Abstract 16755: Personalized Virtual Cardiac Resynchronization Therapy Simulation Based on a 12-Lead ECG and CMR Predicts Electrical Activation Changes After Cardiac Resynchronization Therapy

Abstract

Introduction: Simulating electrical remodeling after cardiac resynchronization therapy (CRT) based on individual patient characteristics could enhance the effectiveness of the procedure, help select optimal CRT candidates, and yield important prognostic data. Hypothesis: A Virtual CRT (V-CRT) simulation based on the 12-lead ECG and cardiac magnetic resonance (CMR) for evaluation of biventricular structure and extent of scar with late gadolinium enhancement (LGE) would predict changes in electrical activation after CRT. Methods: V-CRT uses cardiac geometry and scar (LGE) from CMR, a torso avatar from CMR scouts, and an electrophysiology model derived from the baseline ECG, the left ventricular (LV) pacing site, and programmed CRT pacing parameters. Ventricular activation times and the resulting QRS duration (QRSd) after CRT were estimated. The predicted QRS durations were then compared with the actual QRS durations after CRT pacing. Results: 20 CRT patients (age 67.3 ± 10.2 years, 30% female, 50% with LV scar) had a mean pre-CRT QRSd of 151 ± 12.8 ms. This predicted post-CRT QRSd depended significantly on the LV pacing site with a median maximum-minimum difference per patient over all possible pacing sites of 39.8 ms (IQR 32.5-44.2 ms). The post-CRT QRSd predicted by V-CRT based on the actual LV pacing site correlated well with the observed post-CRT QRS duration (r=0.51, p=0.02) (Figure). Predicted changes in the QRS duration were strongly associated with changes in LV activation times (r=0.74, p=0.0002), but not RV activation times (no correlation). Greater predicted reductions in LV activation times were more common in patients with greater LV end-diastolic volume indices (LVEDVIs) by CMR (r=-0.58, p=0.01). Conclusions: V-CRT effectively predicts changes in electrical activation after CRT. These findings have important implications for patient selection for CRT, prognosis after CRT, and CRT implementation strategies.