Comparison of prognosis and outcomes of catheter ablation versus drug therapy in patients with atrial fibrillation and stable coronary artery disease: A prospective propensity‐score matched cohort study

Abstract Background Atrial fibrillation (AF) and stable coronary artery disease (SCAD) frequently coexist. Hypothesis To investigate the prognosis of catheter ablation versus drug therapy in patients with AF and SCAD. Methods In total, 25 512 patients with AF in the Chinese AF Registry between 2011 and 2019 were screened for SCAD. 815 patients with AF and SCAD underwent catheter ablation therapy were matched with patients by drug therapy in a 1:1 ratio. Primary end point was composite of thromboembolism, coronary events, major bleeding, and all‐cause death. The secondary endpoints were each component of the primary endpoint and AF recurrence. Results Over a median follow‐up of 45 ± 23 months, the patients in the catheter ablation group had a higher AF recurrence‐free rate (53.50% vs. 18.41%, p < .01). In multivariate analysis, there was no significant difference between the strategy of catheter ablation and drug therapy in primary composite end point (adjusted HR 074, 95%CI 0.54–1.002, p = .0519). However, catheter ablation was associated with fewer all‐cause death independently (adjusted HR 0.36, 95%CI 0.22–0.59, p < .01). In subgroup analysis, catheter ablation was an independent risk factor for all‐cause death in the high‐stroke risk group (adjusted HR 0.39, 95%CI 0.23–0.64, p < .01), not in the low‐medium risk group (adjusted HR 0.17, 95%CI 0.01–2.04, p = .17). Conclusions In the patients with AF and SCAD, catheter ablation was not independently associated with the primary composite endpoint compared with drug therapy. However, catheter ablation was an independent protective factor of all‐cause death

Conclusions: In the patients with AF and SCAD, catheter ablation was not independently associated with the primary composite endpoint compared with drug therapy. However, catheter ablation was an independent protective factor of allcause death K E Y W O R D S atrial fibrillation, catheter ablation, coronary artery disease

| INTRODUCTION
As the most common sustained arrhythmia, atrial fibrillation (AF) has a high possibility to occur together with coronary artery disease (CAD), which is another common cardiovascular disease. Over 20 percent of patients with AF suffered from CAD, 1 and about 19 percent of patients with CAD had AF. 2 However, the optimal treatment of AF with CAD remains unclear. Most of previous studies usually focused on the antithrombotic protocol for patients with AF and CAD.
In recent years, catheter ablation has become an important choice for AF treatment. EAST-AFNET 4 trial reported that early rhythmcontrol therapy was associated with a lower risk of a composite of death from cardiovascular causes, stroke, or hospitalization than usual care among patients with early AF and cardiovascular conditions. 3 Which emphasized the advantages of early rhythm-control therapy in the management of AF. Despite the fact that catheter ablation was effective for treating AF, whether CAD could affect the outcome of AF catheter ablation remained unclear. In a study from the Leipzig Heart Center, it was found that neither the presence nor severity of CAD could affect the recurrence within 12 months after AF ablation. 4 Dennis 5 et al. also found that CAD did not increase the recurrence within 12 months after AF ablation. However, some studies showed the opposite conclusion. Hiraya 6 et al. found that CAD was an independent risk factor for recurrence of catheter ablation of AF after an average of 44-months of follow-up. A retrospective study found that the presence of CAD has no impact on AF recurrence after cryoablation. 7 The aforementioned studies explored the impact of CAD on the success rate of catheter ablation of AF. It was more important to address the issue of the impact of catheter ablation on the prognosis in the patients with AF and CAD. However, there were only a few studies to address this issue. In EAST-AFNET4 study, early rhythm control did not reduce the hospitalization with acute coronary syndrome. 3 A small retrospective cohort study showed that catheter ablation could improve the long-term prognosis of AF patients who underwent percutaneous coronary intervention (PCI). 8 The small sample size and the exclusion of patients without PCI lead to low reliability and poor generalization. Here, we aimed to examine the effect of catheter ablation versus drug therapy on long-term prognosis in the patients with AF and stable coronary artery disease (SCAD) in a prospective cohort study.

