Long-Term Outcomes of Anticoagulation Monotherapy Versus Combination Therapy in Atrial Fibrillation Patients with Complex Coronary Artery Disease

Introduction

The strategy of using combined oral anticoagulant (OAC) and antiplatelet therapy (APT) for patients with atrial fibrillation (AF) and complex coronary artery disease (CAD), including left main disease, multivessel disease, and chronic total occlusion, remains a challenging issue in long-term clinical management. Beyond the initial 12 months following percutaneous coronary intervention (PCI), determining the optimal antithrombotic regimen for patients with AF and multivessel disease is particularly difficult. This population faces competing risks of thromboembolism, stent-related ischemia, and major bleeding, with pivotal studies guiding a shift toward a minimalist, individualized approach. The AFIRE trial provided high-level evidence that rivaroxaban monotherapy was non-inferior to rivaroxaban plus single APT for efficacy endpoints and significantly reduced the risk of major bleeding [1]. Similarly, the EPIC-CAD trial demonstrated that edoxaban monotherapy significantly lowered the composite endpoint of death, myocardial infarction (MI), stroke, systemic embolism, unplanned revascularization, and major bleeding, compared with dual therapy (OAC plus APT) [2]. A meta-analysis by Ahmed et al confirmed these results, showing that OAC monotherapy is associated with a 45% reduction in major bleeding risk, without sacrificing protection against ischemic events [3]. However, a Korean nationwide study showed that patients receiving OAC plus APT had a lower incidence of major adverse cardiovascular events than did those on OAC monotherapy, while bleeding risks were similar between the groups [4]. These discrepancies can reflect residual confounding or differences in patient selection, particularly the inclusion of higher-risk patients in real-world practice. A meta-analysis by Hayek et al further supports the notion that monotherapy beyond 1 year reduces bleeding risk without significantly affecting major adverse cardiovascular events, although a slight increase in stent thrombosis remains a concern, particularly in patients with recent PCI [5]. From a pathophysiologic standpoint, modern drug-eluting stents (DES) and improved re-endothelialization profiles have reduced the need for prolonged dual APT, especially in individuals at high bleeding risk [6]. Looking ahead, the ongoing ADAPT AF-DES trial aims to provide definitive evidence comparing direct OAC monotherapy with direct OAC plus clopidogrel in patients with AF more than 1 year after DES implantation [7]. Until then, clinical guidelines recommend OAC monotherapy as the standard long-term strategy for most patients with AF and stable CAD beyond 1 year after PCI [8,9]. Most of the populations in previous studies were patients with stable CAD, and less than half had a history of prior MI or multivessel disease. Therefore, the evidence remains limited regarding optimal antithrombotic strategies beyond 1 year in patients with AF undergoing complex PCI. In clinical practice, the combination of APT is often still maintained when physicians are confronted with highly complex CAD. Therefore, in this study, we aimed to investigate the effect of OAC plus APT beyond 1 year after PCI in an AF population with complex CAD.

Material and Methods

PATIENT POPULATION:

Between January 2018 and December 2023, a total of 179 patients with AF and multivessel disease received diagnosis at our hospital. After excluding patients who underwent coronary artery bypass grafting, patients with a low SYNTAX score (≤22), patients requiring additional revascularization within 1 year, patients lost to follow-up, and patients who died, 56 patients with complex CAD and a SYNTAX score >22 were included in this study. All patients completed 1 year of dual therapy (OAC plus clopidogrel); thereafter, 32 patients switched to OAC alone (group 1), while the remaining 24 patients continued dual therapy of OAC and APT (group 2) (Figure 1).

Data on general demographics, comorbidities, indications, severity of CAD, types of AF, mean CHA2DS2-VA and HAS-BLED scores, and echocardiographic findings were compared between the 2 groups. Baseline echocardiographic parameters were obtained close to the date of PCI. Patients attended outpatient clinic visits at regular 3-month intervals.

ETHICS STATEMENT:

This retrospective study was conducted in accordance with the ethical principles of the 1975 Declaration of Helsinki and was approved by the institutional review board of our hospital (approval number: 11406–014).

DEFINITIONS:

In this study, complex CAD was defined as multivessel disease with a SYNTAX score >22, consistent with previous studies [10,11]. AF was classified according to standard clinical criteria as paroxysmal, persistent, or permanent, based on the duration and pattern of arrhythmia [8,9]. Dual therapy was defined as the concomitant use of an OAC and APT, whereas OAC alone referred to OAC without concurrent APT following the initial 1-year combination regimen [8,9].

