Where does atrial fibrillation surgery fail? Implications for increasing effectiveness of ablation

Objective: Surgical ablation of atrial fibrillation is generally safe and effective, but atrial fibrillation redevelops in approximately 20% of patients. We sought to determine anatomic factors, technology factors, or both that contribute to these failures. Methods: Four hundred eight patients underwent 5 types of atrial fibrillation ablation depending on their atrial fibrillation history and need for concomitant surgical intervention: the classic maze procedure, high-intensity focused ultrasound, the left atrial maze procedure, the biatrial maze procedure, and pulmonary vein isolation. Ninety-five percent of patients with preoperative atrial fibrillation underwent surgical ablation. Results: Patients undergoing high-intensity focused ultrasound had a high rate of late postoperative percutaneous ablation (37.5%) after surgical intervention (P < .001 vs the other groups). At last follow-up, freedom from atrial fibrillation and need for ablation was as follows: classic maze procedure, 90%;high-intensity focused ultrasound, 43%; left atrial maze procedure, 79%; biatrial maze procedure, 79%; and pulmonary vein isolation, 69% (P < .001 between groups). For those with atrial fibrillation, mapping and ablation were performed in 23.6% (n = 27), and all patients with high-intensity focused ultrasound had failure of the box lesion around the pulmonary veins. Of those with just the left atrial maze procedure or pulmonary vein isolation, the right atrium was the source for failure in 75% (6/8). Conclusions: Patients undergoing high-intensity focused ultrasound had a high need for postoperative ablation and low freedom from atrial fibrillation. The classic maze procedure had the best results. Left atrial ablation might allow failure from right atrial foci. Matching the technology and lesion set to the patient yields good results and can be applied in 95% of patients. We suggest others obtain late catheter ablation to correct remaining atrial fibrillation, and add to the paucity of late data regarding failure mode. (J Thorac Cardiovasc Surg 2010; 139: 860-7)