Optimal duration of anticoagulation after successful catheter ablation for atrial fibrillation: insights from recent randomized trials

Abstract

The optimal duration of oral anticoagulation (OAC) after successful catheter ablation (CA) for atrial fibrillation (AF) remains controversial. Current guidelines recommend indefinite OAC based on stroke risk profiles, but long-term therapy adherence is suboptimal due to bleeding concerns. Recent randomized trials—ALONE-AF and OCEAN—provide critical insights into this debate. The ALONE-AF trial demonstrated that discontinuing OAC in patients without arrhythmia recurrence ≥1 year post-ablation (mean CHA₂DS₂-VASc score 2.1) significantly reduced major bleeding without increasing thromboembolic risk. However, the positive composite endpoint was predominantly driven by bleeding reduction, and the stroke comparison alone was underpowered. Similarly, the OCEAN trial showed no additional benefit of rivaroxaban over aspirin in stroke prevention among patients with successful ablation, while highlighting higher bleeding risks with anticoagulation. Importantly, the observed event rate in OCEAN was only ≈0.66 per 100 patient-years—far below the anticipated rate—yielding confidence intervals too wide to conclusively establish thromboembolic safety. These findings support individualized OAC management, particularly for patients with low-to-intermediate stroke risk (CHA₂DS₂-VASc 2–4) and no recurrence. Nevertheless, the persistently low event rates across trials, including OPTION (stroke rate ≈1.2–1.3% despite mean CHA₂DS₂-VASc of 3.5), suggest that post-ablation thromboembolic risk may have been overestimated, and traditional risk scores may no longer accurately stratify risk in this population. Emerging technologies like pulsed‑field ablation (PFA) may offer the potential to shorten OAC duration, but current evidence is derived almost entirely from thermal‑ablation cohorts and limited to small, underpowered PFA studies. Future studies should focus on high-risk subgroups and novel ablation strategies to refine personalized antithrombotic therapy.

Catheter ablation (CA) has become a cornerstone in the management of atrial fibrillation (AF), with rapidly increasing procedure volumes worldwide¹. However, the optimal duration of oral anticoagulant (OAC) therapy post-ablation remains a subject of intense debate. Current guidelines, such as those from the ACC/AHA and ESC, recommend indefinite OAC based on stroke risk profiles (e.g., CHA₂DS₂-VASc score) rather than CA success. This approach is predicated on the recognition that AF recurrence is common even after apparently successful procedures, and subclinical arrhythmias may persist. Furthermore, current clinical studies and multiple meta-analyses have failed to yield consistent evidence regarding the safety of discontinuing anticoagulation therapy, contributing to the ongoing clinical dilemma^1–3. Nonetheless, long-term OAC adherence is often poor due to bleeding concerns and lifestyle demand, particularly in patients with intermediate stroke risk (CHA₂DS₂-VASc scores of 2–4), where the net clinical benefit of continued anticoagulation is uncertain.

Thromboembolic events after AF ablation are relatively rare but carry significant morbidity. Registry data indicate that the annualized stroke rate post-ablation ranges from 0.3% to 1.4%, depending on patient risk factors and monitoring intensity⁴. Embolic events predominantly occur in patients with documented AF recurrence or high baseline stroke risk². A meta-analysis included 5877 patients from 18 eligible randomized controlled trials (RCTs) founding that CA was associated with a significant reduction in stroke (RR 0.63, 95% confidence interval 0.45–0.87, P = 0.006)⁵. Another meta-analysis of prospective studies involving over 11,000 patients found that the incidence of thromboembolism was 1.1% in those who discontinued OAC versus 1.4% in those who continued, with no significant difference—suggesting that successful ablation may mitigate stroke risk to levels comparable to the general population⁴. However, high-risk subgroups, such as those with prior stroke or CHA₂DS₂-VASc scores ≥4, experience elevated event rates. In a German ablation registry, patients with a history of stroke of paroxysmal AF had a thromboembolism rate of 4.3% during 489 (453–782) days of follow-up, compared to 0.3% in those without, underscoring the need for tailored strategies⁶.

Guidelines universally advocate for indefinite OAC in patients with CHA₂DS₂-VASc scores ≥2 post-ablation, regardless of rhythm status. This recommendation stems from studies showing that AF recurrence can occur years after ablation, with a 40–80% recurrence rate at 10 years. However, long-term OAC adherence is suboptimal; real-world data indicate that only 31.3% of patients continue OAC at 12 months post-ablation. Meta-analyses confirm that ongoing OAC therapy significantly increases major bleeding risks^4,7. A pooled analysis of the NOAH-AFNET 6 and ARTESIA trials demonstrated that direct OACs reduced stroke risk (relative risk [RR] 0.65 for stroke or systemic embolism) but increased major bleeding (RR 1.62)⁸. This risk-benefit profile is particularly relevant for patients with intermediate scores (2–4), where net benefit may be marginal.

