Risk stratification of left bundle branch block after transcatheter aortic valve implantation: electrophysiological study versus ambulatory monitoring

Background

Left bundle branch block (LBBB) is a common conduction disturbance after transcatheter aortic valve implantation (TAVI) and has been associated with delayed high-grade AV block and increased permanent pacemaker implantation (PPI). Management varies between invasive evaluation with electrophysiological study (EPS) and conservative approaches employing ambulatory continuous ECG monitoring (AECG). The optimal approach balancing safety, diagnostic yield, and resource utilization remains undefined.

Aim

To compare EPS versus AECG for risk stratification of patients with persistent LBBB after TAVI.

Methods

We conducted a single-centre retrospective cohort study of all patients who underwent TAVI between January 2020 and January 2025. Patients with persistent LBBB before discharge were managed either with EPS or AECG for 7–15 days, according to availability and clinical judgement. Baseline clinical, procedural, and PPI data were recorded.

Group comparisons used χ² or Fisher’s exact tests for categorical variables and t-tests or Wilcoxon rank-sum tests for continuous variables, according to data distribution. Negative predictive value (NPV) for late PPI was calculated for each strategy among patients with normal EPS or AECG.

Results

Of 1076 TAVI patients, 131 with persistent LBBB underwent risk stratification (EPS n=53, AECG n=78). Median age was 81 years; 62% were female, 30% received balloon-expandable valves and 24% had atrial fibrillation. Baseline comorbidities were similar, but EPS patients had longer mean PR (192 vs. 181 ms, p < 0.001) and QRS (157 vs. 149 ms, p = 0.04) intervals.

After a median follow-up of 26 months, PPI was significantly more frequent in the EPS group (38% vs. 10%, p < 0.001), reflecting the higher-risk profile of patients selected for EPS. In this group, most PPI procedures were performed at baseline due to a prolonged HV interval (90%). Among 35 patients with normal EPS, 2 (6%) required late PPI (NPV 94%). Both cases were older women treated with self-expanding valves, admitted with paroxysmal advanced AV block.

In the AECG group, 4 patients (5%) had high-grade AV block detected early after discharge, leading to prompt PPI; 2 had severe presentations requiring emergency assessment, including 1 with cardiopulmonary arrest. Extending AECG beyond 7 days did not improve diagnostic yield. Among 74 patients with normal AECG, 4 (5%) required late PPI (NPV 95%).

Conclusions

Both strategies showed high NPV for late PPI after TAVI. EPS allows early identification of patients at higher risk, but likely incurs higher costs and more frequent PPI. AECG carries a residual risk of severe late-presenting bradyarrhythmias, although most clinically significant events were detected early after discharge. These findings support a tailored approach based on patient risk profile and local resources.