Philipp Lurz

Background: Recent trial data suggest that stratification of patients with heart failure with preserved ejection fraction (HFpEF) according to left ventricular ejection fraction (LVEF) provides a means for dissecting different treatment responses. However, the differential pathophysiologic considerations have rarely been described. Methods: This prospective, single-center study analyzed consecutive symptomatic patients with HFpEF diagnosed according to the 2016 European Society of Cardiology heart failure guidelines. Patients were grouped into LVEF 50% to 60% and LVEF >60% cohorts. All patients underwent cardiac magnetic resonance imaging. Transfemoral cardiac catheterization was performed to derive load-dependent and load-independent left ventricular (LV) properties on pressure–volume loop analyses. Results: Fifty-six patients with HFpEF were enrolled and divided into LVEF 50% to 60% (n=21) and LVEF >60% (n=35) cohorts. On cardiac magnetic resonance imaging, the LVEF >60% cohort showed lower LV end-diastolic volumes ( P =0.019) and end-systolic volumes ( P =0.001) than the LVEF 50% to 60% cohort; stroke volume ( P =0.821) did not differ between the cohorts. Extracellular volume fraction was higher in the LVEF 50% to 60% cohort than in the LVEF >60% cohort (0.332 versus 0.309; P =0.018). Pressure-volume loop analyses demonstrated higher baseline LV contractility (end-systolic elastance, 1.85 vs 1.33 mm Hg/mL; P <0.001) and passive diastolic stiffness (β constant, 0.032 versus 0.018; P =0.004) in the LVEF >60% cohort. Ventriculo-arterial coupling (end-systolic elastance/arterial elastance) at rest was in the range of optimized stroke work in the LVEF >60% cohort but was impaired in the LVEF 50% to 60% cohort (1.01 versus 0.80; P =0.005). During handgrip exercise, patients with LVEF >60% had higher increases in end-systolic elastance (1.85 versus 0.82 mm Hg/mL; P =0.023), attenuated increases in indexed end-systolic volume (−1 versus 7 mL/m²; P <0.004), and more exaggerated increases in LV filling pressures (8 vs 5 mm Hg; P =0.023). LV stroke volume decreased in the LVEF >60% cohort ( P =0.007) under exertion. Conclusions: Patients with HFpEF in whom LVEF ranged from 50% to 60% demonstrated reduced contractility, impaired ventriculo-arterial coupling, and higher extracellular volume fraction. In contrast, patients with HFpEF and a LVEF >60% demonstrated a hypercontractile state with excessive LV afterload and diminished preload reserve. A LVEF-based stratification of patients with HFpEF identified distinct morphologic and pathophysiologic subphenotypes.