CI-452767-1 THE ASSOCIATION BETWEEN QRS NARROWING WITH CARDIAC RESYNCHRONIZATION THERAPY, MORTALITY AND HEART FAILURE: A POST-HOC ANALYSIS OF THE RESYNCHRONIZATION IN AMBULATORY HEART FAILURE TRIAL (RAFT)

Abstract

Amongst patients with left ventricular systolic dysfunction and prolonged QRS, cardiac resynchronisation therapy (CRT) can improve cardiac electrical/mechanical synchrony, and prevent adverse clinical outcomes. The baseline QRS duration is associated with improved response to CRT. However, the association between QRS narrowing after implantation and clinical response to CRT is unclear. To determine whether the degree of QRS narrowing is associated with clinical outcome following CRT implantation. The RAFT study enrolled 1798 patients with left ventricular ejection fraction (LVEF) ≤35%, QRS duration ≥120 ms and NYHA II-III to CRT-D or ICD. 894 were randomized to CRT-D of which 821 patients received CRT-D and had electrocardiograms (ECGs) available for review. The degree of QRS narrowing (dQRS) was calculated as the difference between baseline QRS duration and the widest QRS with CRT pacing post implantation. The primary outcome was a composite of death and heart failure hospitalisation. Multivariable Cox proportional hazards models were performed to assess the association between dQRS and the incidence of the primary composite outcome. To model the variation in risk of the outcome by dQRS, a restricted cubic spline function was used. The median age was 67 (IQR 59-73) years, 125 patients (15.2%) were women, median LVEF 24%, 68% had ischemic heart disease and 30% had atrial fibrillation. The baseline median QRS duration was 160 (IQR 140-178.5) ms, 68% had a left bundle branch block (LBBB). The median dQRS was 2 ms (IQR -18-20 ms), and 450 (55%) of patients had QRS narrowing post CRT. dQRS was associated with a reduction in the composite outcome (HR = 0.988, 95% CI 0.984-0.993), all-cause mortality (HR = 0.990, 95% CI 0.984-0.995) and heart failure hospitalisation (HR 0.990, 95% CI 0.985-0.997) in patients with CRT, following adjustment for age, sex, non-ischaemic cardiomyopathy, LVEF, atrial fibrillation/flutter, NHYA class, and peripheral vascular disease (Figure 1A). Similar findings were seen in the group without LBBB at baseline (HR 0.992, 95% CI 0.985-0.999) (Figure 1B). Amongst patients receiving CRT for left ventricular systolic dysfunction with QRS prolongation, any degree of QRS narrowing was found to be associated with an increasing reduction in mortality and heart failure, irrespective of the QRS morphology at baseline. This finding may improve outcomes in patients undergoing CRT implantation, particularly in those with a non-LBBB QRS morphology.