STUDY ENDPOINTS:

The primary endpoint was the incidence of composite adverse cardiovascular and bleeding events during follow-up after the first year. The composite endpoint included stroke (ischemic or hemorrhagic), revascularization (target lesion or target vessel), massive gastrointestinal bleeding, cardiovascular mortality, and all-cause mortality. Secondary endpoints were the individual components of the composite outcome, allowing assessment of the relative contributions of ischemic and bleeding complications to overall risk. Clinical outcomes were compared between patients who continued dual therapy beyond 1 year (group 2) and those who transitioned to OAC alone (group 1).

STATISTICAL ANALYSIS:

Data are presented as mean±standard deviation or as numbers and percentages, as appropriate. Continuous variables were compared between groups using the independent samples t test, while categorical variables were analyzed with the chi-square test or Fisher exact test, depending on expected frequencies. Survival and event-free rates were assessed using Kaplan-Meier analysis, and differences between groups were evaluated with the log-rank test. Univariate Cox proportional hazards regression analysis was performed to identify potential predictors of revascularization events occurring more than 1 year after PCI. Subgroup analyses were conducted to explore the incidence of revascularization according to SYNTAX score, presence of left main disease, HAS-BLED score, and clinical presentation (stable angina vs acute coronary syndrome). All statistical analyses were performed using SPSS for Windows (version 22.0; IBM Corp, Armonk, NY, USA). A 2-sided P value <0.05 was considered statistically significant.

Results

BASELINE CHARACTERISTICS OF THE STUDY PATIENTS:

The mean age and sex distributions were similar (76±9.6 vs 75±7.7 years, P=0.869; 68.8% vs 83.3% male, P=0.350 respectively), with no significant differences (Table 1). Indications for PCI, including stable angina, unstable angina, and non-ST-segment elevation MI/ST-segment elevation MI, were evenly distributed (P=0.623). Comorbidities such as hypertension, diabetes, dyslipidemia, prior stroke or MI, heart failure, chronic kidney disease, and valvular heart disease were comparable between groups. The severity of CAD, reflected by the SYNTAX score (27.6±10.3 vs 27.9±9.3, P=0.941), presence of left main disease (25.0% vs 25.0%, P=1.000), and chronic total occlusion lesions (9.4% vs 25.0%, P=0.151), showed no significant differences. Types of AF (paroxysmal vs non-paroxysmal; P=0.430) and serum creatinine levels (1.18±0.35 mg/dL vs 1.12±0.26 mg/dL; P=0.491) were also similar. Medication usage, including angiotensin converting enzyme inhibitors/angiotensin II receptor blockers/Entresto, beta-blockers, statins, and various anticoagulants, showed comparable distributions. Both groups had similar stroke and bleeding risk scores (CHA2DS2-VA and HAS-BLED) and echocardiographic parameters, including left atrial dimension and left ventricular ejection fraction. The only statistically significant difference was the follow-up duration, which was longer in group 2 (4.9±2.1 vs 3.4±2.0 years, P=0.010).

CLINICAL OUTCOMES OF THE STUDY PATIENTS:

The composite endpoint occurred in 66.7% of patients in group 2, compared with only 28.1% in group 1 (P=0.006), indicating a notably higher overall adverse event rate (Table 2). A key contributor to this difference was the significantly increased rate of revascularization in group 2 (12.5% vs 50.0%, P=0.003), including target lesion (9.4% vs 33.3%, P=0.041) and target vessel (9.4% vs 37.5%, P=0.019) revascularization, suggesting a greater burden of recurrent or progressive coronary disease. In group 1, one patient underwent revascularization for unstable angina and 3 patients for stable angina. In group 2, one patient underwent revascularization for non-ST-segment elevation MI and subsequently experienced MI-related mortality, while 6 patients presented with unstable angina and 5 with stable angina. The proportion of acute coronary syndrome-related revascularization did not differ significantly between the 2 groups (group 1 vs group 2, 25.0% vs 58.3%, P=0.264). Although not statistically significant, group 2 also had higher rates of massive gastrointestinal bleeding (3.1% vs 12.5%, P=0.303) and all-cause mortality (9.4% vs 16.7%, P=0.447). Stroke incidence was slightly lower in group 2 (15.6% vs 8.3%, P=0.686). Cardiovascular mortality was similar between the 2 groups (3.1% vs 4.2%, P=1.000).

The Kaplan-Meier curve of the composite endpoints showed a non-significant trend in group 2 (log-rank P=0.06; Figure 2A). The Kaplan-Meier curve of the revascularization rate also showed a non-significant trend toward worse outcomes in group 2 (log-rank P=0.08; Figure 2B).

