Meta-analysis: Erythropoiesis-Stimulating Agents in Patients With Chronic Kidney Disease Annals of Internal Med : July 6, 2010 vol. 153 no. 1 23-33Suetonia C. Palmer, MBChB; Sankar D. Navaneethan, MD, MPH; Jonathan C. Craig, MBChB, DCH, MM, PhD; David W. Johnson, MBChB(Hons), PhD; Marcello Tonelli, MD, SM; Amit X. Garg, MD, PhD; Fabio Pellegrini, MSc; Pietro Ravani, MD, MSc, PhD; Meg Jardine, MBBS, PhD; Vlado Perkovic, MBBS, PhD; Giusi Graziano, PhD;Richard McGee, BMedSci, MBBCh, MM; Antonio Nicolucci, MD; Gianni Tognoni, MD; and Giovanni F.M. Strippoli, MD, PhD, MPH, MM
• Context :To summarize the effects of ESA treatment on clinical outcomes in patients with anemia and CKD• Recent studies suggest that treating anemia of chronic kidney disease can sometimes cause harm.• This review compiles 27 randomized trials of erythropoiesis-stimulating agents (ESAs) in patients with anemia and chronic kidney disease. Treatment with ESAs that resulted in higher hemoglobin levels increased risks for stroke, worsening hypertension, and vascular access thrombosis more than strategies that resulted in lower hemoglobin levels (placebo, no treatment, or lower ESA dose).• Effects on all-cause mortality, cardiovascular events, and quality of life were unclear.
BackgroundAnemia frequently complicates chronic kidney disease (CKD), in part because ofdeficient renal erythropoietin releasePhase 1 and 2 clinical trials first evaluated recombinant human erythropoietinreplacement therapy in 1986 and 1987 in patients receiving hemodialysis. Thesetrials demonstrated dose-dependent increases in hemoglobin levels andavoidance of red blood cell transfusions, but they also reported incidence ofhypertension and vascular thrombosisAn earlier meta-analysis of 9 randomized trials involving 5143 patients with CKDconcluded that targeting a higher hemoglobin concentration (typically 12 to 15g/dL) versus a lower concentration (95 to 115 g/L) was associated with anincreased risk for all-cause mortality, arteriovenous access thrombosis, andhypertensionTREAT : The use of darbepoetin alfa in patients with diabetes, chronic kidneydisease, and moderate anemia who were not undergoing dialysis did not reducethe risk of either of the two primary composite outcomes (either death or acardiovascular event or death or a renal event) and was associated with anincreased risk of stroke
Study SelectionRandomized, controlled trials of at least 3 months duration that allocatedpatients to ESA versus placebo, no treatment, or different ESA doses toachieve a higher hemoglobin target versus a lower hemoglobin target.Trials of erythropoietin-alfa or -beta, darbepoetin, or a continuous erythropoietinreceptor activator at any dose and route of administration were eligibleIt was assumed that placebo-controlled trials were comparing 2 treatmenttargets for hemoglobin, although explicit targets were not always described.Trials of populations with any stage of CKD were included.
Data Synthesis & Statistical AnalysisTwo authors independently extracted data on characteristics of the participants,interventions, and clinical outcomes. Both reviewers also assessed the risks for bias in trials according to standardcriteria (allocation concealment, blinding of patients, investigators and outcomeassessors, completeness to follow-up, and use of intention-to-treat analysis)Any discrepancies in data extraction by discussion with an arbitrator.
Outline of participant populationMost studies enrolled participants between 50 and 60 years of age on average. In the 3largest trials of more than 1000 patients each, patients were generally older than 60 years.Two small studies were conducted in childrenWe identified 3 trials (1413 patients) that included only patients with establishedcardiovascular disease2 trials (4210 patients) that included only patients with diabetes mellitus and CKDAverage Male % was 52Estimated GFR was not calculated in studies with patients on Dialysis14 out of 27 Studies also included patients not on DialysisDoses of ESA used in the trials were not available in one third of the study reportsBaseline Hb and Target Hb levels were variable across 27 trials.
