ESAs in Dialysis Patients

The philosopher Isaiah Berlin1 divides authors and perhaps people in general into two categories: Hedgehogs, who view the world through the lens of one defining idea, and foxes, for whom the world cannot be boiled down to a single idea. Looking at erythropoiesis-stimulating agents (ESAs) for the treatment of anemia in chronic kidney disease (CKD), should there be a single conclusion about the safety of ESAs? In a previous editorial, I discussed the use of ESAs in nondialysis patients.2 Here, I consider their continued use for the anemia of dialysis patients. ESAs have proved effective in correcting anemia in patients with ESRD. No one relishes blithely turning the clock back to the pre-ESA era, when dialysis patients were treated with repeated blood transfusions, iron therapy, anabolic steroids, and other maneuvers; however, mortality and the rate of cardiovascular complications in the dialysis population remain high, and reducing risk would be important in improving outcomes. Randomized, controlled trials (RCTs) demonstrate there is increased risk in correcting anemia with ESAs in all patients with CKD–both dialysis and nondialysis. In the recently published Trial to Reduce Cardiovascular Endpoints with Aranesp Therapy (TREAT),3 a placebo-controlled, double-blind, randomized study comprising 4038 nondialysis patients, there was a significantly higher rate of strokes in patients who were treated with darbepoetin (hazard ratio 1.92; P < 0.001) as well as higher rates of thromboembolism and cancer-related deaths. In the Correction of Hemoglobin and Outcomes in Renal Insufficiency (CHOIR) study,4 the risk was a higher rate of mortality and cardiovascular complications in nondialysis patients targeted to a hemoglobin (Hb) level of 13.0 g/dl with epoetin-alfa (hazard ratio 1.34; P = 0.03). A higher rate of death or myocardial infarction (MI) or vascular access thrombosis was also observed in the Normal Hematocrit study.5 Meta-analyses have reached similar conclusions.6,7 In the meta-analysis by Phrommintikul et al.,6 nine RCTs were selected on the basis of quality, sample size, and follow-up and viewed together for a total sample of 5143 patients, including both dialysis and nondialysis CKD trials. There was a higher risk for all-cause mortality (risk ratio [RR] 1.17; 95% confidence interval [CI] 1.01 to 1.35; P = 0.031) and arteriovenous access thrombosis (RR 1.34; 95% CI 1.16 to 1.54; P = 0.0001) in the higher Hb target group compared with the lower Hb target group. In a National Kidney Foundation (NKF) meta-analysis, published as a part of the revised 2007 NKF anemia guidelines,7 dialysis and nondialysis patients with CKD were evaluated separately. Four studies and 2391 patients were included in the NKF analysis of dialysis patients. The NKF reported increased risk with anemia correction, despite the use of a random-effect rather than a fixed-effect model. The point estimate for risk with the study by Phrommintikul et al.6 was 1.17 (95% CI 1.01 to 1.35; P = 0.031), whereas with the NKF study, the hazard ratio was 1.12 (95% CI 0.91 to 1.37). Some in the academy will argue that because both the CHOIR study and TREAT enrolled nondialysis patients with CKD, these studies are not generalizable to the dialysis population; however, results of two RCTs of dialysis patients counter this assertion. The Normal Hematocrit study5 is the largest anemia correction RCT of dialysis patients. It tested the hypothesis that the correction of anemia with Epogen in hemodialysis patients with clinical evidence of congestive heart failure or ischemic heart disease would improve outcomes. The primary end points were length of time to death or a first nonfatal MI. The study was halted at the third interim analysis on the recommendation of the Data Safety Monitoring Committee. At 29 months, there were 183 deaths and 19 first nonfatal MIs in the group with a normal hematocrit level and 150 deaths and 14 nonfatal MIs in the low hematocrit group (RR 1.3; 95% CI 0.9 to 1.9). Even though these differences did not reach the prespecified statistical stopping boundary, the study was halted for safety reasons. In addition, the incidence of thrombosis of vascular access sites was higher in the normal hematocrit group compared with the low-hematocrit group (243 [39%] versus 176 [29%] patients; P = 0.001). Unpublished data from the Normal Hematocrit study, archived by the Food and Drug Administration (FDA),8 showed that the incidence of nonfatal MI was 3.1 and 2.3% in the normal and low hematocrit groups, respectively. The incidences of vascular thrombosis (39 versus 29%) and all other thrombotic events (22 versus 18%) were also higher in the normal hematocrit group (of note, there was no significant difference in the rate of stroke: 7% in normal and 6% in normal hematocrit group, respectively; personal communication from David Goodkin). The risk for thrombotic events is consistent with the TREAT data. With respect to patient-reported quality-of-life determinations, only one category of eight showed improvement in quality of life with higher Hb concentrations, but between-group comparisons (high versus low Hb) for quality of life were not reported. Because the Normal Hematocrit study recruited high-risk patients, it may not be generalizable to the dialysis population at large; however, consider the Canadian-European Normalization of Hemoglobin with Erythropoietin Trial,9 which enrolled relatively healthy incident hemodialysis patients by excluding patients with symptomatic heart disease as well as those with left ventricular dilation at baseline. Patients were randomly assigned to higher versus lower Hb (13.0 versus 11.0 g/dl, respectively). The primary end point was a change in left ventricular volume index. Changes in parameters of heart failure, stroke, and quality of life were also measured. No significant benefit in either of the cardiac structural or functional parameters was observed in the high- versus low-Hb groups; however, there was a statistically significantly higher rate of stroke in the higher Hb group. Quality of life showed improvement in the vitality score in the high- versus low-Hb group, although the 6-minute walk test and two other measures of quality of life did not improve significantly. The FDA has lumped together the major RCTs on anemia correction in patients with CKD.10 