Atrial fibrillation (AF) is the most common serious cardiac arrhythmia in the whole world (1), it results in heart failure and left atrial thrombosis, which can lead to stroke and peripheral vascular embolism. AF can be extremely life-threatening and increases healthcare costs (2). Drug therapy and percutaneous ablation were once considered the mainstay of treatment for AF. However, the conversion rate of pharmacologic therapy and ablation in the long term may be as low as 20% (3). Cox-Maze surgery is an effective treatment for persistent AF (4, 5), but because of the invasiveness of sternotomy, this procedure is currently performed in patients who schemed concomitant cardiac surgery.

Surgical heart valve replacement is a common type of heart surgery that requires sternotomy. Surgical aortic valve replacement (SAVR) is currently the most common treatment for CAVD. SAVR requires sternotomy, and previous publications have examined the efficacy and safety of concomitant Cox Maze in patients with mitral valve disease and preoperative AF (6). However, CAVD is poorly understood by today's researchers, with no clear identification of the pathogenesis and short follow-up period analyzed (7). From a pathological perspective, heart valve disease can be classified into three pathogenesis: congenital malformations; rheumatic valve disease; and calcific aortic valve disease (CAVD) (8). At present, CAVD is the third leading cause of adult heart disease and the most common form of acquired valvular disease in western countries (9). Due to the strong association between CAVD and age, combined with the rapid aging of populations worldwide, the number of cases of CAVD will increase (10). The etiology and pathogenesis of CAVD are still unclear, and its relationship with AF remains controversial (11–15). Persistent AF in individuals with CAVD have increased risk for congestive heart failure. As AF can masks these symptoms, delayed reporting of CAVD symptoms were harmful due to later aortic valve interventions (12). However, despite the fact that certain studies have demonstrated that AF is associated with a poor prognosis in patients with severe CAVD, the apparent difference in outcomes in CAVD patients with AF or sinus rhythm (SR) can also be explained by factors other than AF (16).

The effectiveness and safety of Cox-Maze surgery combined with valve replacement have been extensively studied (7, 13, 16–18), however, these studies did not distinguish between rheumatic valve disease and CAVD in studied patients. Because of their closely spaced anatomy, there may be some potential association between the atria and the aortic valves (12). Therefore, we designed a randomized clinical trial to investigate the clinical efficacy and safety of the Cox-Maze IV procedure for AF in patients with CAVD.

This prospective randomized controlled clinical study was registered on the Chinese clinical trial website (ChiCTR1900023–775) and was approved by the ethics committee of Second Xiangya Hospital, Central South University [2019-No.054]. Patient informed consent was obtained by an independent investigator before surgery. This research was conducted in adherence to the principles of the Declaration of Helsinki.

This study was funded by the National Key Research and Development Program (2018YFC1311204).

In this study, we reported outcomes of a randomized controlled clinical trial of Cox-maze surgery in calcific aortic valve disease (CAVD) patients combined with AF. We found that compared with patients in the no Cox-maze group, those in the Cox-Maze group experienced the same hospital stay time and the possibility of perioperative complications. Meanwhile, the patients in the Cox-maze group had better cardiac remodeling changes, and a higher possibility of relief from AF during the follow-up period. Our results agree with previous publications. Churlya et al. (7) conducted a large retrospective clinical study and found that concomitant Cox-maze surgery and AVR is safe but with no advantage in postoperative survival. The authors discuss that inclusion of patients with mitral valve replacement may be a source of confusion and explain controversial results (6). Nevertheless, despite the small sample size of our study, we reported that Cox-maze surgery improves the postoperative survival. Meanwhile, a similar study conducted by Malaisrie et al. in 2012 also suggested that AVR combined with Cox-maze surgery slightly increased operative time but did not significantly increase postoperative complications and mortality (20). Combined with the above findings, we believe that the question of whether Cox-maze surgery is beneficial to patients' postoperative survival remains controversial and deserves continued discussion in future studies.

There is a high incidence of AF in heart valve disease. A total of 80% of patients with mitral stenosis and systemic embolism suffer from AF, and it is estimated that AF is responsible for 25% of deaths from systemic embolism when surgery and anticoagulation are not available (21). However, to date, there are fewer studies on the relationship between CAVD and AF. We can only know from the clinical database (Loire Valley Atrial Fibrillation Project) that aortic valve stenosis is present in 32% of patients who suffer from AF combined with valvular disease (13, 22). The previous study has found that even in asymptomatic or minimally symptomatic patients with severe AS, AF is a strong predictor of mortality. After aortic valve replacement (AVR), AF is still associated with a lower survival rate than sinus rhythm (23). But Hongju Z et al. hold the opposite view. They reveal that other factors than AF explained apparent differences in outcomes when compared to sinus rhythm, including concomitant cardiac abnormalities and postponement of AVR due to AF-related cardiac symptoms (16). In the present study, patients' baseline characteristics were equal among the two groups. In this condition, we treated AF using the Cox-maze IV procedure, which resulted in a significantly better mid-term prognosis for the AVR procedure compared to the no Cox-maze group. Our conclusions support the views of Levy F et al. (23) and Kubala M et al. (13). Different etiologies of valvular disease might account for this disparity. In previous studies, they did not, however, distinguish between rheumatic heart disease and CAVD in included patients. Rheumatic heart disease is the leading cause of valve disease in developing countries, but in western countries, AS has now become the most common valvulopathy. There is a pathological difference between these two diseases: if the rheumatism is repeatedly active, and the exudate is not easily absorbed, it can form a redundancy, making the valve fibrotic and adhesive, which eventually leads to valve stenosis and insufficient. As for CAVD, however, the mechanism involved in its pathogenesis of it remains controversial: lipoprotein deposition, inflammation, and osteogenic transition of cardiac valve interstitial cells may play a role in its pathogenesis (24). Notably, there are differences between males and females in the pathogenesis of CAVD. This leads to the result of women appearing to have lower levels of calcification but still having poorer outcomes after AVR and higher rates of atrial fibrillation and stroke (25, 26). Research has shown that compared to men, women have fewer calcified valves but more fibrosis (27). The mechanisms may include the aldosterone/mineralocorticoid receptor pathway and sex-dependent expression of neutrophil gelatinase-associated lipocalin (28, 29). Previous studies have also shown women with AF may have more side effects on AAD, higher stroke risk, more disabling strokes, and higher cardiovascular mortality. But interestingly, in research about Cox-maze surgery, no significant sex differences could be found in the outcome freedom from AF without AAD treatment at last follow-up (30, 31). This is consistent with the findings of our study (Table 5).