Two recent RCTs—the ALONE-AF trial⁹ and the OCEAN trial¹⁰—provide groundbreaking insights into this debate. Both studies focused on patients without documented atrial arrhythmia recurrence for at least 1 year after ablation, but they differed in design and outcomes, offering complementary evidence on whether OAC can be safely discontinued post-ablation.

The ALONE-AF trial randomized 840 patients without documented atrial arrhythmia recurrence for at least 1 year (3.6 years (SD, 3 years)) post-ablation to discontinue or continue OAC. Patients had a mean CHA₂DS₂-VASc score of 2.1 and were assigned to either discontinue OAC or continue direct OACs (e.g., apixaban 5 mg twice daily or rivaroxaban 20 mg once daily). At 2 years, the primary composite outcome (stroke, systemic embolism, or major bleeding) occurred in 0.3% of the discontinuation group versus 2.2% of the continuation group (absolute difference −1.9 percentage points; P = 0.02), driven predominantly by a reduction in major bleeding (0% vs. 1.4%). Ischemic stroke rates were low and comparable between groups (0.3% vs. 0.8%), suggesting that discontinuation is safe in this population. The trial’s design censored patients upon AF recurrence.

The OCEAN trial provides complementary evidence by comparing rivaroxaban (15 mg daily) to aspirin (70–120 mg daily) in 1284 patients (a mean CHA₂DS₂-VASc score of 2.2) with successful ablation ≥1 year prior. After 3 years, the primary outcome (stroke, systemic embolism, or new covert embolic stroke on MRI) occurred in 0.8% of the rivaroxaban group and 1.4% of the aspirin group (0.31 vs. 0.66 events per 100 patient-years; relative risk 0.56; P = 0.28), with no significant difference in efficacy. Crucially, rivaroxaban was associated with higher rates of major bleeding (1.6% vs. 0.6%) and clinically relevant nonmajor bleeding (5.5% vs. 1.6%). The inclusion of MRI-detected covert embolic strokes strengthens the conclusion that anticoagulation offers no additional protection in this population. These results indicate that in a population with successful ablation, intensified anticoagulation does not provide superior protection against thromboembolism but unequivocally increases hemorrhagic complications. The low overall event rates—96% of patients had no new cerebral infarcts on MRI—underscore the reduced stroke risk post-ablation.

While both trials support discontinuing OAC after successful ablation, their designs highlight critical nuances. ALONE-AF excluded patients with recurrence, emphasizing a “clean” population where OAC cessation appears safe. In contrast, OCEAN included recurrences but still found no benefit from anticoagulation, suggesting that even with occasional AF episodes, the stroke risk remains low if the overall burden is minimal. This resonates with studies on subclinical AF, where anticoagulation offers limited net benefit in low-burden cases.

However, several methodological caveats must be emphasized. First, neither trial was adequately powered to detect differences in thromboembolic events alone. In OCEAN, the observed event rate in the aspirin arm (≈0.66 per 100 patient-years) was a small fraction of the anticipated rate; a post-hoc calculation suggests that >11,000 patients would be required for sufficient statistical power to assess thromboembolic safety. Consequently, the confidence intervals around the primary efficacy outcome were wide (relative risk 0.56; 95% CI [please add if available]; P = 0.28), precluding definitive conclusions about non-inferiority of a less intensive antithrombotic strategy. Similarly, in ALONE-AF, the composite endpoint benefit was driven almost entirely by a reduction in major bleeding—a near-mechanical consequence of withdrawing anticoagulation—while the ischemic stroke comparison (0.3% vs. 0.8%) remained underpowered.

Second, these power limitations should not be interpreted solely as weaknesses. The persistently low event rates across contemporary trials—including OPTION, which reported stroke rates of ≈1.2–1.3% despite a mean CHA₂DS₂-VASc of 3.5—strongly suggest that the residual thromboembolic risk after successful ablation has been systematically overestimated. This implies that traditional risk-stratification instruments, such as the CHA₂DS₂-VA score, may no longer accurately reflect stroke risk in the modern post-ablation setting, where effective rhythm control and left atrial reverse remodeling may fundamentally alter the underlying substrate for thrombogenesis. These observations call for a reappraisal of guideline recommendations and underscore the need for dedicated, adequately powered trials in this population.

