Pacing dependency after TAVI, a lasting hazard in every case?

Abstract Introduction Atrioventricular (AV) conductions abnormalities that require permanent pacemaker (PPM) occur in up to 20% of patients after TAVI. The evolution of conduction abnormalities and the long-time pacing burden have not been fully characterized. We aim to characterize ventricular pacing burden after TAVI assessing also the impact of clinical and periprocedural factors. Methods Single centre retrospective registry including all consecutive patients undergoing TAVI between 01/2014 and 08/2021 that had a PPM implanted during index admission. Patients with in-hospital death were excluded. Baseline and pacing follow-up data were collected prospectively and assessed retrospectively using local electronic health records. We examined pacing burden as the percentage of ventricular pacing (VP%) and divided our cohort in 3 groups: i VP% ≤10%, ii 10% > VP% < 90%, iii VP% ≥90%. We assessed the difference between discharge and 1y VP% and the interaction between baseline characteristics. Results 766 patients underwent TAVI and 109 (14.2%) were included. The median age was 84 years (IQR 80 ; 90) and 44% were male (n = 48). 19 patients (17.4%) had previous 1st degree AV block (AVB) and 37 patients (33.9%) had previous right bundle branch block (RBBB). A self-expandable valve was implanted in 78 patients (71.6%). The median time between TAVI and PPM was 2 days (1;5). Sustained complete AVB (cAVB) occurred in 85 patients (78.0%) and was the most common indication for PPM. Data on VP% at discharge was available for 102 patients (93.5%). The median VP% at discharge was 100% (99 ; 100) with 4 patients (3.7%) having ≤10% VP% and 87 patients (79.8%) having ≥90% VP%. 95 patients (87.2%) completed the 1y follow-up. The VP% decreased significantly between discharge and 1y follow-up - 8.0 (IQR -59 ; 0 / p <0.05 Wilcoxon test) (Figure 1), similarly across valve type. At 1y follow-up, 20 patients (18.3%) had ≤10% VP% and 43 patients (39.4%) had ≥90% VP%. In the 18 patients (19.4%) without periprocedural sustained cAVB, 9 patients (50.0%) had ≤10% VP% and 1 patient (5.6%) had ≥90% VP%. In contrast, among patients with sustained cAVB, 11 patients (14.7%) had ≤10% VP% and 40 patients (53.3%) had ≥90% VP% (p <0.05 Chi-squared) (Figure 2). Conclusions In this study cohort the pacing burden significantly decreased in the 1st year after TAVI. It was significantly lower in patients without periprocedural sustained cAVB which suggests that a more conservative management may be considered in a small subgroup of patients.