A total of 478 consecutive patients who had undergone TAVR for symptomatic severe AS at Severance Cardiovascular Hospital from June 2011 to May 2021 were retrospectively reviewed. We excluded patients with pre-existing intraventricular conduction disturbances (n = 77) (QRS >120 ms, LBBB, and right bundle branch block) before TAVR and patients who had received permanent pacemaker implantation before or during the index hospitalization for TAVR (n = 34). There was no patient with previously implanted intracardiac cardioverter-defibrillator or cardiac resynchronization therapy. The present study also excluded three patients who did not survive immediately after TAVR. Thus, 364 patients were included in the final analysis in this study.

Surgical risk was estimated using the European System for Cardiac Operative Risk evaluation (EuroSCORE II) and the Society of Thoracic Surgeons Predicted Risk of Mortality (STS-PROM) score (11, 12). The decision to use TAVR as the treatment modality was made by a multidisciplinary heart team, as previously reported (13, 14). The transcatheter aortic valve type was chosen at the discretion of the operators based on the anatomical characteristics of the aortic valve, aortic root, and vascular access. The Institutional Review Board of Severance Hospital approved this study, which was conducted in compliance with the Declaration of Helsinki. The requirement for informed consent was waived because of the retrospective study design.

LBBB was defined as a QRS duration > 120 ms, delayed onset of intrinsicoid deflection in leads V5 and V6, broad monophasic R waves that are usually notched in leads I, V5 and V6, and secondary ST- and T-wave changes opposite in direction to the major QRS deflection (15). In this study, new-onset LBBB was defined as persistent LBBB developed during or after the TAVR procedure and documented on the electrocardiogram (ECG) at hospital discharge or 7 days after TAVR.

All echocardiographic studies were performed using commercially available equipment and were reviewed by imaging cardiologists without knowledge of the clinical data. Standard measurements were performed according to current guidelines (16). LV EF was measured using linear measurement or biplane methods. The LV mass index was calculated using the Devereux formula. The left atrial volume index (LAVI) was calculated using the biplane method. Pulmonary artery systolic pressure and right atrial pressure were estimated using tricuspid regurgitation jet velocity and inferior vena cava (16). Right ventricular (RV) systolic dysfunction was defined as in case of tricuspid annular plane systolic excursion <17 mm, tricuspid pulsed Doppler S wave <9.5 cm/s, or fractional area change <35% (16). Preprocedural and postprocedural AV hemodynamic parameters such as the AV peak flow velocity, transaortic pressure gradient, and aortic valve area were calculated using Doppler echocardiography (17). Concomitant at least moderate mitral or tricuspid regurgitation was defined as other valve pathology. The severity of paravalvular regurgitation was semi-quantitatively assessed according to recent recommendations (17). Patients underwent baseline echocardiography before TAVR and regular planned examinations annually after TAVR, according to standard institutional follow-up protocol. To investigate the impact of LBBB on reverse cardiac remodeling, baseline and 1-year echocardiographic parameters were compared according to the presence of LBBB.

The primary outcome was a composite of cardiovascular death or hospitalization for heart failure. Secondary outcomes included all-cause death, cardiovascular death, hospitalization for heart failure, number of hospitalizations for heart failure event, and permanent pacemaker implantation. All clinical outcomes were analyzed according to the Valve Academic Research Consortium-3 consensus (18).

Categorical variables were presented as numbers (percentages) and were compared using chi-squared test or Fisher's exact test. Continuous variables were presented as means ± standard deviation and compared using Student's t-test or the Wilcoxon rank-sum test. Time-to-event variables were presented as Kaplan-Meier event rates and were compared using the log-rank test. The total number of hospitalizations for heart failure was calculated and compared using Poisson regression. Multivariable analysis for clinical outcomes was performed using a multivariable Cox proportional hazard regression model. The covariates included in the adjusted models were variables with clinical relevance, such as age, sex, New York Heart failure Association (NYHA) functional class, comorbidities such as chronic lung disease, end-stage renal disease, coronary artery disease, peripheral artery disease, prior cardiac surgery, atrial fibrillation, EuroSCORE II, STS-PROM score, baseline LV EF, RV systolic dysfunction, pulmonary artery systolic pressure and moderate or severe paravalvular regurgitation. The baseline and 1-year echocardiographic parameters were compared using paired t-test or the Wilcoxon signed-rank test as appropriate. Two-way repeated ANOVA was used to determine differences between the baseline and 1-year echocardiographic parameters according to the study groups. Missing data of 1-year echocardiographic data was not imputated. As sensitivity analysis, baseline and 1-year echocardiographic data was compared with multiple imputation of missing data. All the tests were two-tailed, and p-values <0.05 were considered statistically significant. All statistical analyses were performed using R version 4.1.0 (The R Foundation for Statistical Computing; www.R-project.org).

