Placebo effect in clinical studies in Hyperkalemia: A double-blind, Randomized, Placebo-Controlled phase 3 study of sodium zirconium cyclosilicate (ZS-9)

ZS-9, a nonabsorbed cation exchanger that selectively binds K throughout the GI tract, is being studied for treatment of hyperkalemia (HK), a common disorder associated with significant mortality and that limits the use of RAAS inhibitors. Recent clinical studies have evaluated new therapies for HK, both with and without placebo (PBO) controls. The need for PBO controls addresses the potential for regression to the mean due to randomly intercepted pt-specific variations in K over time. This is a well known phenomenon in studies based on laboratory-based endpoints. For example, in a large Phase 3 study, ZS-9 acutely reduced serum K in pts with HK (K ≥5.1 mEq/L) within 1 h, achieving mean reduction of -1.1 mEq/L by 48h (Packham NEJM 2014). K+-lowering effects of ZS-9 2.5, 5 and 10g were significantly greater than PBO; however, a significant drop was also observed with PBO, suggesting a placebo effect may at least partially be responsible for the observed effects in other studies that are not adequately controlled. Here we present an analysis of mean change in K for 41 pts with baseline (BL) K >5.5mEq/L and given PBO acutely. In the initial study phase, pts (N=753) with K 5.0-6.5 mEq/L were randomized to ZS-9 (1.25, 2.5, 5 or 10g) or PBO orally 3x daily for 48h. Paired t-tests were used to compare K at each timepoint vs. BL. Mean K change from BL in the PBO group are presented (Figure). (Figure presented) Here we report results of a PBO that had a significant effect on K+ reduction in a HK study. Since most acute therapies for HK are reported without PBO controls, caution should be exercised when interpreting results from single-arm or non-PBO-controlled HK studies.