Prediction of Left Ventricular Ejection Fraction Change Following Treatment With Sacubitril/Valsartan

Background: Sacubitril/valsartan (Sac/Val) improves left ventricular ejection fraction (LVEF) in heart failure (HF) with reduced ejection fraction regardless of previous treatments. Improvements in LVEF may change eligibility for primary implantable cardioverter-defibrillator (ICD) placement. Awaiting LVEF improvement may expose patients to potential risks for arrhythmic complications. Objectives: The authors sought to develop a model predicting LVEF change after Sac/Val therapy. Methods: A total of 416 persons with HF and LVEF of <35% were included in this analysis. Following initiation of Sac/Val, echocardiographic parameters were measured serially for 1 year. A machine learning algorithm was implemented to develop a risk model for predicting the persistence of LVEF of <35% after 1 year and was validated in a separate group of study participants. Results: Baseline LVEF, left ventricular mass index, HF duration, age, N-terminal pro–B-type natriuretic peptide concentration at baseline and change by day 14, and body mass index were the most significant factors for identifying lack of LVEF improvement to ≥35% after 1 year. In the training and validation cohorts, the areas under the model curve for predicting lack of LVEF improvement were 0.92 and 0.86, respectively. Three categories of likelihood for LVEF of <35% after 1 year of Sac/Val treatment were developed based on the model predictions: 3.8%, 30.1%, and 83.7%. During follow-up, arrhythmia event rates were 0.9%, 2.9%, and 6.7% in these groups, respectively. Conclusions: Many persons with HF with reduced ejection fraction eligible for ICD insertion experience an increase in LVEF to ≥35% after treatment with Sac/Val. Early identification of those less likely to improve their LVEF might allow for more refined selection of primary ICD candidates.