Incremental Prognostic Value of Ventricular-Arterial Coupling over Ejection Fraction in Patients with Maintenance Hemodialysis

Abstract

Background: Left ventricular ejection fraction (LVEF) is a predictor of adverse outcomes in hemodialysis patients. LVEF is, however, an integral parameter determined by contractility, loading condition, and coupling. We sought to determine whether these components would better predict adverse outcomes and have incremental prognostic value over a validated clinical score and EF.

Methods: Two hundred thirty-four hemodialysis patients were prospectively followed up for primary composite endpoint: all-cause death, nonfatal myocardial infarction, and hospitalization due to worsening heart failure (HF). Load-independent contractility (end-systolic elastance [Ees] and preload recruitable stroke work [PRSW]) and arterial afterload (arterial elastance [Ea]) were noninvasively estimated. Ventricular-arterial coupling was assessed using the Ea/Ees ratio. LV global longitudinal strain (GLS) and mitral E-wave over annular velocity E' ratio (E/E') were also measured.

Results: During a median follow-up of 776 days, 30 patients developed the primary endpoint. Ees, PRSW, GLS, S', Ea/Ees, E/E', and EF were independently associated with the outcome after adjusting for the clinical score and prior HF hospitalization, whereas end-diastolic volume index or arterial afterload parameters were not. The nested Cox models indicated that Ea/Ees had independent and incremental predictive value over the model based on the score and either EF or E/E'. Furthermore, Ea/Ees continued to have predictive value after adjusting for GLS. The classification and regression analysis stratified event rates ranging from 4.2% to 68.8%.

Conclusions: LV contractility and Ea/Ees were independently associated with adverse outcome in hemodialysis patients. Ea/Ees had an incremental prognostic value over the clinical score and EF.

Keywords: Contractility; Hemodialysis; Load-independent; Prognosis; Ventricular-arterial coupling.