Outcome of Cox Maze procedure concomitant with mitral valve operation in treatment of atrial fibrillation

Atrial fibrillation (AF) is fairly common in patients with chronic mitral valve disease and a markedly dilated left atrium. These patients rarely return to sinus rhythm spontaneously, even after a successful mitral valve operation. The Maze procedure, developed by Cox and associates,1-5 has been used in patients undergoing a mitral valve operation. Restoration of normal sinus rhythm and atrial contraction in such patients have been reported by some groups. This study will summarize retrospectively a group of patients who underwent the Cox Maze procedure for the treatment of AF associated with rheumatic mitral valve disease in China. METHODS From 1994 to 2003, 63 patients with AF associated with rheumatic valve disease underwent the Cox Maze procedure with concomitant valve replacement. Of these, 47 patients underwent a Maze-III procedure, and 16 patients underwent a cryosurgical Maze-I procedure (Table 1). For each case, the protocol of anesthesia, cardiopulmonary bypass (CPB) and myocardial protection was determined by the major surgeon, Professor CHEN Ru-kun. A longitudinal median sternotomy with hypothermic (25°C) extracorporeal circulation and distal cannulation of the venae cavae was performed on all the patients. After clamping of the aorta, crystalloid cardioplegic solution at 4°C (St. Thomas') was infused. The infusion was repeated at 20-minute intervals. Encouraged by Cox and associates1-5 we applied the complete Maze procedure (the same circuit), using cryosurgical technique rather than the old "cut and sew" technique to avoid leaving suture lines in the posterior left atrium, as we previously reported.6-8 Postoperative follow-up ranged from 1 to 70 months (average, 16 months). Patients had electrocardiograms during this time. Postoperative left atrial transport function was documented by multiple tests including transthoracic echocardiography and dynamic magnetic resonance imaging. Complications such as bleeding, anticoagulantrelated hemorrhage, shock and other thromboembolism were recorded. Data were expressed as mean ± standard deviation (SD) or percentage. Independent t test was used to determine statistical significance between these two Maze groups since our sample count was not large enough, and the discontinuous data were not distributed randomly. The SigmaStat software package (SPSS 10.0) was used for statistical analyses.Table 1: Comparison of patients demographics for the two types of maze proceduresRESULTS AF disappeared in all patients postoperatively except one early death (1/63). For patients undergoing a Cox Maze-III procedure, AF reappeared in 2 cases five years later, but could be converted to sinus rhythm with a single anti-arrhythmic drug. Normal sinus rhythm was restored in most of the cases (46/47), and a nodal rhythm appeared in only one patient. No blunted sinus node response to maximal exercise occurred, and no pacemaker implantation was needed. Atrial contractility was restored in most patients (46/47). For patients undergoing a Cox Maze-I procedure, an appropriate sinus tachycardia could not be generated in response to maximal exercise in 3 cases, however no pacemaker implantation was required. In 4 cases, decreased LA contractility and transport function were observed by echocardiography or MRI three months after operation. No thromboembolism or serious hemorrhage related to anticoagulant therapy occurred (Tables 2 and 3 ).Table 2: Perioperative morbidity and early results (less than 1 month after the two types of Maze procedures)Table 3: Comparison of late results (1 to 70 months after the operation) for the two types of Maze procedures (n)DISCUSSION Over the past 50 years, multiple-circuit re-entry has been the dominant conceptual model of AF. However, the pathogenesis of AF in mitral valve patients is incompletely understood, and the extent to which mechanisms of focal activity and re-entry contribute to the initiation and maintenance of AF is disputed. Nevertheless, available mapping studies support the importance of the left atrium as the driving chamber in mitral valve patients.9,10 For many patients with permanent AF, regular and repetitive activation can be identified in the posterior left atrium in the regions of the pulmonary vein orifices and the left atrial appendage.10 However, the spectrum of AF is complex; such foci are not identified in all mapped patients, and some patients also manifest right atrial focal or re-entry activation.9 In patients undergoing mitral valve surgery, persistence of postoperative AF is linked to increased mortality and morbidity that includes stroke, other thromboembolism, anticoagulantrelated hemorrhage,11 and reduced