Embolic Stroke of Undetermined Source: Current State of Evidence

Approximately 30% of all ischemic strokes are caused by an unknown etiology, also sometimes referred to as cryptogenic stroke1. Cryptogenic stroke has been defined as a cerebral infarct without a clear attributable etiology despite extensive vascular, cardiac and laboratory evaluations2. The possible underlying mechanisms for these ischemic strokes are varied and may include occult paroxysmal atrial fibrillation, paradoxical embolism with a patent foramen ovale (PFO), hyper coagulable state, non stenotic atherosclerotic disease.

A more recently coined term has been used to describe cryptogenic strokes which appear embolic on brain imaging: Embolic Stroke of Undetermined Source (ESUS). ESUS accounts for more than half of all cryptogenic strokes and has been defined as a cortical infarct without ipsilateral proximal arterial stenosis or a cardio embolic source3. Antiplatelet medications are generally used to prevent recurrent stroke in the absence of atrial fibrillation. However, clinical studies with long term cardiac rhythm monitoring have demonstrated a higher prevalence of atrial fibrillation in the cryptogenic stroke population4,5. With this background two clinical trials were initiated to investigate whether the direct oral anticoagulants were more effective than aspirin for secondary stroke prevention in patients with ESUS.

Results from the RESPECT-ESUS were presented at the World Stroke Congress in October 20186. The trial showed similar stroke rates with the two drugs at 19 months follow up; 4.8% per year with aspirin and 4.1% per year with dabigatran ( p=0.1 HR=0.85). The major bleeding events occurred at a similar rate with the two drugs- 1.4% per year with aspirin and 1.7% per year with dabigatran. Even though these results did not show a statistically significant difference, there is a slight indication of benefit in favor of dabigatran when patients are followed for long term. A post hoc subgroup analysis showed a statistically significant reduction in secondary stroke rates with dabigatran among patients older than 75 years.

NAVIGATE-ESUS7 trial was designed to compare rivaroxaban 15mg/day to aspirin 100mg/day in patients with a recent ESUS event. This was a multi-national trial which enrolled about 7000 patients who were then followed for a median time of 11 months after randomization. The trial was terminated early due to an increased risk of bleeding associated with rivaroxaban without a significant benefit of stroke risk reduction. Recurrent stroke or systemic embolism occurred at 5.1%/year in the rivaroxaban group as compared to 4.8%/year in the aspirin group. In the rivaroxaban group, there were 13 hemorrhagic stroke events as compared to 2 in the aspirin group. Major bleeding rates were also higher in the rivaroxaban group: 1.8%/year vs. 0.7%/year with aspirin.

An exploratory subgroup analysis was subsequently performed and the results were published earlier this month in JAMA Neurology8. The patients were stratified according to clinical characteristics predictive of atrial fibrillation including left atrial size and frequency of premature atrial contractions. Nine percent of the total trial population had a left atrial diameter size greater than 4.6cms. Among this prespecified group, rivaroxaban was associated with a 1.7%/year rate of recurrent ischemic stroke as compared to 6.5%/year with aspirin. This is a statistically significant difference in favor of rivaroxaban- hazard ratio =0.26, P=0.02. This result indicates that rivaroxaban may be beneficial in patients with ESUS who are found to have at least moderate enlargement of the left atrium as these are the patients most likely to have underlying occult atrial fibrillation.

While these results are promising, the findings need to be confirmed with a randomized clinical trial. This is the premise of the ongoing ARCADIA trial9 which will enroll patients with a recent ESUS event and evidence of atrial cardiopathy defined by specific ECG changes, serum biomarkers and left atrial dilatation. The patients will be randomized to receive apixaban 5mg twice daily or aspirin 81mg daily. The results of this trial will hopefully guide treatment decisions for this patient group in the future. In the meantime, based on the current evidence, a broad utilization of anticoagulants cannot be recommended for prevention of ESUS. Physicians should strive to monitor these patients with longer term cardiac rhythm monitoring techniques such as implantable loop recorders.



  1. Infarcts of undetermined cause: the NINCDS Stroke Data Bank. Ann Neurol. 1989 Apr;25(4):382-90
  2. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke. 1993 Jan;24(1):35-41
  3. Embolic strokes of undetermined source: the case for a new clinical construct. Lancet        2014 Apr;13(4):429-38
  4. Atrial fibrillation in patients with cryptogenic stroke. N Engl J Med. 2014 Jun 26; 370(26):2467-77
  5. Cryptogenic Stroke and Underlying Atrial Fibrillation. N Engl J Med 2014; 370:2478-2486
  6. Rivaroxaban for Stroke Prevention after Embolic Stroke of Undetermined Source. N Engl J Med 2018; 378:2191-2201
  7. https://journals.sagepub.com/doi/full/10.1177/1747493018789543
  8. Recurrent Stroke with Rivaroxaban Compared With Aspirin According to Predictors of Atrial Fibrillation: Secondary Analysis of the NAVIGATE ESUS Randomized Clinical Trial. JAMA Neurol. 2019 Apr 8
  9. The AtRial Cardiopathy and Antithrombotic Drugs In prevention After cryptogenic stroke randomized trial: Rationale and methods. International Journal of Stroke 0(0) 1–8