| Study population
All the patients were screened from the Chinese Atrial Fibrillation Registry (CAFR) study between August 2011 and December 2019. CAFR has been described in details previously. 9 Briefly, CAFR is a prospective registry study with ongoing enrollments and follow-up involving 19 tertiary and 12 non-tertiary hospitals in Beijing, China. Eighteen of the 31 centers had the ability to perform AF ablation. Written consents were obtained from all patients when they enrolled in the CAFR, and the ethics committee approved this study. All the data were collected from the medical record system or through telephone interviews.
Patients would be enrolled in this study if meeting all the following inclusion criteria: (1) age ≥ 18 years; (2) diagnosis of AF; (3) suffered from SCAD. AF was diagnosed by 12-lead electrocardiogram or 24 hours-Holter with a record lasting ≥30 seconds. SCAD was defined as a clinical condition with at least one of the following inclusion criteria: myocardial infarction (MI) ≥3 months ago; coronary artery bypass grafting (CABG) or PCI ≥3 months; stable chest pain with proven myocardial ischemia; or previous coronary angiography showing ≥1 coronary stenosis >50% and not require revascularization. 10 Patients should be excluded if meeting any of the exclusion criteria as the followings: (1) valvular AF; (2) a history of catheter ablation or surgery for AF. In the CAFR, there were 12 104 patients underwent catheter ablation and 13 408 patients underwent drug therapy. Totally, 2665 patients were selected according to the inclusion criteria and exclusion criteria, including 1921 patients with catheter ablation and 844 patients with drug therapy. After propensity-score matching, 815 pairs of patients in each group were enrolled in the study. The patient selection flow diagram was shown in Figure 1A.

| Interventions
In the Catheter Ablation Group, all antiarrhythmic drugs (AAD) except amiodarone were stopped for at least five half-lives before catheter ablation. The procedure was performed by experienced physicians. AF ablation strategy of our study has been described previously. 11 The procedures were performed in patients under conscious sedation. A continuous irrigated radiofrequency ablation was performed along each pulmonary vein antrum in order to encircle the ipsilateral pulmonary veins. Procedural end-points were electrical isolation of all pulmonary veins in patients with paroxysmal AF. In patients with persistent AF, LA roofline, mitral isthmus, and cavotricuspid isthmus were routinely targeted. Pulmonary vein isolation and linear block were identified in sinus rhythm.
Patients underwent catheter ablation were treated with anticoagulant drugs (warfarin, non-vitamin K antagonist oral anticoagulant) and AADs (amiodarone, sotalol, propafenone) for 3 months after the procedure. AADs were withdrawn in the patients without recurrence 3 months after the catheter ablation. The patients with high risk for stroke were encouraged to continue taking anticoagulation. Antithrombotic strategy was determined by the discussion between physicians and the patients according to the patients' thromboembolism and bleeding risk and the patients' intention. The long-term antithrombotic and AAD therapy in the ablation group were adopted the data after 3-month follow-up.
In the drug therapy group, the long-term plan of drug therapy was determined by the professional physicians at the first visit.
According to the patients' thromboembolism and bleeding risk, antiplatelet drugs (aspirin, clopidogrel, and ticagrelor) or/and anticoagulant drugs (warfarin, non-vitamin K antagonist oral anticoagulant) would be prescribed by the physicians. According to the patients' symptom and rhythm, the physicians would prescribe AADs (amiodarone, sotalol, propafenone) or/and rate control drugs (betablockers, non-dihydropyridine calcium channel blockers, digoxin).

| Follow-up
Scheduled follow-up was implemented at 3, 6, and 12 months after the initiation and every 6 months thereafter. Three strategies were applied to monitor heart rhythm: (i) regular reexamination: 24 hours-Holter was performed monthly in the first 3 months, which was followed by an ECG and/or 24 hours-Holter every 6 months thenceforth; (ii) symptom triggered reexamination: patients would record ECGs whenever experienced AF symptoms; and (iii) opportunistic screenings: ECGs recorded for routine examinations or other diseases were involved.

