Skip to main content

Anticoagulants compared with aspirin in patients with embolic strokes of undetermined source: an individual participant data meta-analysis of randomised trials

In brief

Anticoagulants do not prevent more recurrent strokes than aspirin after ESUS

In a pooled analysis of four randomized trials involving nearly 14,000 people with embolic stroke of undetermined source, recurrent stroke rates were similar with anticoagulants and aspirin: 4.6 versus 4.8 events per 100 patient-years. Major bleeding also did not differ significantly, though results varied across trials; a possible benefit for people with enlarged atria remains unproven and needs further study.

Journal
The Lancet. Neurology (Q1)
Published
20 November 2026
Study design
Systematic review / meta-analysis of RCTs
Evidence level
Level 1, High (CEBM 1a)
Authors
Aristeidis H Katsanos, Mukul Sharma, Hans Christoph Diener, Hooman Kamel, Sven Poli, Geoffrey A Donnan, et al.
PMID
43063957
DOI
10.1016/S1474-4422(26)00366-2

Why clinicians should know about it

  • Picked for Neurology (clinical) (top studies of the week, 22 November 2026): Individual-participant meta-analysis informs ESUS secondary prevention

Abstract

BACKGROUND: Direct-acting oral anticoagulation did not reduce recurrent stroke compared with aspirin in randomised controlled trials (RCTs) of embolic stroke of undetermined source (ESUS), and subgroup benefits suggested by analyses of individual RCTs were inconsistent across trials. We aimed to evaluate the effect of anticoagulation on the risks of recurrent stroke and major bleeding in patients with ESUS, overall and in prespecified subgroups using individual participant data from RCTs. METHODS: In this systematic review and meta-analysis, we searched MEDLINE and Scopus for RCTs published between April 1, 2014, and March 1, 2025, without language restriction, comparing anticoagulants with aspirin in patients with ESUS. Non-randomised studies and studies not enrolling patients with ESUS were excluded. We requested individual participant data from the principal investigators of all eligible trials. The primary outcomes were recurrent stroke and major bleeding. For each outcome of interest, we estimated the treatment effect for the overall trial population separately within each trial and the resulting overall treatment estimates were subsequently pooled across trials. Risk of bias was assessed with the Cochrane Risk of Bias tool (version 2.0) for the primary efficacy outcome of stroke recurrence. Hazard ratios (HRs) with corresponding 95% CIs were estimated following adjustment for prespecified covariates and pooled by use of the random-effects model. This study is registered with PROSPERO (CRD42025644724). FINDINGS: Our systematic database search identified 1170 records. Of these studies, four were eligible for inclusion in the analysis, including a total of 13 970 participants. Mean age of participants was 66 years (SD 10·6); 5486 (39·3%) participants were female and 8484 (60·7%) were male. Across all trials, 6989 patients were assigned to receive a direct-acting oral anticoagulant and 6981 were assigned to receive aspirin. The hazard for recurrent stroke was similar between participants allocated to anticoagulant or aspirin (399 of 6989 patients [4·6 events per 100 patient-years] in the anticoagulant group vs 417 of 6981 [4·8 events per 100 patient-years] in the aspirin group; adjusted HR 0·94, 95% CI 0·78-1·14), without evidence of heterogeneity across studies (I2=5%, p for Cochran's Q test=0·37). The hazard for major bleeding was also similar between the two groups (149 of 6989 patients [1·6 events per 100 patient-years] in the anticoagulant group vs 106 of 6981 [1·1 events per 100 patient-years] in the aspirin group; adjusted HR 1·19, 95% CI 0·52-2·73), but heterogeneity was present across studies (I2=79%, p for Cochran's Q test=0·0023). INTERPRETATION: Anticoagulation did not reduce the risk of recurrent stroke after ESUS when compared with aspirin, and the risk of major bleeding did not differ significantly between the two treatments. A possible benefit from anticoagulation in patients with ESUS and left atrial enlargement, as shown in a hypothesis-generating subgroup analysis, should be investigated in future trials. FUNDING: Division of Neurology Innovation Fund, McMaster University.

Abstract as published, via PubMed.

View on PubMedFull text at the publisherOpen in the app

For healthcare professionals. The summary is generated by AI from the published abstract, and the evidence level is assigned automatically from the study design on the Oxford CEBM hierarchy. Neither is medical advice. Read the full paper before changing practice.