Acute procedural outcomes
As shown Table 2 , 12 (23%) patients in LINE-group and 5 (10%) patients in EGM-group were in SR at the beginning of the procedure. In the EGM-group, AF was induced by burst pacing in all patients with SR at the beginning of the procedure. PVI was performed as a first step in all patients, with a 99% of successful isolation rate. In the LINE-group electrical cardioversion was performed after PVI in all patients. In the EGM-group AF persisted post-PVI in all but 1 patient.
In LINE-group, roof and MI lines were successfully blocked in 49 (94%) and 42 (81%) patients, respectively. Both lines were completed in 39 (75%) patients. Epicardial ablation within the CS for the MI line was added in 28 (54%) patients. Mean RF time for roof and MI line was 6.6±3.5 and 14.0±7.3 minutes, respectively. In EGM-group, AF was terminated by RF in 21 (40%) patients (conversion to AT in 11 patients and termination to SR in 10 patients). Mean RF time for EGM-guide ablation was 47±30 minutes. The right atrium was targeted in 36 (69%) patients. Roof and/or MI lines were created due to induced roof and/or MI line-dependent ATs in 5 patients.
Non-PV foci ablation, including SVC isolation, was performed in 7 (13%) in the LINE-group and 5 (10%) patients in the EGM-group (p=0.76). Procedure time and total RF time was significantly longer in EGM-group (LINE-group vs. ECG-group; procedure time 252±65 vs. 291±56 minutes, p=0.002; RF time 65±27 vs. 91±34 minutes, p<0.001). There were no differences in the incidence of procedure-related major complications between the two groups. Very early recurrence rates (within 3 days) were also comparable between two groups (Table 2 ).