| Study endpoints
The primary endpoint was the composite of thromboembolism, coronary events, major bleeding, and all-cause death. The thromboembolism included ischemic stroke (IS), transient ischemic attack (TIA), and systemic embolism (SE). The coronary events included MI and coronary revascularization. The major bleeding was defined according to the International Society on Thrombosis and Hemostasis (ISTH). 12 The secondary endpoints included thromboembolism, coronary events, major bleeding, allcause death, and AF recurrence. AF recurrence was defined as AF, atrial flutter, or atrial tachyarrhythmia lasting ≥30 seconds record by 12-lead electrocardiogram or 24 hours-Holter after a 3-month blanking period from the initiation in both groups. Once a patient underwent catheter ablation during the follow-up period, the follow-up data after the procedure would not be considered in the survival analysis.

| Statistical analysis
Propensity-score matching was used to reduce the selection bias. To generate the propensity score, we made a multiple regression model enrolling 12 baseline variables, including the age, gender, type of AF, MI, heart failure, hypertension, diabetes, history of bleeding, history    With a mean duration of 54.5 ± 24.0 months follow-up, after a single ablation, the AF recurrence-free rate was higher in the catheter ablation group than the drug therapy group (53.50 vs. 18.41%; p < .01; Figure 1B).

| Study endpoints
Clinical adverse events were shown in

| Subgroup analysis
The patients were divided into two different stroke risk subgroups based on the CHA2DS2-VASc score. The patients with CHA2DS2-VASc score ≤ 1 in men or ≤ 2 in women were grouped into the low-medium risk group, and the patients with CHA2DS2-VASc score ≥ 2 in men or ≥ 3 in women were grouped into the high-risk group. As it was shown in Table 4, catheter ablation was not a risk factor for the primary endpoint in either group. For all-cause death, catheter ablation was an independent risk factor in the high-risk group

| AF and CAD
AF and CAD were closely related, and they shared some same risk factors, such as hypertension, diabetes, sleep apnea syndrome, obesity, smoking, and inflammation. CAD could affect the blood supply of atria, and coronary artery revascularization could reduce the recurrence of AF. 13 Alasady 14 et al. found that CAD affecting the atrial artery was independently associated with AF. By measuring coronary artery blood with a microcatheter, it was found that left atrial hypoperfusion existed in patients with lone AF. 15 The atrial ischemia caused by CAD could lead to ion channel disorder, electrical and mechanical reconstruction, fibrosis and scarring, and even local conduction block that could induce and maintain AF. 16 On the other hand, the incidence rate of CAD was higher in patients with AF than heathy persons. AF could predict CAD, independent of conventional risk factors. 17 Besides, AF could predict MI in patients with and without CAD. 18,19 AF impaired myocardial perfusion, which could be improved by cardioversion. 20 With controlled ventricular rate, AF still could independently reduce atrial hypoperfusion by its irregularity. 21 Furthermore, AF could damage the endothelial cell by its special hemodynamics, and these findings were reversible after electrical cardioversion. 20,22 Endothelial dysfunction was probably attributed to the rise of the asymmetric dimethylarginine levels and the downregulation of endothelial nitric oxide synthase expression, which were caused by AF. [23][24][25] AF could also lead to increase in platelet activation and the level of plasma inflammatory mediators, which were risk factors for coronary events. 25 These pathological states for CAD could also be reversed by maintaining sinus rhythm. 23,25