Trial Methodology qualitySuboptimal , according to current standardsMany trials were not reported according to the CONSORT (ConsolidatedStandards of Reporting Trials) statementNone of the 27 trials reported the details of allocation concealment (the methodby which the person responsible for treatment allocation is made unaware of aparticipants treatment assignment)Twelve of 16 trials published before 2004 did not conduct analysis by theintention-to-treat principle, whereas all but 1 of 11 trials from 2004 onwardanalyzed data according to randomized treatment allocationLarger trials that explored major patient-level outcomes (all-cause mortality ormajor adverse cardiovascular events) were generally of better methodologicalquality
Meta-AnalysisRisks for fatal and nonfatal stroke, vascular access thrombosis, and worsening hypertensionwere increased with a higher hemoglobin target than with a lower hemoglobin targetPatients randomly assigned to a higher hemoglobin target were less likely to require a bloodtransfusion (8 trials; 6482 patients) (RR, 0.61 [95% CI, 0.49 to 0.77]) but were more likely toreceive intravenous iron therapy (6 trials; 2283 patients) (RR, 1.57 [CI, 1.13 to 2.20])No statistically significant difference in the risk for all-cause mortality, serious cardiovascularevents, or fatal and nonfatal myocardial infarction between a higher hemoglobin target and alower targetIn the 10 studies conducted in people with CKD who were not dialysis-dependent, the risk forend-stage kidney disease requiring renal replacement therapy did not statistically significantlydiffer when hemoglobin target groups were comparedNo statistically significant differences between treatment groups in left ventricular mass at theend of treatment (4 trials; 1544 patients; mean difference, 0.14 g/m2 [CI, −4.60 to 4.88 g/m2])
Cumulative Meta-AnalysisSmall placebo-controlled trials conducted before 1998 showed effect estimates favoring a higherhemoglobin target for all-cause mortality, serious cardiovascular events, and nonfatal and fatal stroke.Larger, active comparator trials with higher hemoglobin targets conducted from 1998 onward favored alower hemoglobin target and showed risk for harm when targeting higher values.Across the following decade, the accrual of 12 further trials with 8233 additional patients and 1063 eventsdid not materially influence CI for excess mortality that were first identified in 1998.A probable increased risk for serious cardiovascular events with a higher hemoglobin target was evident in2006 after trials had randomly assigned 2670 patients (433 events), and this risk remained consistentacross subsequent trials that enrolled 4210 patients with 1245 additional events.A probable increase in the risk for stroke with ESA treatment to a higher hemoglobin target was firstindicated in 1998 and was confirmed by the placebo-controlled trial by Pfeffer and colleagues Evidence for increased risks for worsening hypertension and vascular access thrombosis with a higherhemoglobin target was evident from trials conducted in 1990
ResultsTreatment with either erythropoietin or darbepoetin to higher target hemoglobin levelsin patients with CKD (any stage) increased the risk for stroke, worsening hypertension,and vascular access thrombosis, compared with treatment to a lower hemoglobintargetNo statistically significant differences between higher and lower hemoglobin targets forthe risk for all-cause mortality, serious cardiovascular events, or end-stage kidneydiseasePoint estimates for all outcomes favored a lower hemoglobin target and effectivelyexcluded the likelihood of any clinically relevant benefit for a higher hemoglobin targetTargeting a higher hemoglobin level was linked to a reduction in need for red blood celltransfusions but greater prescription of intravenous iron therapyStudy identified no clinically important effects of ESA therapy on quality of life
ResultsCumulative meta-analysis demonstrated that evidence for harm associatedwith ESA therapy to achieve higher hemoglobin targets has remained stablesince it became apparent in 1998.findings are also consistent with those of a previous meta-analysis of 8 trialsevaluating target hemoglobin levels in CKD that showed an increase in the riskfor all-cause mortality with higher hemoglobin levels
LimitationsThe evidence for ESA therapy to improve quality of life is based largely onselective reporting of outcomes and is generally of low qualityThe available trial data do not identify the mechanisms by which ESA therapymay be harmfulTrials reported insufficient information to allow analysis of the independenteffects of ESA dose on outcomes
What now?Strong Evidence that Targeting Hb at 13 Gm/dL and achieving Hb over 12 by use of ESAs for both dialysis and pre-dialysis CKD patients increases risk of death and cardiovascular complications like stroke. Furthermore, there isreasonably strong evidence that high ESA dose in patients with resistance to ESAs is a risk for these samecomplications.Evidence for harm when targeting higher hemoglobin values in CKD has now been available for more than 20 years,and this evidence should be incorporated in guidelines and adopted in clinical practice.Study findings indicate that, on the basis of the available trial evidence, the rationale for the recommended hemoglobintarget of 110 to 120 g/L in the National Kidney Foundation clinical practice guideline on recommendations for anemia inCKD is unclearKDIGO :Updated anemia guideline in light of new study results, particularly the data from “Trial to ReduceCardiovascular Events with Aranesp Therapy” (TREAT) is under development. A study published in JAMA by Brookhart and colleagues, USRDS data were analyzed by examining ESA prescriptionpatterns in dialysis facilities, They found that when patients had Hct < 30%, centers using higher doses of ESAs andthose using more frequent doses of Fe had lower mortalityQuality of life & Exercise tolerance, 2 studies, Gandra and colleagues,Johansen and co-authors, found a correlationbetween ESA treatment, higher Hb levels, and exercise capacity and physical functioning.????
What Now?Should ESA be used in CKD, non Dialysis Pt?Should ESA be used in Dialysis Pt?When ESA should be started?What should be the Target Hb in CKD non Dialysis andDialysis Pt?