UNIVARIATE COX REGRESSION ANALYSIS IDENTIFIED THE SIGNIFICANT PREDICTORS OF REVASCULARIZATION MORE THAN 1 YEAR AFTER PCI:

Table 3 presents the univariate Cox regression analyses to identify potential predictors of revascularization more than 1 year after PCI. Among all tested variables, only the combination of APT plus OAC reached statistical significance, with a hazard ratio (HR) of 4.003 (95% CI: 1.287–12.450, P=0.017), indicating that continuation of dual therapy was associated with approximately a 4-fold higher risk of revascularization in univariate analysis. Other factors, including demographic parameters (age, sex), clinical comorbidities (hypertension, diabetes, prior stroke, chronic kidney disease, heart failure), and procedural characteristics (SYNTAX score >33, presence of left main or chronic total occlusion, number of stents >3), did not demonstrate significant predictive value. Notably, prior MI showed a trend toward significance (HR: 4.310; 95% CI: 0.915–20.297; P=0.065), suggesting a possible association of revascularization.

SUBGROUP ANALYSIS OF REVASCULARIZATION INCIDENCE BY CLINICAL AND ANATOMICAL CHARACTERISTICS:

In the comparison of subgroups with SYNTAX score <33 (group 1 vs group 2, 16.7% vs 52.6%, P=0.021), no left main disease (group 1 vs group 2, 16.7% vs 50.0%, P=0.041), HAS-BLED score ≥3 (group 1 vs group 2, 0% vs 66.7%, P=0.021), and acute coronary syndrome (group 1 vs group 2, 5.0% vs 66.7%, P<0.001), there were significantly higher incidences in group 2 (Figure 3).

Discussion

STUDY LIMITATIONS:

Limitations of this study were its retrospective design and reliance of data from a single medical center, which can introduce selection bias and limit the generalizability of our findings. The sample size was relatively small, particularly for the subgroup analyses, which can limit the statistical power to detect meaningful differences between the groups. Given the limited number of patients, we performed only univariate Cox regression, as multivariate analysis was not statistically feasible without risking overfitting. A formal power analysis to determine adequacy of power was not performed, as this was a retrospective study; instead, we included all eligible AF patients with complex PCI and high SYNTAX scores during the study period, to maximize case capture. Accordingly, our findings should be regarded as exploratory and hypothesis-generating rather than definitive. The dual-therapy group had a significantly longer follow-up period, which may have contributed to the higher cumulative event rates observed. To address this, we performed Kaplan-Meier survival analyses and Cox regression, which adjusted for differences in follow-up duration. Even after accounting for time-to-event, the dual-therapy group continued to show higher adverse outcomes, suggesting that longer follow-up alone did not fully explain these findings. Additionally, physicians may have been more likely to continue dual therapy in patients perceived to have greater anatomical complexity or higher ischemic risk, introducing selection bias that could not be fully adjusted for. Such inherent differences in baseline risk or procedure-related complexity may partly explain the higher incidence of major adverse cardiovascular events observed in the dual-therapy group. Hence, these questions are more appropriate to address in randomized controlled trials rather than in observational studies. Our work should therefore be considered hypothesis-generating, as stated above, and we emphasize the need for adequately powered prospective randomized controlled studies to confirm the optimal long-term antithrombotic strategy in this high-risk population. Another limitation was that AF burden was not systematically recorded in this study. While AF type (paroxysmal vs non-paroxysmal) was included in the analysis and did not emerge as a significant predictor of outcomes, more granular assessment of AF burden (eg, arrhythmia duration, frequency, or device-detected AF) was unavailable. This may have limited our ability to explore potential interactions between AF burden, CAD complexity, and antithrombotic strategy. Finally, we speculated that discontinuation or interruption of antithrombotic therapy in the dual-therapy group may have contributed to adverse outcomes. However, because adherence and discontinuation were not systematically recorded in our retrospective dataset, we could not confirm whether therapy interruption directly led to events. Despite these limitations, we believe this study provides important preliminary insights into patients with AF undergoing complex PCI with high SYNTAX scores – a population often underrepresented in randomized controlled trials. Prospective multicenter studies with larger cohorts will be required to validate our findings and clarify the optimal long-term antithrombotic strategy in this setting.

Conclusions

In AF patients with complex CAD, continuation of OAC plus APT beyond 1 year after PCI was associated with a higher rate of adverse clinical outcomes and a greater need for repeat revascularization, compared with OAC alone. These findings suggest that long-term OAC monotherapy can represent a safer and equally effective alternative for selected high-risk patients. However, given the retrospective design and limited sample size, our findings suggest that OAC monotherapy warrants prospective validation in AF patients with complex PCI, as these retrospective results should be regarded as hypothesis-generating.