Higher hemoglobin levels using darbepoetin alfa and kidney outcomes in advanced chronic kidney disease without diabetes: a prespecified secondary analysis of the PREDICT trial

Background In the primary analysis of the PREDICT trial, a higher hemoglobin target (11–13 g/dl) with darbepoetin alfa did not improve renal outcomes compared with a lower hemoglobin target (9–11 g/dl) in advanced chronic kidney disease (CKD) without diabetes. Prespecified secondary analyses were performed to further study the effects of targeting higher hemoglobin levels on renal outcomes. Methods Patients with an estimated glomerular filtration rate (eGFR) 8–20 ml/min/1.73 m 2 without diabetes were randomly assigned 1:1 to the high- and low-hemoglobin groups. The differences between the groups were evaluated for the following endpoints and cohort sets: eGFR and proteinuria slopes, assessed using a mixed-effects model in the full analysis set and the per-protocol set that excluded patients with off-target hemoglobin levels; the primary endpoint of composite renal outcome, evaluated in the per-protocol set using the Cox model. Results In the full analysis set (high hemoglobin, n = 239; low hemoglobin, n = 240), eGFR and proteinuria slopes were not significantly different between the groups. In the per-protocol set (high hemoglobin, n = 136; low hemoglobin, n = 171), the high-hemoglobin group was associated with reduced composite renal outcome (adjusted hazard ratio: 0.64; 95% confidence interval: 0.43–0.96) and an improved eGFR slope (coefficient: + 1.00 ml/min/1.73 m 2 /year; 95% confidence interval: 0.38–1.63), while the proteinuria slope did not differ between the groups. Conclusions In the per-protocol set, the high-hemoglobin group demonstrated better kidney outcomes than the low-hemoglobin group, suggesting a potential benefit of maintaining higher hemoglobin levels in patients with advanced CKD without diabetes. Clinical trial registration Clinicaltrials.gov (identifier: NCT01581073).