Does prior catheter based ablation influence the outcome of subsequent stand-alone atrial fibrillation surgery?

Objectives: Failed catheter ablation of atrial fibrillation (AF) is a common indication for stand-alone AF-surgery. Whether or not these patients are less susceptible for restoration of sinus rhythm (SR) remains unclear. Here we compare the outcomes of patients with and without prior catheter intervention. Methods: Stand-alone surgery for paroxysmal or persistent AF was performed in 85 patients, including 35 (41%) patients with at least one prior catheter ablation. The following approaches were used: video-assisted, bilateral (n = 64) or unilateral (n = 5) mini-thoracotomy; totally endoscopic, bilateral (n = 15) closed chest; and full sternotomy (n = 1). Energy sources were bipolar radiofrequency and cryothermy in 79 and 6 patients, respectively. Lesions sets included bilateral pulmonary vein isolation and various other ablation lines of the left atrium, as well as left atrial appendage removal. Heart rhythm was monitored by serial 24-h Holter EKGs up to 3 years of follow-up. Results: All patients survived AF surgery and were discharged from the hospital in good condition. Major adverse events included intraoperative conversion to sternotomy due to bleeding (n = 1) and reexploration for bleeding (n = 2). Overall success rate was 85% of patients in sinus rhythm at last follow up visit. There was no difference between patients that did or did not have prior catheter ablation as to restoration of sinus rhythm (87% vs. 84%, p NS) or adverse events. One patient required re-do catheter ablation due to atypical left atrial flutter after AF surgery. This patient did have prior catheter ablation. Conclusions: Patients with prior catheter ablation are just as good of a candidate for stand-alone AF surgery as those without prior catheter interventions. Overall success rates are promising and justify the procedure to treat recurrent AF. Whether or not AF surgery should be preferred over repeat catheter intervention will require a randomized trial.