The Cox maze procedure in mitral valve disease: Predictors of recurrent atrial fibrillation

Objectives: The Cox maze procedure is the gold standard for ablation of atrial fibrillation in patients undergoing mitral valve surgery, and new approaches to atrial fibrillation ablation must be compared with it. Therefore, we sought to determine the time-related prevalence of atrial fibrillation and its risk factors after combined Cox maze and mitral valve surgery. Methods: From November 1991 through January 2004, 263 patients (mean left atrial diameter, 5.8 +/- 1.2 cm) underwent combined mitral valve surgery (repair in 71%) and a cut-and-sew Cox maze procedure for atrial fibrillation (permanent, 74%; persistent, 7%; paroxysmal, 16%). Rhythm documented on 2367 postoperative electrocardiograms was used to estimate the prevalence of atrial fibrillation across time. Results: Hospital mortality was 1.9%. Postoperative atrial fibrillation prevalence peaked at 36% at 2 weeks, decreasing to 21% at 5 years. Risk factors for higher postoperative atrial fibrillation prevalence varied with time and included longer duration of preoperative atrial fibrillation (P = .003), larger left atrial diameter (P = .01), older age (P = .0002), and higher left ventricular mass index (P = .02). Conclusions: in some patients undergoing mitral valve surgery and a Cox maze procedure, atrial fibrillation recurs over time, mandating close, long-term follow-up of heart rhythm. Earlier operation and left atria] size reduction should be considered to improve results in selected patients.