Abstract 11860: Long Term Outcomes and Risk Stratification in Patients With Heart Failure With Recovered Ejection Fraction

Abstract

Introduction: Data are lacking on the natural history of heart failure with recovered ejection fraction (HFrecEF) with no available risk model for future heart failure events. This study aimed to understand the long-term outcomes of patients with HFrecEF, identify predictors of adverse events, and develop a risk stratification model. Methods: From an academic healthcare system we retrospectively identified patients who had an improvement in left ventricular ejection fraction (LVEF) from <40% to ≥53%. Significant predictors of all-cause mortality, hospitalization and recurrent reduction in LVEF were identified via Cox regression analysis. A Cox proportional hazard risk model was developed for all-cause mortality. Results: 133 patients were included (median age 67, 39% female, 30% ischemic etiology). Kaplan-Meier survival was 70% at 5 years. Freedom from hospitalization was 58% at 1 year, and the risk of recurrent LVEF reduction to <40% was 28% at 3 years. Diuretic dose and BNP at the time of LVEF recovery were the strongest predictors of mortality and hospitalization in multivariate adjusted analysis (BNP HR per 100pg/mL increase 1.13 [p<0.01]; furosemide equivalent dose HR per 40mg increase 1.19 [p=0.02]). An all-cause mortality risk model incorporating NYHA functional class, BNP and diuretic dose at the time of recovery showed excellent risk discrimination (c-statistic 0.79) and calibration. Conclusions: Patients with HFrecEF have significant heterogeneity in clinical outcomes ranging from persistent recovery to requiring advanced heart failure therapies or death. A risk model utilizing NYHA functional class, BNP and diuretic dose can accurately stratify mortality risk.