Key takeaways
- The best approach to prevent strokes in high-risk patients with atrial fibrillation who have had an intracranial haemorrhage (brain bleed) remains uncertain.
- The large ENRICH-AF trial compared oral anticoagulation with edoxaban vs. no anticoagulation.
- Edoxaban did not reduce stroke or systemic embolism but increased major bleeding.
- These findings support individualised decision-making and highlight the need for safer stroke-prevention strategies in this vulnerable population.
Munich, Germany – 29 August 2026: Edoxaban did not reduce stroke or systemic embolism but increased major bleeding in high-risk patients with atrial fibrillation who had had an intracranial haemorrhage, according to results presented in a Hot Line session today at ESC Congress 2026.[1]
Atrial fibrillation (AF) is the most common form of irregular heart rhythm and a leading cause of ischaemic stroke, where a blood clot cuts off the blood supply to part of the brain. Oral anticoagulants such as edoxaban are highly effective at preventing ischaemic stroke in patients with AF, but they come with concerns over bleeding risk. These concerns are amplified in patients with AF who have had a prior intracranial haemorrhage, which includes bleeding within or around the brain.
“The ideal strategy to prevent ischaemic stroke is uncertain in patients with AF and a prior intracranial haemorrhage, as these patients have been excluded from the large landmark trials assessing oral anticoagulants,” noted Professor Ashkan Shoamanesh from McMaster University, Hamilton, Canada, the Principal Investigator of the ENRICH-AF trial. “This leaves us with a challenging clinical dilemma that is becoming more frequent with our ageing population. Data from small studies indicate that oral anticoagulation may result in net overall benefit among survivors of intracranial haemorrhage with AF.[2] We conducted the large ENRICH-AF trial to evaluate the efficacy and safety of oral anticoagulants further in these patients,” he explained.
This was an investigator-initiated, open-label, blinded endpoint, event-driven trial conducted across 174 sites in 20 countries. Patients with high-risk AF and prior intracranial haemorrhage were enrolled. Following Data and Safety Monitoring Board review in 2023, no further patients with lobar intraparenchymal haemorrhage or convexity subarachnoid haemorrhage were included because of safety concerns.
Eligible participants were randomised (1:1) to edoxaban 60 mg once daily (with dose adjustments to 30 mg daily according to the label) or no anticoagulation (no antithrombotic therapy or single antiplatelet therapy, as determined by the clinician). The study population included 948 patients who had a mean age of 77 years and 39% were women.
Over an average of 28 months of follow-up, edoxaban did not significantly reduce the primary endpoint of stroke (including ischaemic and haemorrhagic stroke) or systemic embolism compared with no anticoagulation (11.8% vs. 12.8%; hazard ratio [HR] 0.88; 95% confidence interval [CI] 0.61 to 1.26; p=0.48). Ischaemic stroke and myocardial infarction were significantly reduced with edoxaban vs. no anticoagulation; however, this benefit was offset by an almost three-fold excess in haemorrhagic stroke.
The primary safety outcome of International Society on Thrombosis and Haemostasis major bleeding occurred in 11.6% of patients with edoxaban and 5.2% with no anticoagulation (HR 2.23; 95% CI 1.39 to 3.59; p<0.001).
“Our findings do not support the use of edoxaban in unselected patients with AF after intracranial haemorrhage, but highlight the need for an individualised decision-making approach,” commented Professor Shoamanesh. “Ongoing trials, including the double-blinded, randomised ASPIRE trial comparing apixaban vs. aspirin, will add to the evidence base. Additionally, a prospective, individual participant data meta-analysis of all trials (COCROACH) will ultimately provide useful additional insights on how best to individualise optimal stroke prevention in this very vulnerable patient population,” he concluded.
ENDS