| Catheter ablation for AF
In the recently published CABANA study, catheter ablation was not superior to drug therapy on the primary endpoint, which was a composite of death, disabling stroke, serious bleeding, and cardiac arrest. 26 In the CABANA study, only about 19% of participants suffered from CAD, the outcomes of patients with CAD and AF have not been analyzed in the CABANA study. The CASTLE-AF study, a randomized controlled trial, confirmed that catheter ablation was superior to drug therapy in patients with concurrent AF and heart failure.
However, the information about impacts of CAD on the prognosis was also limited. 27 Accordingly, the data of catheter ablation on the prognosis in the patients with concurrent AF and SCAD were very limited. This study added some information about the issue of on the outcomes in the patients with AF and SCAD.
A small observational trial of catheter ablation therapy vs. medical therapy in the patients with AF and prior coronary intervention provided evidence that catheter ablation may reduce adverse events, The patients with CHA2DS2-VASc score ≤ 1 in men or ≤ 2 in women were grouped into the low-medium risk group, and the patients with CHA2DS2-VASc score ≥ 2 in men or ≥ 3 in women were grouped into the high-risk group.
including acute coronary syndrome requiring hospitalization, stroke, pulmonary embolism and mortality. 8 Our study found that catheter ablation could not improve the composite primary endpoints in the patients with AF and SCAD. However, catheter ablation could reduce all-cause mortality. The difference between the two studies might be due to the different study population. The exclusion of patients without PCI in the aforementioned study led to poor generalization. Our study with larger sample size and longer follow-up may influence the treatment choice for the patients with AF and SCAD. Furthermore, our study showed that whether catheter ablation therapy was associated with fewer all-cause mortality depended on the stroke risk based on CHA2DS2-VASc scores.

| Drug therapy
According to recent guidelines, patients with AF and CAD were suggested to take an anticoagulant or at last an antiplatelet drug. 28 29 In our study, only 21.5% of patients in catheter ablation group and 28.6% of patients in drug therapy group took oral anticoagulation therapy, which was not in accordance with the guidelines. 30 Previous study showed that only 36.5, 28.5, and 21.4% of patients in the CAFR with CHA2DS2-VASc scores ≥2, 1, and 0 underwent oral anticoagulation therapy. 31 In other Chinese AF centers, according to studies conducted between 2017 and 2020, the proportions of oral anticoagulation use were from 13.9 to 35.6%. [32][33][34][35] As a current status, oral anticoagulation was significantly underused in patients in China. In addition, the application of antiplatelet drugs or anticoagulation in Chinese patients was hard to sustain. 31,36 The problems with patients' compliance, the misgiving about the risk of bleeding, and the high cost of the non-vitamin K antagonist oral anticoagulants were possible reasons for the lack of anticoagulation therapy. This study revealed the gap between the clinical practice in the real world and the guidelines in China, which might contribute to the high incidence of adverse events in our study.

| Thromboembolism events and major bleeding events
In most registries on AF, stroke was much less common than bleeding.
However, the incidence of thromboembolism events was higher than that of major bleeding in this study. Compared with bleeding, Chinese were more likely to suffer from thromboembolism events. It was reported in an observational study enrolling 9806 AF patients with anticoagulation therapy, the incidence of IS and SE was 6.5%, and the incidence of intracranial hemorrhage and gastrointestinal bleeding was 4.9%. 38 In another retrospective study, the incidence of IS and SE was 2.8%, and the incidence of major bleeding events was 1.1% in the patients with AF treated with dabigatran in China. 39 These studies from China provided the evidences that there might be racial difference in the incidence of adverse events. In this study, about 37.5% of patients did not take any antithrombotic drugs, which might cause a further increase in thromboembolism events and reduce major bleeding events.

| Limitations
Some main limitations existed in this study. (1) Our research was a prospective cohort study rather than a randomized controlled trial.
Selection bias could exist for the inherent deficiency of observational study. For balancing patients' characteristics, propensity-score matching was used to reduce the bias. There were also some slight differences of the baseline characteristics between the two groups.
Multivariate analysis was used to adjust for potential confounding fac-

| CONCLUSION
In conclusion, in the patients with AF and SCAD, compared with drug therapy, after adjusted the cofounders, catheter ablation was not significantly associated with fewer primary composite endpoints of thromboembolism, coronary events, major bleeding, and all-cause death. However, catheter ablation could lead to fewer all-cause death.

DATA AVAILABILITY STATEMENT
Research data are not shared.