Atrial Fibrillation at TAVR and the Clinical Significance of Permanent Pacing in Contemporary Practice

Atrial fibrillation is common in TAVR and pacemaker decisions are frequent [1, 2], but the field still argues about whether permanent pacemaker implantation (PPI) worsens long-term outcomes [3, 4]. The study by Osovsky [5], recently published in Cardiology, isolates patients with preexisting AF and asks two focused questions: who needs a pacemaker, and does PPI carry a mortality penalty over 3 years. The answers are pragmatic. Conduction risk is still anatomy and platform; mortality tracks age, not PPI [6, 7].This was a single-center historical cohort of 1,579 TAVR patients (2008–2022); 405 (25.6%) had prior AF (17.1% paroxysmal, 8.4% persistent). Mean age in AF was 82 ± 6 years, 49.9% male. Thirty-day PPI among AF patients was 21.7%. Independent predictors of 30-day PPI were preexisting right bundle branch block (RBBB; aOR: 5.34, 95% CI: 2.79–10.21) and the CoreValve/Evolut self-expanding platform (aOR: ∼2.5). In the subgroup with AF rhythm on admission, deeper implantation independently increased PPI risk (OR: 1.12 per mm). PPI was not associated with 3-year mortality; older age was the dominant predictor (aHR: ∼1.55 per decade). Kaplan-Meier curves showed no mortality separation by PPI status (shown in Fig. 1).What should readers take from this? First, AF does not alter the basic conduction-risk playbook: baseline RBBB remains the strongest signal; self-expanding supra-annular valves raise PPI risk and implant depth matters when AF is present at the time of implant. Second, in this dataset PPI itself did not translate into higher 3-year mortality. That runs against a common assumption in counseling and supports a shift in endpoints away from mortality alone toward heart-failure events, pacing burden, and remodeling.The analytic choices are sensible for the questions asked. Logistic regression identified 30-day PPI predictors; Cox regression with a 30-day landmark handled longer-term mortality to avoid early survival bias. Proportional hazards were assessed; missingness was modest overall, though pre-procedural CT was unavailable in 19% due to era effects, limiting assessment of membranous septum length and annulus-to-valve ratio in the full cohort. Device programming and pacing percentages were not captured, which constrains inference on mechanism. Still, platform signals, RBBB, and depth are robust and consistent with prior sinus-rhythm literature [8, 9], now confirmed in an AF cohort of meaningful size.Why might PPI not increase mortality here? One explanation is substrate: in a very old cohort, competing risks dominate and any adverse pacing effects are diluted by age-driven hazards. Another is practice evolution: cusp-overlap, shallower targets, and selective platform choice may have reduced injurious conduction trauma over time, even while PPI rates remained nontrivial. A third is outcome choice: mortality is blunt. If right ventricular pacing burden drives heart failure rather than death, then mortality will miss clinically relevant differences without device-level data. The study’s own univariate findings hint at care signals: post-procedure AF rhythm and acute kidney injury tracked with death before adjustment, again consistent with non-pacing drivers of risk.The platform and depth signals deserve to change practice tomorrow. Baseline ECG screening for RBBB and QRS duration should be routine pre-TAVR triage in AF. Where anatomy allows, a balloon-expandable valve may reduce PPI risk compared with self-expanding devices. If a self-expanding valve is chosen, depth targets should be conservative, using cusp-overlap views and intraprocedural attention to conduction changes. The subgroup analysis is useful: in AF rhythm at admission, every millimeter deeper made PPI more likely, that is, both a caution and an opportunity for teams to tighten implantation targets.The ECG phenotype is also actionable. RBBB increased absolute PPI risk dramatically (about 52% vs. 17% without RBBB; number needed to “harm” ∼3). This is the conversation to have with patients: with preexisting RBBB, pacemaker risk is high regardless of AF type; platform and depth choices can modify but not erase that risk. Post-procedure surveillance should prioritize conduction changes in this subgroup and avoid unnecessary delays in pacing when high-grade block emerges.What about outcomes we are not measuring? If PPI does not move mortality here, it may still influence heart-failure hospitalization, left ventricular function, and quality of life through pacing burden. Follow-up should also optimize contemporary HF therapy – including SGLT2 inhibitors, whose cardiorenal actions extend beyond glucosuria [10] – to reduce HF events regardless of pacing status. The registry did not report device programming, %RV pacing, or adoption of conduction-system pacing (CSP). Those details are increasingly central. In AF with anticipated high pacing burden, early CSP may mitigate dyssynchrony and heart-failure risk compared with standard right ventricular apical pacing. Without device-level data, mortality neutrality should not be over-read as clinical neutrality.Generalizability needs a clear boundary. This is a high-volume center across 15 years, with evolving platforms and techniques; CT became routine partway through the period. Decision processes for PPI and thresholds for implantation likely varied over time. The authors note these constraints; they also show that their PPI rates align with external series, which supports face validity. Still, era and technique confounding can bias both the predictors and the mortality analysis, especially if later-era patients received more contemporary valves, shallower depth targets, or different pacing algorithms. Calendar-time-adjusted sensitivity analyses would strengthen confidence in neutrality of PPI on mortality.How should this study change how we counsel and operate? First, reset expectations: in AF patients, PPI after TAVR is common; it is chiefly predicted by RBBB and platform/depth; and in this study it did not increase 3-year mortality. That supports more precise risk communication and less reflexive pessimism when a pacemaker is needed. Second, prevention remains worthwhile: platform selection and surgical-style attention to implantation depth can reduce PPI without sacrificing valve performance. Third, follow-up should be tailored: in patients with high expected pacing burden, consider pathways that evaluate CSP, monitor %RV pacing, and track heart-failure status closely.The research agenda is straightforward. First, multicenter validation with mandatory CT metrics (membranous septum length, annulus-to-valve ratio), documented implantation depth, and predefined depth targets by platform is needed. Second, calendar-time-adjusted models are needed to account for adoption of cusp-overlap and newer valve iterations. Third, device-level follow-up with programming, %RV pacing, and CSP use is needed to connect PPI to heart-failure trajectories. Fourth, emulated target trials in AF with RBBB should test platform selection or depth targets against PPI and clinical outcomes. These steps would turn a clear clinical message into a tested pathway.The bottom line is practical. Expect high pacemaker rates in AF after TAVR; reduce risk where you can with platform choice and shallow, precise implantation; and stop assuming the device itself portends worse survival. In this cohort, mortality hinged on age and intercurrent illness, not PPI. Measure what matters to patients: heart-failure events, pacing burden, and function and invest procedural attention where it changes the conduction story: RBBB at baseline, platform choice, and implantation depth.The authors have no conflicts of interest to declare.This study was not supported by any sponsor or funder.Nils Grimm: writing – original draft and writing – review and editing. Stephan Peronard Mayntz: conceptualization and writing – review and editing.