Prognostic Importance of Left Ventricular Global Longitudinal Strain in Patients with Severe Aortic Stenosis and Preserved Ejection Fraction.

BACKGROUND:Impaired left ventricular (LV) speckle-tracking-derived global longitudinal strain (GLS) magnitude (GLS worse than 14.7%) has been associated with poor outcome in patients with severe aortic stenosis (AS) and preserved LV ejection fraction (EF). OBJECTIVES:To test the hypothesis that GLS magnitude ≤ 15% obtained with vendor-independent speckle-tracking strain software may be able to identify patients with severe AS who are at higher risk of death, despite preserved LVEF and no or mild symptoms. METHODS:GLS was retrospectively obtained in 332 patients with severe AS (aortic valve area indexed [AVAi] < 0.6 cm2/m2), no or mild symptoms, and LVEF ≥ 50%. Absolute values of GLS were collected. Survival analyses were carried out to study the impact of GLS magnitude on all-cause mortality. RESULTS:During a median follow-up period of 42 (37-46) months, 105 patients died. On multivariate analysis, and after adjustment of known clinical and/or echocardiographic predictors of outcome and aortic valve replacement as a time-dependent covariate, GLS magnitude ≤ 15% was independently associated with mortality during follow-up (all P < .01). Adding GLS magnitude ≤ 15% (adjusted hazard ratio = 1.99 [1.17-3.38], P = .011) to a multivariate model including clinical and echocardiographic variables of prognostic importance (aortic valve replacement, aortic valve area, LV stroke volume index < 30 mL/m2, and LVEF<60%) improved the predictive performance with improved global model fit, reclassification, and better discrimination. After propensity score matching (n = 196), increased risk of mortality persisted among patients with GLS magnitude ≤ 15% compared with those with GLS > 15% (hazard ratio = 2.10; 95% confidence interval, 1.20-3.68; P = .009). CONCLUSIONS:In this series of patients with severe AS, no or mild symptoms, and LVEF ≥ 50%, GLS obtained with vendor-independent speckle-tracking strain software was an effective tool to identify patients with a poor outcome. Detection of myocardial dysfunction by identifying GLS magnitude < 15% in patients with severe AS, no or mild symptoms, and LVEF ≥ 50%, can aid in risk assessment.