluation of the variability in populations can be problematic.

  4. Quality of life (QOL) is an important aspect of care of CKD patients. The measures used have differed, but the results are relatively consistent. CHOIR would be the only study to demonstrate no differences in QOL between groups. This might also bring into question whether or not the CHOIR cohort is truly representative of CKD patients, or were they more ill?[2]

Meta-analyses are important methods by which to try to appreciate disparate or large bodies of evidence. However, they are highly influenced by the size of the studies done and their quality, as it is difficult to adjust for study overall quality or particular shortcomings, and the size of any one study will overpower the smaller studies. CHOIR and NHCT are the 2 largest studies, differ from all others, and contribute over 70% of the effect size in any meta-analysis. Thus the meta-analysis published recently in The Lancet demonstrates harm but is heavily weighted toward 2 studies that may not be generalizable to all cohorts.[5,15]

Do all patients with CKD need ESAs? Do they all need the same Hgb target? Do they all need high doses of ESA and iron to achieve those targets? The answer to all of these is, of course, no.

Do CKD patients need close monitoring for complications related to CKD, CVD, and the impact of all medications that are prescribed, including ESAs? The answer is clearly yes.

Is there a strong signal that attempting to increase the Hgb with ESAs is associated with adverse outcomes? Yes, in specific populations (kidney patients who are ill and respond poorly to ESAs and some cancer patients), but not in all persons within those populations. This signal seems to be seen when attempting to achieve Hgbs higher than currently recommended by all of the current clinical practice guidelines from around the world.

Perhaps in the interest of evidence-based medicine, and in the absence of enough large randomized controlled trials in nephrology, we have overemphasized the largest trials, ignored their shortcomings or generalizability, and have now tried to apply a "one-size-fits-all" approach. Over the last decade, numerous individuals have benefited from the use of ESA therapy, there has been a steady decrease in mortality of dialysis patients despite their increasing age and number of comorbidities,[16] and there has been an increasing understanding of the complexity of clinical care and disease processes in CKD patients, including those on dialysis. It is important for the nephrology community to understand the recent trials in aggregate and in context and to better identify to whom those trials are applicable.

We would hope that the myriad of unanswered questions could be logically articulated and that the benefits of ESA therapy in the majority of individuals receiving reasonable doses could be maintained. Who does benefit from higher Hgb? Can we define the phenotype of those individuals more clearly and then apply that to populations? Given the lack of data in the different populations that constitute the CKD group, both those on and off dialysis, it would be premature to limit studies or discourage further research based on the few studies currently available, which answer specific questions, and which are not generalizable to all CKD populations whether on dialysis or in earlier stages of CKD.

The practice of evidence-based medicine requires an understanding and appropriate interpretation of evidence, contextualized within the knowledge of the complexity of human disease and conditions. The nephrology community needs to re-examine the complexity of anemia pathophysiology and treatment aspects before supporting any initiatives that may limit access to medications that have numerous, well-documented positive effects for our patients.

As physicians, we must always remember: First do no harm. In science, no single study provides "the" answer to a complex question or hypothesis. This is true in the field of anemia management. While randomized controlled trials are essential to test hypotheses, observational trials should not be dismissed out of hand.[17] We must keep in mind that not all studies are generalizable to all patients. The nephrology community as well as regulators, including the FDA and Centers for Medicare and Medicaid Services, need to understand the details of all published studies as well as consider the experiential information available from nephrologists and patients before considering policy changes in this area. Application of all data to all patients is inappropriate and potentially dangerous since it is clear that there are differences in benefits and risks of any given target Hgb based on patient characteristics. The overwhelming benefits of anemia treatment with ESAs in CKD and end-stage renal disease patients over the past 18 years should be acknowledged, and policy decisions that might set the care of patients back nearly 2 decades should not be made without overwhelming evidence that these are appropriate and based on an accurate understanding of the available evidence.


