Highlight
– In the randomized OCEAN trial (n=1284), rivaroxaban (15 mg daily) did not significantly reduce a composite of stroke, systemic embolism, or new covert embolic stroke on MRI compared with aspirin over 3 years.
– Event rates were low in both arms (0.31 vs 0.66 events per 100 patient‑years), producing wide confidence intervals and limited statistical power for modest differences.
– Fatal or major bleeding was numerically higher with rivaroxaban (1.6% vs 0.6%; HR 2.51), though the estimate was imprecise.
Background
Atrial fibrillation (AF) is a major cause of ischemic stroke. For patients with AF, oral anticoagulation (OAC) reduces stroke risk and is generally recommended on the basis of CHA2DS2‑VASc score rather than the presence or absence of symptoms or rhythm control success. Catheter ablation reduces AF burden and can restore and maintain sinus rhythm in many patients. Whether durable rhythm control following successful ablation allows safe discontinuation or replacement of OAC is an important and unresolved clinical question.
Observational reports and small imaging studies have shown both symptomatic and small covert cerebral infarcts after ablation procedures, but data on long-term thromboembolic risk after successful ablation are limited. Guideline panels have generally advised that anticoagulant decisions should be based on stroke risk (CHA2DS2‑VASc) rather than rhythm status, but this remains an area of active debate, especially for patients who are asymptomatic and have had no recurrent AF for prolonged periods.
Study design
The OCEAN trial was an international, open‑label, randomized, blinded‑outcome‑assessment study enrolling 1,284 patients who had undergone successful catheter ablation for AF at least 1 year earlier and had a CHA2DS2‑VASc score of ≥1 (or ≥2 for women or for patients in whom vascular disease was a risk factor). Patients were randomized 1:1 to aspirin (70–120 mg daily, locally available dose) or rivaroxaban 15 mg daily and followed for 3 years. Brain MRI was obtained at enrollment and again at 3 years to detect covert (subclinical) embolic infarcts. The primary outcome was a composite of clinical stroke, systemic embolism, or new covert embolic stroke defined as ≥1 new infarct ≥15 mm on MRI at 3 years. Major bleeding and fatal bleeding formed key safety outcomes.
Key features that strengthen trial validity included randomized allocation and blinded adjudication of outcomes (including imaging). The trial was open label for treatment assignment, a practical constraint when comparing aspirin with an oral anticoagulant. The inclusion criterion of at least 1 year since successful ablation selected a population with sustained procedural success and presumably lower AF recurrence risk.
Key findings
Of 1,284 randomized patients, 641 were allocated to rivaroxaban and 643 to aspirin. Over 3 years the primary composite outcome occurred in:
- Rivaroxaban group: 5 patients (event rate 0.31 per 100 patient‑years)
- Aspirin group: 9 patients (event rate 0.66 per 100 patient‑years)
The relative risk for the primary outcome with rivaroxaban versus aspirin was 0.56 (95% CI, 0.19 to 1.65); the absolute risk difference at 3 years was −0.6 percentage points (95% CI, −1.8 to 0.5), P = 0.28. These point estimates numerically favored rivaroxaban but were not statistically significant and had wide confidence intervals that included both clinically meaningful benefit and no effect.
Smaller, sub‑15 mm cerebral infarcts (covert lesions below the primary threshold) were detected in 22 of 568 patients (3.9%) in the rivaroxaban group and 26 of 590 (4.4%) in the aspirin group (RR 0.89; 95% CI, 0.51–1.55), again showing no significant difference.
On safety, the composite of fatal or major bleeding occurred in 10 patients (1.6%) receiving rivaroxaban versus 4 patients (0.6%) receiving aspirin (hazard ratio 2.51; 95% CI, 0.79–7.95). Although numerically higher with rivaroxaban, the bleeding estimate was imprecise.
Interpretation of effect size and statistical power
Event rates were substantially lower than historical estimates for unselected AF populations, reflecting patient selection (successful ablation ≥1 year earlier) and possibly underling low AF recurrence and stroke risk. Because events were uncommon, the trial was underpowered to detect small absolute differences. The inclusion of covert MRI infarcts in the primary composite increased event capture compared with clinical stroke alone but still yielded few events meeting the ≥15 mm threshold.
Notable aspects of the intervention
Rivaroxaban was dosed at 15 mg daily in this trial. In many regions, the licensed dose for stroke prevention in nonvalvular AF is 20 mg daily (or 15 mg for reduced renal function). The use of a 15 mg dose may reflect local regulatory practices in participating countries but is an important consideration when extrapolating results to settings where higher dosing is standard.
Expert commentary and clinical implications
The OCEAN trial addresses a highly practical question: can anticoagulation be safely modified after apparently successful AF ablation? The data do not support routine replacement of OAC with aspirin in patients with risk factors for stroke who are at least one year out from successful ablation. Several key points inform interpretation and application.
