Atrial fibrillation cycle length as a predictor for the extent of substrate ablation

Aims Atrial fibrillation (AF) cycle length (CL) has been demonstrated to be one of the predictors for termination during ablation for AF. We evaluated the AF CL gradient between right atrium (RA) and left atrium (LA) and their mean AF CL in predicting the extent of substrate ablation. Methods and results One-hundred and thirty-six patients undergoing first ablation for persistent AF were studied. Stepwise ablation, sequentially in the following order: pulmonary veins (PV), LA, and RA, was performed to achieve AF termination. Stepwise ablation terminated AF in 110 patients (81%). In the AF termination group, AF was terminated by PV isolation (PVI) (Group P), PVI plus LA ablation (Group L), and PVI plus LA plus RA ablation (Group R) in 14 patients (13%), 49 patients (44%), and 47 patients (43%), respectively. Group R had much shorter mean AF CL than Group L (156 ± 18 vs. 174 ± 24 ms, P < 0.001) and mean AF CL in Group L was much shorter than Group P (174 ± 24 vs. 209 ± 36 ms, P = 0.004). The RA to LA AF CL gradient was not significantly different between left-side ablation (Group P + Group L) and additional RA ablation (Group R) (P = 0.177). Mean AF CL >180.50 ms predicted AF termination by PVI (Group P) with 79% sensitivity and 84% specificity while mean AF CL >165.25 ms predicted AF termination by left-side ablation (Group P + Group L) with 67% sensitivity and 75% specificity. After a mean follow-up of 15 ± 7 months, freedom from arrhythmia recurrence was significantly higher in left-side ablation (Group P + Group L) than additional RA ablation (Group R) (P = 0.024). Conclusion Baseline mean AF CL may identify the subset of patients in whom persistent AF can be terminated by different extent of substrate ablation, which may in turn predict the chance of recurrence. However, baseline RA to LA AF CL gradient cannot predict the need for additional RA ablation.