BACKGROUND: This study aimed to evaluate the association between left atrial (LA) structural remodeling and very late recurrence [VLR; initial recurrence >12 months after catheter ablation (CA)] after successful CA for non-paroxysmal atrial fibrillation (AF). METHODS: We retrospectively evaluated 63 patients who underwent initial, single ablation for drug-refractory persistent or long-standing persistent AF and those who had no recurrence in the first year after CA. We followed patients for a mean of 3.2±1.5 years and divided them into VLR and no-recurrence (NR) groups. Before and 3 months after ablation, all patients were subjected to 64-slice multidetector computed tomography scanning to estimate LA volume, including maximum and minimum volume during the cardiac cycle (LAMaxV and LAMinV, respectively), and the LA emptying fraction. RESULTS: VLR occurred in 21 patients. The reduction rate of LAMaxV after CA was significantly larger in the NR group than in the VLR group (25±19% vs. 5±18%, p=0.0002). Receiver operating characteristic analysis was performed to determine the best cut-off values in the prediction of VLR. The highest area-under curve was obtained with post-CA LAMinV [0.828 (95% confidence interval, 0.712-0.912), p<0.0001], with a best cut-off value of 44mL (sensitivity 81.0%, specificity 81.0%). CONCLUSIONS: Persistent LA structural remodeling after initially successful CA for non-paroxysmal AF may be an important risk factor for VLR.
BACKGROUND: This study aimed to evaluate the association between left atrial (LA) structural remodeling and very late recurrence [VLR; initial recurrence >12 months after catheter ablation (CA)] after successful CA for non-paroxysmal atrial fibrillation (AF). METHODS: We retrospectively evaluated 63 patients who underwent initial, single ablation for drug-refractory persistent or long-standing persistent AF and those who had no recurrence in the first year after CA. We followed patients for a mean of 3.2±1.5 years and divided them into VLR and no-recurrence (NR) groups. Before and 3 months after ablation, all patients were subjected to 64-slice multidetector computed tomography scanning to estimate LA volume, including maximum and minimum volume during the cardiac cycle (LAMaxV and LAMinV, respectively), and the LA emptying fraction. RESULTS: VLR occurred in 21 patients. The reduction rate of LAMaxV after CA was significantly larger in the NR group than in the VLR group (25±19% vs. 5±18%, p=0.0002). Receiver operating characteristic analysis was performed to determine the best cut-off values in the prediction of VLR. The highest area-under curve was obtained with post-CA LAMinV [0.828 (95% confidence interval, 0.712-0.912), p<0.0001], with a best cut-off value of 44mL (sensitivity 81.0%, specificity 81.0%). CONCLUSIONS: Persistent LA structural remodeling after initially successful CA for non-paroxysmal AF may be an important risk factor for VLR.