Prevalence and Discrimination of On-going Fibrillation in Isolated Pulmonary Vein in Non-paroxysmal AF Ablation

Background: It is unclear whether on-going fibrillation in isolated pulmonary veins (OFIP) may coexist with AF in patients undergoing ablation for non-paroxysmal AF (PAF). This study explored the prevalence of OFIP in non-PAF ablation with the aim of developing an algorithm to distinguish between OFIP and isolation failure, thereby minimising excessive ablation. Methods: Non-PAF patients undergoing first-time ablation were divided into a retrospective induction cohort (IC; n=300) and a prospective validation cohort (VC; n=263). In the case of on-going pulmonary vein (PV) fibrillation after ablation, the completeness of PV isolation was assessed based on the change in PV potential and after additional ablation. Results: Among 563 patients, 20 (3.55%) had OFIP (9 in the IC; 11 in the VC). No significant predictors of OFIP were identified in terms of baseline characteristics or procedural factors. Based on three right-sided, five left-sided and one bilateral case of OFIP in the IC, the disappearance of PV potential (either spontaneous or mechanical bumping-induced) or a combination of bipolar voltage reduction >80%, local activation time later than PV recordings and no response to carina ablation were adopted as major criteria in the algorithm to discriminate between OFIP and isolation failure. Among the 20 patients in the VC with on-going PV fibrillation, the algorithm identified true OFIP in 11, false OFIP in three and true incomplete PV isolation in six. The sensitivity, specificity and positive and negative predictive values of the algorithm were 100%, 66.7%, 78.6% and 100%, respectively. Conclusion: The rate of OFIP was not negligible in non-PAF patients undergoing ablation procedures to achieve efficient and safe PV isolation. A novel four-step algorithm was developed to discriminate OFIP from isolation failure, with good predictive value.