A composite metric for predicting benefit from spironolactone in heart failure with preserved ejection fraction

Abstract

Aims: The TOPCAT trial showed no benefit for spironolactone in heart failure patients with preserved ejection fraction (HFpEF). Post-hoc, spironolactone helped participants from the Americas, but not Eastern Europe. Determining which patients with HFpEF could respond like TOPCAT's responders should help guide their care. We aimed to develop a TOPCAT Trial Score (TS) as a composite metric to identify such patients. Methods and results: From the TOPCAT individual-level data, we calculated a TS of age, body mass index, systolic blood pressure, heart rate, creatinine, potassium, glucose, left ventricular ejection fraction, and left atrial volume for each participant as a weighted distance in multidimensional space from the theoretical perfectly average Americas participant. Logistic regression was used to measure TS and spironolactone as predictors of TOPCAT's primary outcome. The relationship between TS and the H2FPEF score was also determined in TOPCAT and a registry cohort of real-world patients in the U.S. with HFpEF. A bimodal distribution of TS separated American (n = 1766) and Eastern European (n = 1,677) participants. Those with lower TS showed no significant response to spironolactone. Spironolactone's benefit rose with rising TS [βinteraction = -0.28 (P < 0.01)]. Significantly more American participants had benefit from spironolactone based on higher TS (> 1.14), in addition to higher likelihood of HFpEF based on higher H2FPEF scores (≥3). The cohort of real-world patients with HFpEF had even higher TS than American TOPCAT participants. Conclusions: Patients with HFpEF can be quantified by the TS to capture the likelihood of benefit from spironolactone.