Periprocedural implant predictors for successful left bundle branch pacing
F Yang, S Ng, X Y Teoh, K Tan, X J Ge, P L Chia, D FooAbstract
Background/Introduction
Left bundle branch pacing (LBBP) is widely adopted as an established physiological pacing method. Contemporary criteria for successful LBBP focuses on final paced parameters, particularly stimulation to LV activation time (LVAT) and presence of incomplete RBBB morphology transition. Data on periprocedural implant predictors of left bundle branch (LBB) capture remain limited.
Purpose
This study aims to discover additional periprocedural implant parameters that may serve as novel predictors of successful LBB capture.
Methods
192 patients who underwent LBBP were analysed, success being defined as selective or non-selective LBB capture, indicated by LVAT ≤ 75ms in patients with normal or right bundle branch block morphology, or ≤ 85ms in those with baseline left bundle branch block, together with an incomplete right bundle branch pattern in V1 and a V1–V6 inter-peak interval ≥ 33ms. Initial LVAT (prior to septal lead advancement) was measured in addition to final LVAT measurement. Baseline characteristics, left ventricular ejection fraction (LVEF), pacing indications, QRS morphology, and intrinsic QRS duration were also recorded. Univariate and multivariate analyses were performed to identify successful predictors by logistic regression. ROC curve analysis was performed to identify optimal cutoffs from significant predictors.
Results
A total of 192 patients (mean age 74.3 ± 9.4, male 56.2%, LVEF 53.7 ± 9.8%, BMI 23.9 ± 4.7 kg/m²) were included (Table 1). Pacing indications included sick sinus syndrome (29.7%), atrioventricular (AV) block (66.1%), and the rest with binodal disease, atrial fibrillation post AV node ablation or cardiac resynchronisation therapy (4.2%). Baseline QRS morphology was normal in 59.9%, RBBB in 27.6%, LBBB in 7.8%, and alternating bundle branch block in 4.7%, with a mean intrinsic QRS duration of 117.0 ± 29.6ms. Successful LBBP was achieved in 72.9% (23.4% selective, 76.6% non-selective). Both initial LVAT and LVAT/QRS ratio had a significant positive correlation with successful LBBP. ROC analysis identified an optimal initial LVAT cutoff ≤ 96.5ms (AUC 0.73, sensitivity 75.7%, specificity 63.5%) for predicting successful final LBB capture (Figure 1). Initial LVAT had better discriminative performance as compared to LVAT/QRS ratio (AUC 0.61) , the initial LVAT cutoff ≤ 96.5ms performed similarly among all different subgroups of baseline QRS morphologies.
Conclusions
Initial LVAT correlated positively with successful LBBP, with a cut-off of ≤ 96.5ms predicting successful final LBB capture with sensitivity 75.7%, specificity 63.5%. Larger prospective studies are warranted to validate these findings, as these novel predictors may help streamline implantation strategy, reduce procedural time, and enhance overall success rates.