Comparison of pulsed field and radiofrequency ablation for non-paroxysmal atrial fibrillation: a propensity score-matched analysis

Abstract

Abstract Background Ablation of non-paroxysmal atrial fibrillation (AF) is challenging and often requires left atrial substrate ablation. While the pentaspline, multi-electrode pulsed-field ablation (PFA) catheter is approved for pulmonary vein isolation (PVI), its role in procedures involving substrate ablation for nonparoxysmal forms of AF remains unknown. Purpose We conducted a retrospective study to assess the safety and long-term outcomes of PFA for non-paroxysmal AF. Methods Procedural and follow-up data were retrospectively collected for all patients undergoing a first-time ablation for non-paroxysmal AF across two centers, between November 2021 and December 2022. Patients undergoing PFA consisting of PVI, isolation of the posterior wall (PWI) and ablation of the mitral isthmus (MI) with the pentaspline multi-electrode PFA catheter, were compared with patients undergoing radiofrequency ablation (RFA) involving PVI, a roof line or PWI, alcohol ablation of the vein of Marshal (VOM) and MI ablation. Patients were matched using propensity scores to minimize potential confounders. The primary endpoint was arrhythmia (AF/atrial flutter/atrial tachycardia) recurrence after a 90-day blanking period. Results The baseline characteristics of the 116 matched patients are detailed in Table 1. PVI was successful in all patients. In the PFA group, PWI was additionally performed in all patients and MI ablation in 83% of cases, resulting in acute success rates of 97% and 96%, respectively. In the RFA group lesions beyond PVI included: a roof line (90%), PWI (10%), VOM alcohol ablation (93%) and MI ablation with ablation in the coronary sinus (CS) if needed (97%). Procedural block across the roof and the MI was achieved in 96% and 95%, respectively and complete PWI in 83% of cases. The PFA group had a shorter total procedure (84.2±22.3 vs 117.3±24.8 min, p<0.001) but a longer fluoroscopy time (23.1±7.2 vs 16.8±6.3 min, p<0.001). The difference in radiation dose, however, was not statistically significant (57.7±74.1 Gy.cm2 in the PFA vs 49.8±53.6 Gy.cm2 in the RFA group, p=0.24). The overall periprocedural complication rates were similar in both groups (8.6% PFA vs 12.1% RFA, p=0.76), with no cardiac tamponades or deaths. Group-specific complications included 1 coronary spasm in the PFA group and 1 CS dissection in the RFA group. During a median follow-up of 447 days (IQR: 364-552), 1 stroke occurred in each group and 1 sudden cardiac death in the RFA group, 13 months post-procedure. Both groups had similar rates of recurrences (32.8% PFA vs 25.9% RFA, p=0.54). Arrhythmia-free survival probability at 16 months was estimated at 67.2% for PFA and 77.6% for RFA by Kaplan-Meier analysis (p=0.16, Figure 1). Conclusions In this propensity-matched cohort of patients with non-paroxysmal AF, PFA and RFA demonstrated similar complication rates and arrhythmia-free survival at follow-up. The procedure duration, however, was significantly shorter with PFA.Figure 1