Site-Specific Differences in Latency Intervals during Biventricular Pacing: Impact on Paced QRS Morphology and Echo-Optimized V-V Interval

Methods: We recorded digital 12-lead electrocardiograms in 40 CRT patients during RV, LV, and biventricular pacing at three output settings. Stimulus-to-earliest QRS deflection (latency) intervals were measured in all leads. Echocardiographic atrioventricular (AV) and V-V optimization was performed using aortic velocity time integrals. Results: Latency intervals were longer during LV (34 +/- 17, 29 +/- 15, 28 +/- 15 ms) versus RV apical pacing (17 +/- 8, 15 +/- 8, 13 +/- 7 ms) for threshold, threshold x3, and maximal output, respectively (P < 0.001), and shortened with increased stimulus strength (P < 0.05). The echo-optimized V-V interval was 58 +/- 31 ms in five of 40 (12%) patients with LV latency >= 40 ms compared to 29 +/- 20 ms in 35 patients with LV latency < 40 ms (P < 0.01). During simultaneous biventricular pacing, four of five (80%) patients with LV latency >= 40 ms exhibited a left bundle branch block (LBBB) pattern in lead V-1 compared to three of 35 (9%) patients with LV latency < 40 ms (P < 0.01). After optimization, all five patients with LV latency >= 40 ms registered a dominant R wave in lead V-1. Conclusions: LV pacing from the lateral cardiac vein is associated with longer latency intervals than endocardial RV pacing. LV latency causes delayed LV activation and requires V-V interval adjustment to improve hemodynamic response to CRT. Patients with LV latency >= 40 ms most often display an LBBB pattern in lead V-1 during simultaneous biventricular pacing, but a right bundle branch block after V-V interval optimization. (PACE 2010; 1382-1391).