Few studies are reporting a direct relationship between AF and CAVD, however, we can still find clues from independent studies. According to Dai W et al.'s research, HOTAIR, which acts as a ceRNA by sponging miR-613, is a significant factor in the remolding of Cx43 in AF (32). Carrion K et al. demonstrated that HOTAIR is mechanoresponsive and repressed by WNT b-CATENIN signaling. This is the role of HOTAIR in the calcification of human aortic valve interstitial cells (33). The above study found a protective effect of the long non-coding RNA HOTAIR for both AF and CAVD. These results suggest that there may be a molecular mechanism associated with AF and CAVD. This also explains the high conversion rate of Cox-maze surgery in patients with CAVD combined with AF (100% in the present study vs. 60%–70% in the previous study): this may be due to the removal of calcification from the aortic valve, which mitigates the risk factors for the development of AF (34–40).

Previous studies suggested that the major independent predictors of ablation failure are atrial fibrillation/flutter at discharge, preoperative right atrial diameter, hypertension, diabetes, and smoking (41). However, there is limited evidence from patients with CAVD combined with AF. Our study supports part of the conclusions of previous studies: the multivariate Cox regression analysis revealed that SBP (hazard ratio, 1.096; 95% CI, 1.004–1.196; P = 0.04) and pre-operative RADs (hazard ratio, 1.755; 95% CI, 1.182–2.604; P = 0.005) were linked to AF recurrence. Because fewer of our included patients had a history of diabetes and smoking, therefore, the present study failed to conclude the predictive value of these two variables for the recurrence of AF.

Similar to the studies reported in hypertrophic obstructive cardiomyopathy by Meng Y et al. (42), the univariate Cox proportional hazards regression analysis revealed that large post-operative LADs (hazard ratio, 1.156; 95% CI, 1.027–1.303; P = 0.017) can predict the recurrence of atrial filiation after Cox-maze surgery. However, we did not obtain this outcome in the subsequent multi-factor regression analysis ([hazard ratio, 1.144; 95% CI, 0.928–1.409; P = 0.207] in our research vs. [hazard ratio, 1.099; 95% CI, 1.024–1.409; P = 0.009] in Meng's research). The following explanations could explain this difference: there was continuous variable collinearity (LADs, RADs, and left ventricular end-diastolic diameters [LVEDDs]) in the data for which we performed multifactorial regression analysis, which leads to disparity in results. In conclusion, our findings indicate that the AF recurrence rate in patients who underwent the Cox-maze IV procedure was lower than in patients who did not receive the procedure, confirming that the Cox-maze IV procedure should be actively performed in such patients to maintain post-operative SR.

There are some limitations to this study. This was a single-center prospective control study with a relatively small sample size and insufficient follow-up data. In the present study, although G power calculation revealed a sufficient sample size, this only means that the minimum standards are met. As a result, performing a meaningful and extensive statistical analysis is difficult. Since few previous studies have separately examined the surgical treatment of CAVD combined with AF, our total surgical volume was relatively high. Previous studies on the surgical treatment of patients with heart valve disease and AF were mostly retrospective observational studies, so the prospective character of our study provides an opportunity to enhance its importance. Another limitation is that, due to the surgeon's experience and the patient's conditions, we used a variety of ablation equipment and valve prostheses, including Medtronic and AtriCure, while the present study lacked relevant subgroup analysis. Large-scale prospective studies are required in the future to confirm the efficacy and safety of various ablation strategies, equipment, and valve prosthesis. Moreover, these study's findings might be conditioned by the cohort's peculiarities. For example, women represent more than 50% of the two groups and yet the mean age is around 60 years old, normal range of BMI. Therefore, these factors may interfere with the external validity of the results of this study.

In conclusion, our study demonstrated that in patients with CAVD and AF, the Cox-maze IV technique in conjunction with aortic valve replacement increased mid-term survival and decreased mid-term AF recurrence. In patients with CAVD and AF, the concurrent Cox-maze IV procedure is linked to a decreased AF recurrence rate.