Surgical Ablation of Atrial Fibrillation during Mitral-Valve Surgery

Atrial fibrillation, which is associated with reduced survival and increased risk of stroke, is present in 30 to 50% of patients presenting for mitral-valve surgery.1,2 The development of open surgical procedures for the ablation of atrial fibrillation has led to their widespread application during cardiac operations, but their effectiveness and safety have not been rigorously established. It is hypothesized that long-term outcomes can be improved by successful ablation in patients with preexisting persistent or long-standing persistent atrial fibrillation who are undergoing mitral-valve surgery.

The Cox maze III operation (sometimes called the “cut-and-sew” maze operation) is a complex surgical procedure for the control of atrial fibrillation. Developed in 1992, the procedure involves the creation of a “maze” of surgical incisions in both the right and left atria to interrupt macro-reentrant circuits that are thought to be responsible for the propagation of atrial fibrillation. Newer insights into the pathophysiological mechanisms of atrial fibrillation and the development of other tissue-ablation technologies (e.g., radiofrequency ablation and cryoablation) have encouraged frequent attempts at atrial fibrillation ablation during heart surgery. Almost all current approaches to ablation during surgery include pulmonary-vein isolation, which is the simplest, most rapidly completed set of ablation lesions; it involves the creation of circumferential ablation lesions around the pulmonary veins. In contrast, the biatrial maze lesion set, which is performed with contemporary ablation devices, requires right and left atriotomies, a longer duration of cardiopulmonary bypass, and the creation of endocardial ablation lesions extending to the mitral and tricuspid annuli.

Current American Heart Association (AHA)–American College of Cardiology (ACC)–Heart Rhythm Society (HRS) guidelines state that it is reasonable to perform atrial fibrillation ablation in selected patients undergoing other types of cardiac surgery but acknowledge that there are limited data on which to base this recommendation (level C evidence).3 Moreover, although pulmonary-vein isolation is used more frequently than the biatrial maze procedure, data on the comparative effectiveness of the two procedures are also limited. Uncertainty regarding both the benefits of surgical ablation and the choice of lesion sets has led to wide variation in practice among surgeons.4 In this randomized trial involving patients with persistent or long-standing persistent atrial fibrillation who were undergoing mitral-valve surgery, we sought to determine the effect of surgical ablation on the recurrence of atrial fibrillation in the first year after surgery and to explore the effects of two different ablation procedures (pulmonary-vein isolation or biatrial maze procedure) on freedom from atrial fibrillation during the same period.

METHODS

Trial Design and Oversight

This trial was performed at 20 centers in the Cardiothoracic Surgical Trials Network (CTSN); the design of the trial has been published previously.5 The trial was conducted under an investigational-device exemption, because only devices using cryoablation were approved for the treatment of atrial fibrillation when the study began; other energy sources (e.g., radiofrequency) were not approved at that time. The trial was designed to evaluate ablation as a therapeutic approach, not to support Food and Drug Administration approval for any individual device.

The CTSN investigators designed the trial protocol, which was approved by the institutional review board at each participating center. The investigators also collected and analyzed the data and wrote the manuscript. A coordinating center, an independent adjudication committee, and a data and safety monitoring board appointed by the National Institutes of Health oversaw trial progress. There was no support from a commercial entity in this trial. The investigators vouch for the accuracy and completeness of the data and for the fidelity of this report to the trial protocol, which is available with the full text of this article at NEJM.org.

Patients and Interventions

This trial enrolled adult patients with persistent or long-standing persistent atrial fibrillation who also had mitral-valve disease requiring surgical intervention. According to the 2012 HRS guidelines in place during the trial,6 persistent atrial fibrillation was defined as non–self-terminating atrial fibrillation lasting more than 7 days, or less than 7 days if cardioversion was required. The definition was revised per AHA-ACC-HRS guidelines in 2014 and limited simply to continuous atrial fibrillation for more than 7 days.3 Long-standing persistent atrial fibrillation was defined as continuous atrial fibrillation for more than 12 months.6 Detailed eligibility criteria have been reported5 and are described in the Supplementary Appendix, available at NEJM.org. All participating patients provided written informed consent.

Eligible patients were randomly assigned, in a 1:1 ratio, to undergo either surgical ablation or no ablation (control group) during the mitral-valve operation after the induction of anesthesia. Patients in the ablation group underwent further randomization to one of two lesion sets: pulmonary-vein isolation or biatrial maze. Randomization was performed after intraoperative transesophageal echocardiography confirmed the absence of a left atrial thrombus. Randomization was stratified according to center. All patients also underwent closure of the left atrial appendage to reduce the risk of formation of a left atrial thrombus. The surgical-ablation procedures and postoperative management are described in the Supplementary Appendix, including Fig. S1.

End Points

The primary end point was freedom from atrial fibrillation at both 6 months and 12 months after surgery, as assessed by means of 3-day continuous Holter monitoring. Patients who died before the 12-month assessment or who were too ill to undergo assessment of atrial fibrillation were considered not to have had a response to treatment, as were patients who underwent any ablation therapy for atrial fibrillation after the index procedure.

Secondary end points included a composite of major cardiac or cerebrovascular adverse events (death, stroke, hospitalization for heart failure, worsening heart failure [as defined by an increase of one or more classes in the New York Heart Association classification], or mitral-valve reintervention), mortality, the need for rhythm-related interventions, quality of life (as assessed by means of the Atrial Fibrillation Severity Scale [AFSS] and the physical and mental subscales of the Medical Outcomes Study 12-Item Short Form Health Survey [SF-12]), and rehospitalization.

The primary safety end point was a composite of death, stroke, heart failure, myocardial infarction, rehospitalization for cardiac causes, transient ischemic attack, pulmonary embolism, peripheral embolism, excessive bleeding, deep sternal-wound infection or mediastinitis, damage to the specialized conduction system necessitating implantation of a permanent pacemaker, or damage to peripheral structures such as the esophagus, within 30 days after the procedure or hospital discharge (whichever was later).

Follow-up assessments were conducted by telephone interview at 3, 6, and 9 months and in person at 12 months. Investigators were unaware of the trial outcomes, and the personnel at all core laboratories were unaware of the treatment-group assignments. Definitions of end points are provided in the Supplementary Appendix.

Statistical Analysis

The primary null hypothesis was that there would be no difference between randomization groups in the proportion of patients free from atrial fibrillation. We tested this hypothesis in an intention-to-treat analysis using a chi-square test, at a 0.05 alpha level. Patients with missing data (not owing to death) had their primary end point imputed (with the use of a multiple-imputation model with five iterations), on the assumption that data were missing at random. We calculated that enrollment of 260 patients would give the study 90% power to detect an absolute increase of 20 percentage points (from 25% to 45%) in the proportion of patients free from atrial fibrillation with ablation therapy.5,7 We conducted a planned interim analysis using the Lan–DeMets approach with an O'Brien–Fleming–type spending function when approximately 50% of the patients had reached their 1-year follow-up; therefore, a P value of less than 0.049 was considered to indicate statistical significance in the final analysis of the primary end point. No inferential hypotheses were specified to compare the ablation subgroups (pulmonary-vein isolation vs. biatrial maze procedure).

The hazards of major cardiac or cerebrovascular adverse events and death from any cause were compared between groups with the use of Cox proportional-hazards models. Between-group differences in adverse-event and hospitalization rates were tested with the use of Poisson regression, differences in quality-of-life scores with the use of Student's t-tests, and differences in categorical outcomes with the use of chi-square tests.