Abstract 14586: Total Scar Burden Predicts Cardiac Resynchronization Therapy (CRT) Response Beyond Left Bundle Branch Block (LBBB) and QRS Duration (QRSd) and Regardless of the Type of Cardiomyopathy (CM)

Introduction: Prior studies have demonstrated that up to 30% of patients do not benefit from CRT and that response may be affected by the extent and location of myocardial scar. We sought to determine the predictive value of the scar burden on CRT response after controlling for the type of cardiomyopathy (Ischemic=ICM vs. Non-Ischemic=NICM), the presence of LBBB and QRSd. Methods: We reviewed 71 pts (49% NICM, age 71 +/- 11, 34% women) who underwent cardiac MRI (CMR) pre-CRT. Total and lateral wall % scar were measured and pre-CRT ECG features (LBBB and QRSd) were noted. The change in EF on echo >6 months post-CRT was calculated and a positive response was defined as a ≥10% increase in left ventricular ejection fraction (EF). The relationship between LBBB, QRSd, total and lateral scar % and change in EF were investigated using Pearson (r) correlation. Hierarchical multiple regression was used to assess the ability of total and lateral scar burden to predict CRT response after controlling for the type of CM, the presence of LBBB and QRSd. Results: Mean EF pre-CRT was 23.6%+/-8 and correlated significantly with EF by CMR (mean =24.4 +/-10; r = 0.5, p<0.01). LBBB was present in 52% of pts and mean QRSd was 152ms. A positive response to CRT was noted in 52% of pts. The change in EF correlated significantly with the presence of LBBB (0.31; p=0.01), QRSd (0.27; p=0.03); total scar % (-0.39; p<0.01) and lateral scar % (-0.28;p=0.02). After controlling for the type of CM, and presence of LBBB and QRSd, total scar % was still significantly predictive of the response to CRT [R squared change (tot scar) = 0.06, F change (1,63) = 16, p=0.01] - see table 1. With lateral scar, the same analysis was not significant [R squared change (lat scar)=0.04, F change (1,62)=3.5 p=0.06] Conclusion: Assessment of total scar burden provides incremental predictive value to CRT response after controlling for the type of CM (ICM vs. NICM), the presence of LBBB and QRSd.