PROGNOSTIC UTILITY OF DIASTOLIC DYSFUNCTION AND SPECKLE TRACKING IN PATIENTS WITH HEART FAILURE WITH REDUCED EJECTION FRACTION

Background: We hypothesized that grading of diastolic dysfunction (DDF) yields prognostic information on all-cause mortality in patients with heart failure with reduced ejection fraction (HFrEF). Methods: We studied ambulatory HFrEF (n = 917) patients (mean age 67 ± 12 years, EF 27 ± 9%) who underwent echocardiography and speckle tracking. Patients were divided in categories of DDF (grade 1-3) according to two algorithms (Fig A-B). We assessed prognostic performance by Harrel's C-statistics. Results: Concordance between the algorithms was moderate (Kappa = 0.48) and reclassification was 33%. During median follow-up 3.5 years (IQR 1.9, 4.8 years), 142 (15%) died. DDF algorithm 1 (HR 1.45 per grade increase, 95%CI 1.19-1.77, P<0.001) and algorithm 2 (HR 1.61 per grade increase, 95%CI 1.29 to 2.00, P<0.001) yielded prognostic information. We observed no difference in performance between the algorithms (C-stat 0.61 vs 0.62, P=0.24). When the three most powerful predictors, assessed by C-stat, were combined (algorithm 1: E/e', TR and LAVI and algorithm 2: E/e', E/A and LAVI) the prognostic performance increased (C-stat 0.63 and 0.64), however, no difference was observed (P difference=0.11). When global longitudinal strain, by speckle tracking, was added, C-stat for both algorithms increased (P<0.05 for both) (Fig C). Conclusion Grading of DDF in HFrEF patients yields prognostic information on all-cause mortality. Addition of GLS to contemporary algorithms of DDF adds novel prognostic information.