Background: Catheter ablation targeting isolation of the pulmonary veins (PVI) is the most effective rhythm control treatment for atrial fibrillation (AF). Despite its high overall effectiveness, one or more repeat AF ablation procedures (re-do procedures, RDP) are often required to maintain sinus rhythm.

Purpose: To determine predictors associated with multiple or complex RDP, to evaluate reference values for radiation exposure during PVI, and to estimate the risk of intraprocedural pericardial effusion (PE) for index PVI (iPVI) and RDP.

Method: Data mining was applied to identify all AF patients with at least one RDP performed in a large German AF ablation center between September 2008 and September 2021. Procedures were classified for procedure type by an algorithm based on regular expressions. Index PVI procedures were classified as PVI-only or PVI with additional substrate modification (SM). RDP were classified as repeat PVI (Re-PVI) due to reconduction (PV reconduction), ablation of atrial tachycardia (AT) or SM, e. g. defragmentation of fractionated signals, or combinations. All automatically classified results were manually quality controlled. From normally distributed scalars, outliers were removed based on distance to mean greater than three standard deviations. Patients were clustered based on diagnosis at iPVI and type of iPVI.

Results: 934 out of 6848 AF patients (mean age 62.6 ± 12.3 years, 346 females) underwent 2152 iPVI and at least one RDP (out of 8750 total AF-related ablations). At iPVI, AF pattern was classified as paroxysmal AF (PAF) in 387 patients (41%). All others (59%) were classified as non-paroxysmal AF (Non-PAF). Non-PAF was significantly more frequent in males than in females (64% vs. 49%, p<0.01). Median period between first PVI and RDP was 558 days (25th/75th percentiles 244.0/1175.5 days). 724/934 patients (78%) received PVI-only as initial procedure. Of these, 572 (79%) had only 1 RDP, 116 (16%) had 2 RDP and 36 (5%) had 3 or more RDP. This distribution was 77%, 15% and 8% for 1, 2 and 3 or more RDP for patients with complex PVI as iPVI. Non-PAF patients had a significantly higher probability of multiple RDP compared to patients with PAF at iPVI (p<0.01, Figure 1A). 18% (8%) of patients with non-PAF had 2 (3) or more RDP while only 13% (3%) of pat. with PAF had 2 (3) or more RDP. 798/934 (85%) patients required PV re-isolation due to PV reconduction, 298/934 (32%) required ablation for atrial tachycardia (AT) at least once during FU (Figure 1B). Comparing PVI-only iPVI patients with patients who received substrate modification during iPVI, significantly less patients with PVI-only iPVI had RDP for AT compared to those with additional lesions during iPVI (27% vs. 50%, p<0.01). On the other hand, more PVI-only iPVI patients required PV-reisolation at any time during FU (87% vs 79%, p<0.01). Considering PVI-only (+/- CTI) iPVIs only, dose-area product decreased in RDP compared to first PVI, while procedure duration slightly increased (Figure 1C). Data on periprocedural complications will be reported.

Conclusion: Redo AF ablations procedures are mainly required due to reconnected pulmonary veins or atrial tachycardias. Patients with paroxysmal AF are less likely to require more than 1 RDP, providing indirect support for early rhythm control. Substrate modification during the first AF ablation procedure might be a predictor for occurrence of atrial tachycardias after ablation.