However, several caveats merit attention. First, both trials underrepresented high-risk patients (e.g., CHA₂DS₂-VASc ≥4 or prior stroke). In OCEAN study, for instance, only 10.6% of participants had a score ≥4 and both studies, patients with prior stroke comprised only about 5% of the cohort. In such subgroups, the benefit-risk balance may differ. Second, in the OCEAN study, for patients with AF recurrence, the aspirin group had higher embolism rates (0/134 vs. 4/124), highlighting that recurrence status, not baseline score alone, should guide therapy. Third, AF burden—not just recurrence—should be considered; data from control clinical trial indicate that burdens <0.1% may confer low stroke risk, potentially allowing for OAC discontinuation even with occasional arrhythmia¹¹. Fourth, as noted above, the low event rates in OCEAN and ALONE-AF limit the statistical power of their thromboembolic comparisons, meaning that the apparent safety of discontinuation must be interpreted cautiously. Conversely, the consistency of low event rates across trials—including OPTION—supports the hypothesis that post-ablation stroke risk is genuinely lower than historically assumed. Therefore, the definition of “successful ablation” relied on intermittent monitoring, which may miss asymptomatic recurrences; continuous monitoring could refine risk stratification.

The timing of OAC discontinuation is another key consideration. Based on current evidence, stopping OAC 1 year post-ablation in recurrence-free patients seems reasonable. Yet, emerging technologies like pulsed-field ablation (PFA) may alter this paradigm. However, nearly all evidence supporting less intensive post‑ablation anticoagulation comes from thermal‑ablation cohorts (radiofrequency or cryoballoon) and should not be uncritically extrapolated to PFA. A single prospective study comparing PFA to radiofrequency ablation¹² found that discontinuing OAC as early as 1‑month or 2‑month post‑PFA resulted in negligible stroke events (0.5% vs. 5.8% with radiofrequency), suggesting that ablation modality could influence thrombogenicity. Nevertheless, the low absolute event rates make this study underpowered to characterize cerebrovascular safety reliably. Moreover, recent signals temper any assumption of a superior PFA profile: the VARIPULSE U.S. early‑launch phase was paused in January 2025 after a peri‑procedural neurovascular event rate of 3% (4 of 132) versus the 1.1% observed in the pre‑market study (source: FDA enforcement report, 2025)¹³. This highlights that cerebrovascular safety with PFA cannot be assumed superior to thermal ablation and must be evaluated by specific catheter platform rather than across “PFA” as a single category. Until larger, adequately powered, platform‑specific trials are available, any recommendation for earlier OAC cessation after PFA remains investigational and should be approached with caution.

Future directions and conclusion

In light of these trials, clinicians should adopt an individualized approach. For patients with low to intermediate stroke risk (CHA₂DS₂-VASc 2–4) and no recurrence 1 year after ablation, discontinuing OAC is a valid option that reduces bleeding without significantly increasing thromboembolic risk. Shared decision-making is essential, incorporating patient preferences and bleeding risk scores like HAS-BLED. For high‑risk patients, such as those with prior stroke or CHA₂DS₂‑VASc ≥4, continued OAC remains the more defensible default. The OPTION study previously suggested left atrial appendage occlusion (LAAO) as an alternative in this subgroup¹⁴; however, the recent CLOSURE‑AF trial¹⁵ randomized 912 very‑high‑risk patients (CHA₂DS₂‑VASc >2, high bleeding risk) to LAAO versus best medical therapy and did not meet noninferiority, arguing against routine LAAO in exactly this population. Therefore, LAAO cannot currently be recommended as a routine substitute for OAC in high‑risk post‑ablation patients, and continued OAC (or enrollment in dedicated LAAO trials) is preferred until further data emerge. With respect to PFA, the current evidence base is insufficient to recommend routine early OAC discontinuation. Cerebrovascular safety profiles appear to differ across PFA platforms, and the low event rates in existing studies preclude definitive conclusions. Dedicated, adequately powered, platform‑specific trials are urgently needed.

In conclusion, the ALONE-AF and OCEAN trials mark a paradigm shift in post-ablation care, challenging the dogma of indefinite anticoagulation. While they affirm the safety of OAC discontinuation in select populations, future studies must focus on high-risk subgroups and the role of novel ablation technologies -- particularly PFA, where safety data remain immature and platform‑dependent. As we move toward personalized medicine, these findings empower clinicians to tailor antithrombotic therapy, balancing efficacy and safety in the era of catheter ablation.

Author contributions

Conceptualization, M.J.L., B.R., and J.Q.Z; Methodology, M.J.L.; Software, B.R.; Validation, K.Z.; Formal analysis, M.J.L. and B.R.; Investigation, B.R.; Resources, B.R.; Data curation, M.J.L. and K.Z.; Writing—original draft preparation, M.J.L. and B.R.; Writing—review and editing, K.Z. and J.Q.Z.; Visualization, B.R; Supervision, K.Z. and J.Q.Z.; Project administration, K.Z. and J.Q.Z. All authors have read and agreed to the published version of the manuscript.

Peer review

Peer review information

Communications Medicine thanks all reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Funding

This work was supported by the Natural Science Foundation of China (81970282, 82170361, 82270331); the Natural Science Foundation of Shandong Province (No. ZR2021MH100); and the Qingdao Key Clinical Specialty Elite Discipline (QDZDZK-2022008). The funding sources had no roles in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at https://doi.org/10.1038/s43856-026-01944-5.