Prediction of optimal cardiac resynchronization by vectors extracted from electrograms in dyssynchronous canine hearts

Introduction: Proper optimization of atrioventricular (AV) and interventricular (VV) intervals can improve the response to cardiac resynchronization therapy (CRT). It has been demonstrated that the area of the QRS complex (QRSarea) extracted from the vectorcardiogram can be used as a predictor of optimal CRT-device settings. We explored the possibility of extracting vectors from the electrograms (EGMs) obtained from pacing electrodes and of using these EGM-based vectors (EGMVs) to individually optimize acute hemodynamic CRT response. Methods and results: Biventricular pacing was performed in 13 dogs with left bundle branch block (LBBB) of which five also had myocardial infarction (MI), using 100 randomized AV- and VV-settings. Settings providing an acute increase in LV dP/dt_max ≥ 90% of the highest achieved value were defined as optimal. The prediction capability of QRSarea derived from the EGMV (EGMV-QRSarea) was compared with that of QRS duration. EGMV-QRSarea strongly correlated to the change in LV dP/dt_max (R = −0.73 ± 0.19 [LBBB] and −0.66 ± 0.14 [LBBB + MI]), while QRS duration was more poorly related to LV dP/dt_max changes (R = −0.33 ± 0.25 [LBBB] and −0.47 ± 0.39 [LBBB + MI]). This resulted in a better prediction of optimal CRT-device settings by EGMV-QRSarea than by QRS duration (LBBB: AUC = 0.89 [0.86–0.93] vs. 0.76 [0.69–0.83], P < 0.01; LBBB + MI: AUC = 0.91 [0.84–0.99] vs. 0.82 [0.59–1.00], P = 0.20, respectively). Conclusion: In canine hearts with chronic LBBB with or without MI, the EGMV-QRSarea predicts acute hemodynamic CRT response and identifies optimal AV and VV settings accurately. These data support the potency of EGM-based vectors as a noninvasive, easy and patient-tailored tool to optimize CRT-device settings.