The Clinical Implications of Slope of Glomerular Filtration Rate in Clinical Trials of Chronic Kidney Disease

Background: Slope of the glomerular filtration rate (GFR) is considered a validated surrogate endpoint for chronic kidney disease (CKD) trials. However, differing short and long-term treatment effects on GFR slope can create ambiguities concerning the appropriate time period for evaluating slope, in part because current methods cannot separate the distinct contributions of the acute (before three months) and chronic (after three months) slopes for treatment effects on clinical endpoints. Methods: We estimated treatment effects on the acute and chronic GFR slopes and on the established clinical endpoint (CE) of kidney failure or serum creatinine doubling for 66 randomized treatment comparisons from previous CKD clinical trials. We used a novel Bayesian meta-regression framework to relate treatment effects on the established CE to both the acute and chronic slopes in a single multivariable model to determine the independent contributions of the acute and chronic slopes. Results: Treatment effects on both the acute and chronic slopes independently predicted the treatment effect on the established CE with a high median R 2 (95% credible interval) of 0.95 (0.79,1.00). For a fixed treatment effect on the chronic slope, each 1 mL/min/1.73m 2 greater acute GFR decline for the treatment vs. control increased the HR for the established CE by 11.4% (7.9%, 15.0%), against the treatment. The optimal weights for the acute and chronic slopes were consistent with the three-year total slope defined as the average slope extending from baseline to three years. Conclusion: Treatment effects on both the acute and chronic GFR slopes are independent determinants of the effects on the established CE, with variation in acute effects accounting for much of the observed variation in treatment effects on the CE across previous trials. Our results establish that acute effects impact the CE independently of treatment effects on the chronic slope, and support the three-year total slope as the primary slope-based outcome in randomized trials.