The current FDA recommendation states a Hb goal of 10 to 12 g/dl for all patients with CKD, regardless of dialysis status.10 From a safety perspective, there is some rationale for lumping the trials together: The risk of ESA exposure seems insensitive to the patient population. Data from RCTs of patients with CKD,3,4,5,11,12 patients with cancer,11,13 healthy individuals who underwent spine surgery,14 and critically ill patients15 all point to increased risk with ESA exposure. In patients with CKD, data from both observational studies and secondary analyses of RCTs suggest that exposure to high dosages of ESA explains the higher risk for mortality and cardiovascular complications16–19; however, no study has proved causality so far, and the issue continues to be debated.20 Confounding by indication and dosage-targeting bias have been raised as limitations in these analyses.21,22 Furthermore, not all studies implicate ESA exposure in explaining risk; a recent study using marginal structural modeling failed to report an association between high ESA dosage and adverse outcome.23 Conversely, a secondary analysis of the CHOIR study, using landmark analysis as a way to avoid some of the biases inherent in observational studies, demonstrated that high ESA dosage was an independent predictor of the combined end point of mortality, nonfatal MI, heart failure, and stroke.17 In this analysis, targeting a higher Hb or lower Hb with higher dosages of ESA associates with poor outcomes, whereas achieving a higher Hb improves outcomes. A fair appraisal of the evidence is that, although additional studies will be needed to address causality, we should be concerned about the possibility of adverse effects from high ESA dosage. Reducing ESA dosage in dialysis patients will probably result in a higher red cell transfusion rate. In the Normal Hematocrit study,5 129 (21%) patients in the normal hematocrit group received red cell transfusions compared with 192 (31%) patients in the low hematocrit group (P < 0.001). In the Canada-Europe study, transfusion rates were 9.1% with the high target versus 19.3% with the low target. In TREAT, nearly twice as many patients required blood transfusions when assigned to placebo versus darbepoetin (496 [24.5%] versus 297 [14.8%] patients, respectively; P < 0.001). Although the FDA review cited a low rate of transfusion-related complications in the current era, it is unlikely that repeated blood transfusions will be without risk.11 Iron use will also rise with aggressive attempts to reduce ESA dosage. Exposure to large amounts of intravenously administered iron, over the long-term, raises safety concerns,24 although observational studies have been reassuring.25 In the absence of additional RCTs, clinicians are faced with making decisions on what to do with managing anemia in dialysis patients. Maintaining dialysis patients on the lowest possible dosage of ESA, above a Hb threshold of >9 g/dl, is commensurate with the placebo arm of TREAT. This Hb threshold is also consistent with the lower Hb arm of the Normal Hematocrit study. For some patients, this strategy will be well tolerated and should be the goal. These patients probably will be healthier and have less laboratory evidence of inflammation, normal iron stores, well-controlled metabolic bone disease parameters, and good dialysis adequacy; however, a Procrustean one-size-fits-all approach to treating anemia in dialysis patients has its own problems, because dramatically decreasing the ESA dosage in some patients, perhaps the majority, will result in a precipitous fall in Hb concentration, long-term complications of repeated blood transfusion, and intolerable fatigue. In these patients, engaging in a discussion about the risks and benefits of ESA therapy, as well as individualizing goals of therapy, will be important. A computerized “generic” anemia protocol will not do for these patients. Maneuvers that lower ESA dosage but prevent a concomitant drop in the Hb concentration should be aggressively pursued. These include using intravenous iron therapy to keep the patient iron-replete, even when the patient has a high ferritin level.26,27 Switching to subcutaneous ESA will also result in one-third lower ESA dosage use,28 and patients may be willing to trade the discomfort of subcutaneous ESA for greater safety.29 Optimizing adequacy of dialysis or switching patients to alternative modalities, such as peritoneal dialysis30 or nocturnal31 hemodialysis, may also result in lower ESA dosage. Treating inflammation either with drugs, such as pentoxyfilline32 or a statin,33 or treating an underlying infection or treating hyperparathyroidism also seem effective in reducing ESA dosage without causing the Hb to fall. For some patients who are waiting for a kidney allograft, long-term ESA therapy will be necessary because blood transfusion must be avoided to minimize the risk for sensitization. Conversely, avoiding ESAs should be considered for patients who are within 2 years of a diagnosis of a malignancy or are actively undergoing chemotherapy. In conclusion, there is increased risk in targeting a higher Hb level in dialysis patients. Aiming for an Hb level of >9 g/dl is commensurate with the placebo arm of TREAT but also the lower Hb arm of the Normal Hematocrit study; however, a higher rate of red cell transfusion will probably be necessary with such a strategy. The key to ESA therapy in dialysis patients will be individualization of risk and benefit. In seventh century BC, the Greek soldier-poet Archilochus34 wrote, “The fox knows many things but the hedgehog knows one big thing.” In CKD anemia, the one big thing is using the lowest possible dosage of ESA because of safety concerns; however, reducing exposure to ESA and individualizing therapy will require “many things” and undeniably the smarts of a fox. Disclosures A.K.S. was principal investigator of the CHOIR study and a member of the Executive Committee for TREAT. He presented to the FDA Cardiovascular Disease and Renal Advisory Committee in September 2007 and to the US Congress House of Representatives Ways and Means Committee in December 2006 and June 2007. He has received consulting fees from Amgen, Johnson and Johnson, Fibrogen, and Watson and has received grant support from Amgen, Johnson and Johnson, and Watson.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.