2.1. Study Populations
Between January 2006 and March 2015, a total of 362 consecutive patients (aged ≥ 70 years) with valve disease and AF underwent cardiac surgery at our institution. Among them, 48 patients underwent VR with concomitant Cox-maze procedure (42 underwent BRFA procedure and 6 underwent Cox-maze III procedure). The inclusion criteria were as follows: 1) Patients who underwent VR and/or BRFA for AF; and 2) patients with biological valves. The exclusion criteria were as follows: 1) Patients with mechanical valves; 2) patients who had received repeated cardiac surgeries; 3) patients who had received other forms of surgery for AF; and 4) patients who had not completed a follow-up of at least 5 years, unless death occurred. Finally, 276 patients were included in the final analyses. Preoperative demographic and clinical data, perioperative outcomes, and follow-up results were collected for all patients and retrospectively analysed. Of the 276 patients, 42 patients underwent VR with concomitant BRFA, and the remaining 234 patients underwent VR alone. All patients had persistent or long-lasting persistent AF, the diagnosis of which was confirmed by 24-hour Holter monitoring before surgery. PSM was performed to achieve balance in covariates between groups. The process of patient selection is illustrated in Figure 1.
The study was approved by the Ethics Committee of Union Hospital, Fujian Medical University, and the study adhered to the Declaration of Helsinki. The requirement for written informed consent from the patients or their guardians was waived due to the retrospective nature of the study.
2.2. Surgical Procedures
Decisions regarding the indication for AF ablation were made according to the preference of the heart surgeon at our institution (according to the patients’ EuroSCORE and other characteristics). In general, patients with a EuroSCORE ≥ 10, severe valvular heart disease requiring emergency surgery, LA diameter > 55 mm, or long-lasting (> 15 years) persistent AF were not considered suitable candidates for concomitant BRFA.
All surgical procedures were performed through a median sternotomy under general anaesthesia with hypothermic cardiopulmonary bypass (CPB). We used the Medtronic Cardioblate G2 Surgical Ablation System (Medtronic Inc., Minneapolis, MN) with a bipolar radiofrequency clamp for surgical ablation in accordance with the method introduced by Sie et al. (Figure 2). [18] In cases where there was a thrombus in the left atrial, atrial thrombectomy was performed first after aortic cross-clamping. If there was no thrombus, right atrial ablation was performed with the heart beating, and left atrial ablation was performed during cardioplegic arrest. The right pulmonary was isolated, and the circumferential line around the right pulmonary veins was ablated. Two ablation lines were made from the right atrial appendage to the right atrial free wall and the tricuspid annulus (the junction between the anterior and septal leaflets of the tricuspid valve). The incision on the free wall of the right atrium was extended from the atrioventricular groove to the interatrial groove. Two ablation lines were created from the right atriotomy to the inferior vena cava and the superior vena cava. Another ablation line was created around the orifice of the coronary sinus. After aortic cross-clamping, the left pulmonary vein was isolated, and the ligament of Marshall was interrupted. The left atrium was entered via the incision parallel to the interatrial groove, and the circumferential line around the left pulmonary veins was ablated. Other ablation lines were created between the left atrial appendage and the left superior pulmonary vein, between the bilateral superior and inferior pulmonary veins, between the right inferior pulmonary vein and the posterior mitral annulus. The left auricle was finally ligated. Each radiofrequency ablation was performed at least three times. The valve prosthesis was then implanted. The temporary epicardial pacing wire was placed routinely after cardiac surgery.
2.3. Postsurgical Treatment
The postoperative care was similar for both groups. Unless contraindicated, amiodarone infusion was given 5 mg/kg in the first hour and 0.6 mg/kg/h over the following 72 hours, and 200 mg orally twice a day for the next 3 weeks. For patients undergoing BRFA for AF, amiodarone infusion was given 200 mg orally daily until the end of the first 6 months to reduce the risk of early atrial arrhythmia recurrence. Sotalol or metoprolol was given as appropriate to patients with a history of side effects of amiodarone, or other clinical contraindications, in order to control the ventricular rate. Patients were advised to continue use of warfarin for 3 months after surgery, unless there was a specific contraindication. The duration of prothrombin was regularly monitored; the target international normalized ratio was 1.5–2.0 for aortic valve replacement and 2.0-2.5 for mitral or tricuspid valve replacement. In cases where the patient underwent valve replacement alone, and if no attempts were made to restore and maintain sinus rhythm by means of ablation, β-blocker drugs were administered to control the ventricular rate, and warfarin was used for anticoagulation.
2.4. Outcome Measures
The primary outcomes included all-cause death, cardiac death, sinus rhythm restoration, and stroke events. The secondary outcomes of interest were surgical mortality and major postoperative morbidity. Surgical mortality was defined as death during the same hospitalization, and cardiac death was defined as death due to malignant arrhythmia, stroke, congestive heart failure, or sudden unexplained death. Neurological complications are defined here as embolic or hemorrhagic strokes. Stroke was defined as an ischemic event caused by cerebral embolism. After 6 months of surgical ablation, any episode of AF, atrial flutter, or atrial tachycardia lasting more than 30 seconds recorded by electrocardiogram or 24-hour Holter monitoring, was considered as recurrence of AF. Long-term survival was measured as time (months) to death from the date of surgery.
2.5. Follow-up
Follow-up visits at discharge were made 3, 6, and 12 months postoperatively and annually thereafter. The patients’ health history, a physical examination, a 12-lead electrocardiogram, and an echocardiography were obtained at each visit. Patients with palpitations or documented atrial arrhythmias by 12-lead electrocardiography underwent 24-hour Holter monitoring for further confirmation. The period between the day after surgery and death, loss to follow-up, or the predefined date of March 2020 was defined as the follow-up duration.
2.5. Statistical Analysis
Continuous variables are expressed as mean ± standard deviations and categorical variables are presented as count, unless otherwise noted. Differences were compared using the Student’s t-test or Mann-Whitney U test for continuous variables, and the chi-square test for categorical variables. One-to-one pair matching was performed using the caliper-matching method with the variables shown in Table 1, and a 0.02 propensity score tolerance was imposed on the maximum propensity score distance. The overall survival and AF-free survival were analysed using Kaplan–Meier analysis and log-rank methodology. Cox regression was used to test for differences in survival between groups, and hazard ratios (HR) and 95% confidence intervals (CI) were computed. The incidence of postoperative stroke in the matched population was estimated using the cumulative incidence function with death as a competing risk. The hazard of cardiac death was also analysed using the competing risk method, with non-cardiac death as the competing event. The differences between cumulative incidence curves in the competing risk analysis were compared using Gray’s tests. Fine and Gray competing risk regression was used to analyse the determinants of postoperative stroke and cardiac death. The sub distribution hazard ratio (SHR) and 95% CI were computed. A two-sided P-value < 0.05 was considered statistically significant. All statistical analyses were performed using SPSS 26.0 and R 3.6.2.