Cardiac memory following cardiac resynchronization therapy predicts survival in heart failure patients with left bundle branch block

Purpose: Cardiac memory (CM) is considered to reflect electrical remodeling and has been described following right ventricular pacing or intermittent LBBB. We investigated whether CM occurs following cardiac resynchronization therapy (CRT) and whether it influences outcome. Methods: In 41 CRT patients (>95% BiV pacing) 12 lead ECGs were recorded during intrinsic rhythm after interrupting CRT pacing (ECG POST) and compared to the pre-implant ECG (ECG PRE). Mean spatial angles (MSA) between QRS and T vectors were determined with the Kors conversion method. LVEF was determined at CRT POST. Cardiac mortality was the primary endpoint. Results: 41 patients (age 69±11 years, all LBBB PRE CRT, 56% ischemic cardiomyopathy) were included in the study. Following CRT, LVEF improved significantly from 26±10 to 36±14% (p=0.01). Follow-up started after recording of ECG POST and lasted 30±2 months. 9 patients (22%) died. Univariate Cox regression revealed a higher cardiac mortality in the patients with lower LVEF POST CRT (HR 1.10, p<0.01), larger MSA between QRS and T vectors POST CRT (HR 1.03, p0.03), smaller MSA between QRS vectors PRE and POST CRT (HR 0.97, p0.03) and smaller MSA between T vectors pre and post CRT (MSA Tpre-Tpost) (HR 0.96, p<0.01). Using multivariate analysis, only a smaller MSA Tpre-Tpost predicted cardiac death (HR 0.97, p0.02), independent from worse LVEF POST CRT. At a cut-off value of 45°, MSA Tpre-Tpost showed high sensitivity (84%) and specificity (78%) (fig.1) (ROC AUC 0.81, p<0.01). ![Figure][1] Kaplan Meier Plot for MSA TpreTpost Conclusion: A greater change of the mean spatial T-angle after CRT pacing results in significantly higher cardiac survival. This parameter adds prognostic value independent from structural remodeling and therefore deserves further investigation to better understand response to CRT therapy. [1]: pending:yes