Spironolactone in Patients With Heart Failure, Preserved Ejection Fraction, and Worsening Renal Function

Background: Treatment of heart failure with preserved ejection fraction (HFpEF) with spironolactone is associated with lower risk of heart failure hospitalization (HFH) but increased risk of worsening renal function (WRF). The prognostic implications of spironolactone-associated WRF in HFpEF patients are not well understood. Objectives: The purpose of this study was to investigate the association between WRF, spironolactone treatment, and clinical outcomes in patients with HFpEF. Methods: In 1,767 patients randomized to spironolactone or placebo in the TOPCAT (Treatment of Preserved Cardiac Function Heart Failure With an Aldosterone Antagonist Trial)-Americas study, we examined the incidence of WRF (doubling of serum creatinine) by treatment assignment. Associations between incident WRF and subsequent risk for the primary study endpoint of cardiovascular (CV) death, HFH, or aborted cardiac arrest and key secondary outcomes, including CV death, HFH, and all-cause mortality according to treatment assignment, were examined in time-updated Cox proportional hazards models with an interaction term. Results: WRF developed in 260 (14.7%) patients with higher rates in those assigned to spironolactone compared to placebo (17.8% vs. 11.6%; odds ratio: 1.66; 95% confidence interval: 1.27 to 2.17; p < 0.001). Regardless of treatment, incident WRF was associated with increased risk for the primary endpoint (hazard ratio: 2.04; 95% confidence interval: 1.52 to 2.72; p < 0.001) after multivariable adjustment. Although there was no statistical interaction between treatment assignment and WRF regarding the primary endpoint (interaction p = 0.11), spironolactone-associated WRF was associated with lower risk of CV death (interaction p = 0.003) and all-cause mortality (interaction p = 0.001) compared with placebo-associated WRF. Conclusions: Among HFpEF patients enrolled in TOPCAT-Americas, spironolactone increased risk of WRF compared with placebo. Rates of CV death were lower with spironolactone in both patients with and without WRF.