Prognostic impact of new-onset atrial fibrillation and left bundle branch block after TAVR: a propensity-matched landmark analysis

Background

New-onset atrial fibrillation (NOAF) and left bundle branch block (LBBB) frequently occur after transcatheter aortic valve replacement (TAVR). Although both have been linked to adverse outcomes, prior studies are limited by residual confounding and time-dependent bias, leaving uncertainty as to whether these conditions represent independent prognostic factors or simply markers of higher baseline risk.1–3

Purpose

To evaluate the association of NOAF and new-onset LBBB detected at 7 days post-TAVR by 12-channel ambulatory ECG or 24-hour Holter ECG with all-cause mortality using a propensity score–matched landmark analysis.

Methods

This retrospective observational cohort study included patients undergoing TAVR at a high-volume tertiary care center between January 2012 and December 2023 (n = 2,488). The population was elderly (mean age 82 years, 95% CI 79–83), with mean left ventricular ejection fraction (LVEF) 53.3% (95% CI 52.7–53.8) and EuroSCORE II 6.1% (95% CI 5.8–6.3). Most were symptomatic at baseline, with 82% in NYHA class II–III. Mean follow-up was 39 months (95% CI 33–45).

A 7-day landmark analysis was performed to mitigate immortal time bias; NOAF and LBBB were assessed at this time point, including only patients alive and event-free. Propensity scores were estimated using logistic regression (age, sex, body mass index, EuroSCORE II, NYHA class, baseline LVEF), and patients were matched 1:1 using nearest-neighbour matching without replacement. Covariate balance was assessed using standardised mean differences (<0.1), yielding matched cohorts of 1,193 patients for NOAF and 916 for LBBB. The primary endpoint was all-cause mortality. Survival was analysed using Kaplan–Meier estimates and Cox proportional hazards models.

Results

In landmark analyses, NOAF was associated with significantly higher mortality (log-rank p = 0.001; HR 1.46, 95% CI 1.23–1.75). This association persisted after multivariable adjustment in the unmatched cohort (adjusted HR 1.38, 95% CI 1.16–1.65; p < 0.005). Addition of NOAF significantly improved model fit (likelihood-ratio χ² = 46.7, p < 0.001), despite only a modest increase in discrimination (C-index 0.614 to 0.622).

In contrast, LBBB at 7 days was not associated with mortality in the matched cohort (log-rank p = 0.712; HR 0.95, 95% CI 0.73–1.24) or after multivariable adjustment in the unmatched cohort, including among patients with LVEF < 40% (p = 0.811). Furthermore, LBBB was not associated with new pacemaker implantation (adjusted OR 1.00, 95% CI 0.897–1.028; p = 0.694).

Conclusions

In an elderly TAVR population, NOAF at 7 days post-TAVR was independently associated with adverse long-term outcomes and provided incremental prognostic information beyond established risk factors. In contrast, LBBB at 7 days was not associated with adverse outcomes. Prospective randomised studies are warranted to determine whether targeted management of NOAF improves clinical outcomes.Kaplan-Meier Estimate for NOAFKaplan-Meier Estimate for New-LBBB