Male sex and insulin-treated diabetes were independently associated with sudden death (SD) in patients with heart failure with preserved ejection fraction (HFpEF), which accounted for 19 percent of deaths in a TOPCAT trial analysis presented by Muthiah Vaduganathan, MD, MPH, at ACC.18. The results were published simultaneously in JACC: Heart Failure.

The competing risk analysis investigated the incidence rates of the composite of SD or aborted cardiac arrest (ACA), unique predictors of SD or ACA, and response to spironolactone in patients with HFpEF. This analysis was performed in the TOPCAT patient cohort from the Americas, which included 1,767 patients. All eligible patients had a left ventricular ejection fraction (LVEF) ≥45 percent. The patients were randomized to either spironolactone or matching placebo. Predictors of composite SD and ACA were identified using a step-wise backward selection with competing risks regression analysis accounting for non-SD causes of death.

During a median 3.0 years of follow-up, a total of 385 patients died, of whom 19 percent were due to SD and 81 percent to non-SD causes. Six ACA events occurred during that time. During 5,426 person-years of observation, incidence rates of SD/ACA vs. non-SD/ACA were 1.4 and 5.8 events/100 patient-years, respectively.

Patients with SD/ACA were younger, male and had higher body mass indexes and rates of diabetes. The final competing risks regression model showed that male sex and insulin-treated diabetes were independent predictors of SD/ACA with modest discrimination. Male sex and insulin-treated diabetes remained independent predictors of SD/ACA in sensitivity analyses excluding patients with implantable cardioverter-defibrillators and when predicting SD alone.

Incidence rates of SD/ACA in patients receiving spironolactone vs. placebo were 1.2 vs. 1.6 per patient-years, respectively. Competing risks regression analysis showed that the difference between the treatment groups was not statistically significant.

The authors concluded that further studies are needed to define mechanisms contributing to SD, build more robust SD risk prediction models integrating imaging and biomarkers, and develop strategies to attenuate SD risk in patients with HFpEF.

Keywords: ACC18, ACC Annual Scientific Session, Spironolactone, Defibrillators, Implantable, Stroke Volume, Insulin, Incidence, Follow-Up Studies, Cause of Death, Heart Failure, Diabetes Mellitus, Ventricular Function, Left, Death, Sudden, Regression Analysis, Heart Arrest, Biomarkers

< Back to Listings