Response of atrial fibrillation to pulmonary vein antrum isolation is directly related to resumption and delay of pulmonary vein conduction.

Background: The role of pulmonary vein (PV) isolation in ablative treatment of atrial fibrillation (AF) has been debated in conflicting reports. We sought to compare PV conduction in patients who had no AF recurrence (group I), patients who could maintain sinus rhythm on antiarrhythmic medication (group II), and patients who had recurrent AF despite antiarrhythmic medication (group III) after PV antrum isolation (PVAI).

Methods And Results: PV conduction was examined in consecutive patients undergoing second PVAI for AF recurrence. We also recruited some patients cured of AF to undergo a repeat, limited electrophysiological study at >3 months after PVAI. All patients underwent PVAI with an intracardiac echocardiography (ICE)-guided approach with complete isolation of all 4 PV antra (PVA). The number of PVs with recurrent conduction and the shortest atrial to PV (A-PV) conduction delay was measured with the use of consistent Lasso positions defined by ICE. Late AF recurrence was defined as AF >2 months after PVAI with the patient off medications. Patients in groups I (n=26), II (n=37), and III (n=44) did not differ at baseline (38% permanent AF; ejection fraction 53+/-6%). Recurrence of PV-left atrial (LA) conduction was seen in 1.7+/-0.8 and 2.2+/-0.8 PVAs for groups II and III but only in 0.2+/-0.4 for group I (P=0.02). In patients with recurrent PV-LA conduction, the A-PV delay increased from the first to second procedure by 69+/-47% for group III, 267+/-110% for group II, and 473+/-71% for group I (P<0.001). When pacing was at a faster rate, A-PV block developed in all 5 of the group I patients with recurrent PV-LA conduction.

Conclusions: The majority of patients with drug-free cure show no PV-LA conduction recurrence. Substantial A-PV delay is seen in patients able to maintain sinus rhythm on antiarrhythmic medication or cured of AF compared with patients who fail PVAI.