Choice of endpoint in kidney outcome trials: considerations from the EMPA-REG OUTCOME® trial

Abstract

Background Doubling of serum creatinine [equivalent to 57% reduction in estimated glomerular filtration rate (eGFR)] is an established surrogate for end-stage kidney disease (ESKD); however, this endpoint necessitates lengthy follow-up and large sample sizes in clinical trials. We explored whether alternative eGFR decline thresholds provide more feasible surrogate kidney endpoints. Methods The study involved post hoc analysis of the EMPA-REG OUTCOME® trial. Adults with type 2 diabetes, high cardiovascular risk and eGFR ≥30 mL/min/1.73 m2 were assigned empagliflozin 10 mg or 25 mg (n = 4687) or placebo (n = 2333), on top of standard of care. We assessed composite endpoints incorporating different eGFR decline thresholds (≥30, ≥40, ≥50 or ≥57%) combined with initiation of renal replacement therapy (RRT) or renal death. This trial is registered with ClinicalTrials.gov (NCT01131676). Results Empagliflozin versus placebo significantly lowered the risk of decline in eGFR for each threshold listed above, combined with initiation of RRT or renal death, ranging from a hazard ratio (HR) of 0.81 [95% confidence interval (CI) 0.72–0.91] for endpoints based on 30% eGFR decline to an HR of 0.37 (0.23–0.61) for endpoints based on 57% eGFR decline. Lower thresholds (e.g. 30%) were associated with higher event rates but weaker treatment effects. The time to the 95% CI of the HR falling to <1.0 decreased with increasing eGFR threshold. Conclusions The composite of 40% decline in eGFR, ESKD or renal death appears to provide reliable results similar to the traditional 57% decline in eGFR.