Of the 364 patients, 41 (11.3%) had new-onset persistent LBBB after TAVR. The baseline clinical characteristics are shown in Table 1. The two groups, new LBBB group and no LBBB group, were similar in sex distribution, symptom severity, comorbidities, and the surgical risk score. The new LBBB group was younger than the no LBBB group, but the difference did not reach statistical significance. The new LBBB group had a higher prevalence of prior cardiac surgery. The choice of transcatheter aortic valve was not statistically different between the two groups; however, the new LBBB group had a trend toward more frequent use of self-expandable valves than the no LBBB group. Predilation rate was low in the new LBBB group, and the degree of paravalvular regurgitation was comparable in both groups.

The echocardiographic characteristics are summarized in Table 2. The baseline echocardiographic parameters were comparable between the new LBBB and no LBBB groups. One-year follow-up echocardiograms were obtained in 264 (73%) patients. At the 1-year follow-up after TAVR, both groups showed improved and similar AV hemodynamic parameters, such as the peak velocity, pressure gradient, and AV area; however, the new LBBB group showed a lower LV EF (59.1 ± 13.7 vs. 65.8 ± 9.6%; p = 0.018), a larger LV end-diastolic dimension (48.6 ± 5.5 vs. 46.5 ± 5.0 mm; p = 0.037) and end-systolic dimension (33.1 ± 6.7 vs. 30.3 ± 5.0 mm; p = 0.035), and higher E/e′ (28.2 ± 15.1 vs. 20.5 ± 7.9; p = 0.029) than the no LBBB group.

When the baseline and 1-year echocardiographic parameters were compared, the no LBBB group showed a significantly increased LV EF and a decreased LV end-systolic dimension, LV mass index, and left atrial (LA) volume index (all p < 0.001; Figure 1). However, the new LBBB group had significantly decreased LV EF (−6.0 ± 14.5%; p = 0.038) at the 1-year follow-up and no significant changes in the LV end-systolic dimension, LV mass index, and LA volume index. The sensitivity analysis with multiple imputation for 1-year echocardiographic data showed similar results (Supplementary Table 1 and Supplementary Figure 1). In the subgroup of patients with baseline LV systolic dysfunction (LV EF ≤ 50%), the new LBBB group showed no significant change in the LV EF (+8.2 ± 19.9%; p = 0.408), whereas the no LBBB group showed significant LV EF improvement (+20.5 ± 14.5%; p < 0.001) 1 year after TAVR. However, in patients with a preserved LV EF (>50%), the new LBBB group showed a significantly decreased LV EF (−9.0 ± 11.4%; p < 0.002) from the baseline while the no LBBB group had a similar LV EF (+1.0 ± 8.1%; p = 0.092) after TAVR (Supplementary Figure 2).

Patients were followed for a median of 18.1 months (interquartile range: 7.7–30.1). All the clinical outcomes after TAVR and hazard ratios for the adverse clinical events are described in Table 3 and Figure 2. The new LBBB group showed a higher rate of primary composite outcome events (cardiovascular death or hospitalization for heart failure) than the no LBBB group (HR: 5.03; 95% CI: 2.60–9.73; p < 0.001). The new LBBB group had a higher risk of all-cause death (HR: 2.80; 95% CI: 1.38–5.69; p = 0.003), individual events of cardiovascular death (HR: 7.34; 95% CI: 2.35–22.93; p < 0.001), and hospitalization for heart failure (HR: 5.25; 95% CI: 2.57–10.75; p < 0.001). Furthermore, the new LBBB group had more hospitalizations for heart failure (29.4 vs. 5.1 events per 100-person year; p < 0.001) and PPM implantation than the no LBBB group (HR: 5.44; 95% CI: 1.21–24.52; p = 0.010). There was no post-procedural CRT implantation. After multivariable adjustment, the patients with new-onset LBBB still had a significantly higher risk for adverse clinical outcomes, with the exception of PPM implantation. In Cox multivariate regression analysis, new-onset persistent LBBB, end-stage renal disease, atrial fibrillation, and prior caridac surgery were identified as independent predictors for the composite events of cardiovascular death and hospitalization for heart failure (Table 4).

This study has several limitations. First, this study was a single-center retrospective study, which has inherent limitations. Second, the number of subjects with new LBBB was too small for detailed subgroup analysis. Third, this study included subjects with an advanced age and at variable surgical risk, possibly limiting the generalization of our study results to younger patients, who are at a lower surgical risk. Finally, 1-year follow-up echocardiography data were not available for all patients due to the early occurrence of clinical outcomes and the advanced age of the study population. However, we performed the multiple imputation for missing 1-year echocardiographic data and found similar results with main findings.