cardiac output. For these reasons, the presence of AF should be factored into the operative strategy for these patients, but there is no consensus concerning ablation strategy in these patients. In a recently published review, four ablation procedures were prominently mentioned.12 (1) The Maze procedure, developed by Dr. Cox and his research laboratory in Saint Louis, Missouri, cures AF by isolation of the posterior left atrium, including all four pulmonary veins and multiple incisions on the right and left atria, and results in a high success rate of ablation with a significantly low rate of thromboembolism. (2) The radial procedure, in which the atrial incisions radiate from the sinus node toward the atrio-ventricular annular margins and parallel the coronary arteries, was developed as an outgrowth and alternative to the Maze procedure in order to preserve a more physiological activation sequence and atrial transport function. (3) Simplified procedures such as left atrial isolation and even pulmonary vein isolation are advantageous in reducing blood loss and operation time, although the ablation rate is lower than the Maze or radial procedures. (4) Map- guided surgery for AF, in which the intraoperative mapping assists in determining the optimal procedure for each patient, were reported in groups, but are time consuming, technically demanding, and are not feasible for most mitral valve cases. Cox Maze procedure has been considered a milestone in the surgical treatment of AF. Initial efforts of cardiac electrophysiological research allowed complicated cardiac electrophysiological analysis during open heart surgery, and the AF ablation success rate is much higher than with former procedures. But the original Maze procedure that was described for the treatment of atrial fibrillation was later found to be associated with 2 unacceptable problems: (1) the frequent inability to generate an appropriate sinus tachycardia in response to maximal exercise; and (2) occasional left atrial dysfunction.1-3,6-8 In an effort to overcome these problems, the Maze-I procedure was modified twice, and the result was the Maze-III procedure, which is associated with a higher incidence of postoperative sinus rhythm, improved long-term sinus node function, fewer pacemaker requirements, less arrhythmia recurrence, and improved long-term atrial transport function.2,3 In our cases, we also found that compared to the Maze I procedure, the Maze-III procedure resulted in better postoperative sinus node function and left atrial contractility, which indicates a better long-term outcome. In addition, the Maze-III procedure is technically less demanding than either the Maze-I or Maze-II procedures. For experienced surgeons, the duration of aortic clamp time and CPB time was acceptable.6-8 Despite the obvious efficacy of the Maze procedure, it has never been widely adopted by either surgeons or cardiologists because of its complexity and associated complications.5 Several "modifications" of the procedure were attempted. It was once considered that all that was needed to cure AF was to isolate the orifices of the pulmonary veins, an approach that eventually resulted in a predictable 30% failure rate.2 However, the overall AF ablation rate of 70% to 80% that now is being accomplished in some centers certainly is an improvement over simply ignoring AF in patients undergoing mitral valve surgery.5 We continued to apply the complete Maze procedure,6-8 using cryosurgical technique rather than the old "cut and sew" technique to avoid leaving suture lines in the posterior left atrium. We emphasized that the design of incision in the Maze-III procedure was necessary for its high efficacy, but the "cut and sew" procedure for certain locations could be modified to simplify the operation. Groups of successful cases of modifying Cox Maze procedure have been reported in recent years, including alcohol injection, cryosurgery, radiofrequency, electro-coagulation, and highintensity focused ultrasound.13-15 We also realized that complete transmurality for every modified "cut and sew" procedure must be ensured, and also that use of a proper linear cryoprobe was strongly preferred to accomplish the continuous cryoablation, and that these were the key points to retain the same efficacy as the old "cut and sew" technique.8 It was unfortunate that our 63 case experience covered a period of 10 years, and as a retrospective report, these cases were not precisely divided randomly. Nevertheless, for each patient, Professor CHEN Ru-kun, the major surgeon, made up the same protocol for both operations (including surgical technique, anesthesia, cardiopulmonary bypass and myocardial protection) and follow up, and this likely added some validity to the present report.