First, the trial reinforces a core principle in contemporary AF care: stroke risk is driven largely by patient‑level risk factors (age, prior stroke, heart failure, hypertension, diabetes, vascular disease) and by atrial cardiomyopathy that can persist despite rhythm control. Ablation reduces AF burden but may not abolish the substrate for thrombus formation.
Second, low absolute event rates mean that absolute benefits of anticoagulation (or harms from stopping it) are modest in this selected post‑ablation population. Individual patient decisions will therefore hinge on patient preferences, bleeding risk, monitoring strategies for AF recurrence, and regional practice (including DOAC dosing norms).
Third, the numerically higher major bleeding with rivaroxaban highlights the perennial tradeoff: continued anticoagulation reduces thromboembolism but increases bleeding risk. The confidence intervals are wide, so definitive conclusions about bleeding risk cannot be made from OCEAN alone.
Guideline alignment
Current major guidelines (for example, the 2020 ESC Guidelines for the diagnosis and management of AF) recommend that anticoagulation decisions be based on stroke risk as assessed by CHA2DS2‑VASc and not on rhythm status alone. OCEAN provides randomized evidence that supports this conservative approach: stopping anticoagulation or replacing it with antiplatelet therapy after ablation should not be routine for patients who remain at elevated stroke risk.
Limitations and unanswered questions
- Low event rates and wide confidence intervals limit firm conclusions about small but clinically relevant differences.
- The rivaroxaban dose (15 mg daily) differs from the 20 mg dose used in many regions; whether higher-dose DOACs would yield different results is unknown.
- Open‑label treatment assignment could influence care patterns, though blinded outcome adjudication mitigates bias in endpoint assessment.
- AF recurrence monitoring (beyond clinical follow‑up) and AF burden quantification were not central to the primary analysis; whether very frequent monitoring and aggressive re‑initiation of OAC at the time of subclinical recurrence would alter outcomes remains untested.
- Generalisability is limited to patients who had a successful ablation and remained free of clinically apparent AF for at least 1 year; results do not apply to patients shortly after ablation or with recurrent AF.
Practical takeaways for clinicians
– Do not routinely stop anticoagulation after successful AF ablation solely because the patient is in sinus rhythm; decisions should remain guided principally by CHA2DS2‑VASc score and bleeding risk.
– A discussion of individualized risk–benefit tradeoffs is appropriate for patients with low absolute stroke risk who are reluctant to remain on OAC—shared decision‑making should incorporate the patient’s values and preferences and the limited magnitude of absolute benefit in low‑event populations.
– When considering a change in antithrombotic strategy after ablation, ensure adequate and perhaps enhanced rhythm monitoring so that subclinical AF recurrence can prompt reassessment of anticoagulation.
– Where regional dosing of DOACs differs, clinicians should be cautious extrapolating OCEAN’s rivaroxaban 15 mg data to settings where higher doses are standard.
Conclusion
OCEAN provides the first large randomized evidence specifically addressing antithrombotic strategy after successful catheter ablation for AF. Among patients at least one year after successful ablation with stroke risk factors, rivaroxaban 15 mg did not significantly reduce a composite of clinical and covert embolic stroke compared with aspirin over 3 years; event rates were low and confidence intervals broad. The data reinforce guideline recommendations to base anticoagulation on stroke risk rather than procedural success alone. Decisions to discontinue or alter anticoagulation after ablation should be individualized, accounting for patient stroke and bleeding risk, monitoring strategy, and patient preferences. Further research could examine outcomes with different DOAC doses, more intensive rhythm monitoring strategies, and in populations with different baseline risks.
Funding and clinicaltrials.gov
The trial was funded by Bayer and others. ClinicalTrials.gov identifier NCT02168829.
References
1. Verma A, Birnie DH, Jiang C, et al.; OCEAN Investigators. Antithrombotic Therapy after Successful Catheter Ablation for Atrial Fibrillation. N Engl J Med. 2025 Nov 8. doi:10.1056/NEJMoa2509688.
2. Hindricks G, Potpara T, Dagres N, et al. 2020 ESC Guidelines for the diagnosis and management of atrial fibrillation developed in collaboration with the European Association for Cardio‑Thoracic Surgery (EACTS). Eur Heart J. 2021;42(5):373–498.
3. The CABANA Trial Investigators. Catheter Ablation vs Antiarrhythmic Drug Therapy for Atrial Fibrillation. N Engl J Med. 2019;380:2295–2303.
4. Selected prior MRI and imaging studies of cerebral lesions after AF ablation and reviews of anticoagulation strategy are summarized in the ESC Guidelines and CABANA publications above; clinicians should consult guideline documents for detailed recommendations.