Predictors of LVEF improvement in ambulatory patients with severly reduced LVEF

Background: In recent years, a new phenotype of heart failure has been described for patients that recover their left ventricular ejection fraction (LVEF) after initiation of guideline directed medical therapy (GDMT). This phenotype has been associated with a more favorable outcome and has been referred as heart failure with improved ejection fraction (HFimpEF). The aim of this study was to identify predictors of HFimpEF in a longitudinal cohort of ambulatory heart failure patients presenting to our specialized heart failure clinic.

Methods: The trial was based on a retrospective monocentric analysis of 791 heart failure patients enrolled in the heart failure registry Jena. The primary endpoint was the development of HFimpEF defined as an initial severely reduced LVEF with a second measurement of LVEF >40 % and absolute improvement in LVEF of more than 10% according to the current consensus. Clinical, laboratory, echocardiographic and cardio-pulmonary exercise testing (CPET) data were drawn from electronic patient records at baseline and 6 to 24 months later. Using multivariate binary logistic regression a model including the predictive parameters was developed. Cut-offs were determined using Youden J statistic for each parameter.

Results: Of the 791 subjects reviewed, 224 were women, the mean age was 65±14 years. 21,1 % of the 331 subjects with an initial LVEF below 40%, reached an improvement of LVEF as predefined in the primary endpoint. At baseline these patients were less likely to have cardiovascular comorbidities (e.g. coronary heart disease, myocardial infarction), had higher systolic blood pressure, less limitation of physical activity and higher platelet counts. They were more likely to be women (27,1 % vs. 15,7 %) or have a non-ischemic etiology of heart failure (23,2 % vs. 44 %). Patients with HFimpEF were more often treated with ARNI (58,6 % vs. 42,5 %) and showed less dilated left ventricles, regarding the end-systolic and end-diastolic volume (59 ml vs. 119 ml respectively 132 ml vs. 178 ml). In CPET subjects evolving a HFimpEF showed a higher maximum oxygen consumption per minute at baseline (19 ml/min/kg vs. 15 ml/min/kg) than patients with persistent heart failure with reduced ejection fraction (HFrEF). In a multivariate regression model, patients with an ESV <126 ml, a systolic blood pressure >133 mmHg and platelet concentrations >205,5 gpt/L had greater odds to experience an improvement in their LVEF. In contrast the presence of coronary heart disease reduced the odds of developing HFimpEF. While pVO2 showed a significant influence at baseline (OR 3,28; 95% CI: 1,32-8,15; p=0,011) it was non-significant when entered in a multivariate model.

Conclusions: Patients in the presented study cohort who developed HFimpEF in their course of disease are more often women, have a more favorable profile of cardiovascular risk factors and a less severe initial limitation in myocardial function as measured by transthoracic echocardiography.

Tab.1. multivariate logistic regression for development of HFimpEF