Predictors of heart failure and all-cause mortality in asymptomatic patients with moderate and severe aortic regurgitation.

Abstract

Class I indications for aortic valve replacement (AVR) for severe chronic aortic regurgitation (AR) include AR attributable symptoms or left ventricular (LV) ejection fraction <50%. As noninvasive estimates of elevated LV filling pressures (LVFP's) have been noted to predict heart failure (HF) readmission and all-cause mortality (ACM) in HF patients, we hypothesize that elevated LVFP's may also be independent predictors of HF and ACM in chronic AR. We developed a single center patient database of moderate or greater AR diagnoses between 2003 and 2008 and followed each patient through January 2013. We included patients with >30 days follow-up with interpretable Doppler-echocardiograms. We recorded demographic variables, EuroScore II, incident HF and ACM, and Doppler-echo variables of LV size, systolic and diastolic function. Patients with severe AR (105 patients) and moderate AR (201 patients) had similar EuroScore II values and similar incident HF and ACM. For the 180 patients who developed HF, effective arterial elastance (aHR=1.70 (1.01-2.83), p=.041), LV end-diastolic dimension (aHR=1.83, (1.11-3.03), p=.0176), E/e' (aHR=3.04, (1.83-5.05), p<.0001), eccentric hypertrophy (EH) (aHR=2.39, (1.62-5.12), p=.0004), and tricuspid regurgitation (TR) velocity (aHR=5.75, (3.70-10.36), p<.0001) were independent predictors. For the 118 patients with ACM, EH (aHR=1.73, (1.02-3.28), p=.0414), systolic blood pressure (aHR=.58, (.33-.95), p=.0301), left atrial volume index (aHR=1.82, (1.06-3.06), p=.0293), E/e' (aHR=1.83, (1.07-3.08), p=.0280), and TR velocity (aHR=4.14, (2.22-6.49), p<.0001) were independent predictors. Elevated TR velocity and EH were strong markers of HF and ACM in patients with asymptomatic severe AR and in moderate AR.