  1. Singh AK. Anemia and Erythropoietin: The FDA and CRAC. Available at: Accessed September 7, 2007.

  2. Levin A. Understanding recent haemoglobin trials in CKD: methods and lesson learned from CREATE and CHOIR. Nephrol Dial Transplant. 2007;22:309-312. Abstract

  3. Singh AK, Szczech L, Tang KL; for the CHOIR Investigators. Anemia of CKD: the CHOIR study revisited. Nephrol Dial Transplant. 2007 Apr 9; [Epub ahead of print].

  4. Strippoli GF, Craig JC, Manno C, Schena FP. Hemoglobin targets for the anemia of chronic kidney disease: a meta-analysis of randomized, controlled trials. J Am Soc Nephrol. 2004;15:3154-3165. Abstract

  5. Phrommintikul A, Haas SJ, Elsik M, Klum H. Mortality and target haemoglobin concentrations in anemia patients with chronic kidney disease treated with erythropoietin: a meta-analysis. Lancet. 2007;369:381-388. Abstract

  6. Besarab A, Bolton WK, Browne JK, et al. The effects of normal as compared with low hematocrit values in patients with cardiac disease who are receiving hemodialysis and epoetin. N Engl J Med. 1998;339:584-590. Abstract

  7. Singh AK, Szczech L, Tang KL; for the CHOIR Investigators. Correction of anemia with epoetin alfa in chronic kidney disease. N Engl J Med. 2006;355:2085-2098. Abstract

  8. Drueke TB, Locatelli F, Clyne N, for the CREATE Investigators. Normalization of hemoglobin level in patients with chronic kidney disease and anemia. N Engl J Med. 2006;355:2071-2084. Abstract

  9. Ritz E, Laville M, Bilous RW; for the Anemia Correction in Diabetes Study Investigators. Target level for hemoglobin correction in patients with diabetes and CKD: primary results of the Anemia Correction in Diabetes (ACORD) Study. Am J Kidney Dis. 2007;49:194-207. Erratum in: Am J Kidney Dis. 2007;49:562.

  10. Fishbane S, Nissenson AR. The new FDA label for erythropoietin treatment: how does it affect hemoglobin target? Kidney Int. 2007 Jun 27; [Epub ahead of print].

  11. Collins AJ, Li S, St. Peter W, et al. Death, hospitalization, and economic associations among incident hemodialysis patients with hematocrit values of 36 to 39%. J Am Soc Nephrol. 2001;12:2465-2473. Abstract

  12. Collins AJ, Ma JZ, Ebben J. Impact of hematocrit on morbidity and mortality. Semin Nephrol. 2000;20:345-349. Abstract

  13. Regidor DL, Kopple JD, Kovesdy CP, et al. Associations between changes in hemoglobin and administered erythropoiesis-stimulating agent and survival in hemodialysis patients. J Am Soc Nephrol. 2006;17:1181-1191. Abstract

  14. Parfrey PS, Foley RN, Wittreich BH, Sullivan DJ, Zagari MJ, Frei D. Double-blind comparison of full and partial anemia correction in incident hemodialysis patients without symptomatic heart disease. J Am Soc Nephrol. 2005;16:2180-2189. Abstract

  15. Strippoli GF, Navaneethan SD, Craig JC. Haemoglobin and haematocrit targets for the anaemia of chronic kidney disease. Cochrane Database Syst Rev. 2006;(4):CD003967. Review. Abstract

  16. Wolfe RA, Hulbert-Shearon TE, Ashby VB, Mahadevan S, Port FK. Improvements in dialysis patient mortality are associated with improvements in urea reduction ratio and hematocrit, 1999 to 2002. Am J Kidney Dis. 2005;45:127-135. Abstract

  17. Jager KJ, Stel VS, Wanner C, Zoccali C, Dekker FW. The valuable contribution of observational studies to nephrology. Kidney Int. 2007;72:671-675. Abstract