Early Identification of Patients at Risk for Incident Heart Failure With Preserved Ejection Fraction: Novel Approach to Echocardiographic Trends

Background: Heart failure with preserved ejection fraction (HFpEF) continues to increase in prevalence with a 50% mortality rate within 3 years of diagnosis, but lacking effective evidence-based therapies. Specific echocardiographic markers are not typically used to trigger alarm before acute HFpEF decompensation. The goal of this study was to retrospectively track changes in echocardiographic markers leading to the time of incident HFpEF hospitalization.

Methods and results: In a single-center, retrospective analysis, patients with HFpEF admitted between 2007 and 2014 were identified using the International Classification of Diseases, 9th Revision with search refined using the European Society of Cardiology HFpEF guidelines. Using linear mixed effects models, changes in echocardiographic markers preceding acute HF decompensation owing to incident HFpEF were analyzed. We report on an incident HFpEF cohort of 242 patients, extending 18 years retrospectively, and including 675 echocardiograms analyzed from the overall sample at 14 distinct time intervals before acute decompensation. The regression models demonstrated 3 echocardiographic markers with statistically significant increases across multiple time intervals including, arterial elastance (P = .006), right atrial pressure estimate (P < .001), and right ventricular systolic pressure (P = .006). Other echocardiographic markers had individual time intervals with significant increases before acute decompensation, including (a) left atrial diameter, 8 to 10 years before HFpEF diagnosis, (b) left ventricular filling pressure 2 to 6 years before HFpEF diagnosis, (c) ventricular elastance 3 to 6 months before HFpEF diagnosis, and (d) ventricular elastance/arterial elastance as early as 10 to 20 years and as late as 3 to 6 months before HFpEF diagnosis. Furthermore, African Americans presented with incident HFpEF at an average younger age than White patients (65.6 ± 15.2 years vs. 76.7 years ± 11.7, P < .001).

Conclusions: Noninvasive echocardiographic markers associated with incident HFpEF diagnosis showed long, mid, and acute range, significant changes as far back as 10 to 20 years and as close as 3 to 6 months before acute HFpEF decompensation. Including a diverse study cohort is critical to understanding the phenotypic differences of HFpEF. This hypothesis-generating study identified a novel approach to identifying trends in echocardiographic markers that may be used as a signal of impending incident HFpEF.