Surgery for permanent atrial fibrillation: impact of patient factors and lesion set.

Background: Whether a complete Cox-maze procedure is needed to ablate permanent atrial fibrillation in patients undergoing concomitant cardiac surgery is unknown. Our objective was to assess the effectiveness of different lesion sets in such patients.

Methods: From November 1991 to January 2004, 575 patients underwent surgical treatment of permanent atrial fibrillation (duration > 6 months); mitral valve disease was the primary indication for surgery in 74%. Procedures included pulmonary vein isolation alone (n = 68, 12%), pulmonary vein isolation with left atrial connecting lesions (n = 265, 46%), and Cox-maze (n = 242, 42%). Rhythm documented on 5,120 postoperative electrocardiograms was used to estimate time-related prevalence of, and risk factors for, atrial fibrillation.

Results: Prevalence of postoperative atrial fibrillation peaked at 46% two weeks after operation, declining to 24% at one year. Patient-related risk factors for increased prevalence included older age (p < 0.0001), larger left atrium (p < 0.0001), and longer duration of preoperative atrial fibrillation (p = 0.0008). The Cox-maze procedure and lesion sets resembling it created with alternative energy sources had a similarly low prevalence of late postoperative atrial fibrillation; in contrast, pulmonary vein isolation and lesion sets that did not include a lesion to the mitral anulus were less effective.

Conclusions: This study suggests that in cardiac surgical patients with permanent atrial fibrillation the left atrial lesion set should include wide pulmonary vein isolation, at least one connection between right and left pulmonary veins, and a connection to the mitral anulus. Availability of alternative energy sources to create lesions sets has virtually eliminated the need for the cut-and-sew Cox-maze procedure.