By Michael H. Crawford, MD

Professor of Medicine, Lucy Stern Chair in Cardiology, University of California, San Francisco

SYNOPSIS: An MRI study of cerebral circulation in patients with embolic stroke of undetermined etiology showed evidence of atherosclerotic plaque in most patients.

SOURCE: Lin Tao MM, Li XQ, Hou XW, et al. Intracranial atherosclerotic plaque as a potential cause of embolic stroke of undetermined source. J Am Coll Cardiol 2021;77:680-691.

Ruptured non-stenotic intracranial atherosclerotic plaque has been suspected to cause embolic stroke of undetermined source (ESUS), but there are little data to support this hypothesis. Accordingly, Tao et al evaluated the morphology and composition of intracranial plaque in patients with ESUS and small vessel disease (SVD).

They used the 3.0 Tesla MRI to compare the ipsilateral side to the contralateral side of the stroke. Patients with acute ischemic stroke who had undergone a full evaluation for stroke etiology, including intracranial high-resolution MRI and met criteria for ESUS or SVD, were enrolled retrospectively. Patients with bilateral or posterior circulation strokes were excluded, leaving 243 with ESUS and 160 with SVD. Plaque was defined as eccentric focal wall thickening at the point of minimal lumen diameter in the major vessels in the anterior intracranial circulation. Vessel reference sites were adjacent plaque-free areas. The remodeling index (RI) was the ratio of the cross-sectional area of the vessel at the plaque site to the reference vessel area. Plaque burden (PB) was defined as the percent difference between the total vessel area and the luminal area at the plaque site. Among ESUS patients, 69% had any intracranial plaque. Among SVD patients, 40% had any intracranial plaque (P < 0.001 for the difference).

The prevalence of intracranial plaque was higher in the ipsilateral compared to the contralateral side in ESUS patients (64% vs. 43%; OR, 5.25; 95% CI, 2.83-9.73). In SVD patients, this difference was not found (36% vs. 31%; OR, 2.14; 95% CI, 0.87-5.26; P = 0.13). Also, ESUS patients exhibited larger PB and RI in the ipsilateral compared to the contralateral plaque (PB: 64% vs. 60%; P = 0.002 and RI: 1.17 vs. 1.09; P < 0.001). In addition, complicated plaque was ipsilateral more often in ESUS patients (77% vs. 60%; P = 0.003). None of these plaque features were associated with stroke in SVD patients. A multivariate logistic regression analysis excluding overlapping plaque characteristics showed that RI was independently associated with ESUS (OR, 2.30; 95% CI, 1.66-3.17; P < 0.001). Using an RI cutoff of 1.162, the area under the curve was 0.74. The authors concluded these data suggest high-risk, non-stenotic intracranial plaque represents a significant underestimated embolic source in patients with ESUS.


This is an important foundational study. ESUS represents about 20% of ischemic stroke patients,1 and they are at a high risk of recurrence. Also, there is no established therapy to prevent recurrences. Standard stroke prevention therapies have shown mixed results. In NAVAGATE ESUS, the subgroup with non-stenotic intracranial or generalized atherosclerosis showed no difference in recurrent events between treatment with rivaroxaban or aspirin.2 By contrast, in COMPASS patients with cryptogenic stroke, reduced recurrence rates were observed on low-dose rivaroxaban and aspirin vs. aspirin alone.3 Neither study focused exclusively on patients with intracranial atherosclerosis. Techniques for studying intracranial arterial lesions, such as the MRI technique used in the Tao et al study, are relatively new but have shown excellent intra- and interobserver agreement. Thus, now, there are tools to study this issue more carefully. Hopefully, more effective recurrence-preventing therapies will be discovered.

It has been evident for some time that non-stenotic atherosclerotic lesions in extracardiac vessels proximal to the cerebral circulation, such as the proximal aorta, carotids, and vertebral basilar system, could produce emboli to the brain, but evidence of the potential for intracranial vessel plaques to be a source has been limited by technical issues. In the Tao et al study, positive remodeling of ipsilateral arterial plaque sites was shown to be the best independent predictor of ESUS. Thus, intracranial arteries that have enlarged because of the presence of plaque and have a larger RI seem to be the most likely sites of plaque-derived emboli. However, patients with negative remodeling were excluded because they were more likely to be stenotic. Plaque burden was less predictive of ESUS, probably because it can be overestimated thanks to curved vessels and oblique cuts. The authors interpreted images without any clinical knowledge about the patients. This, along with the cutting-edge imaging and sophisticated statistical analysis, made this a compelling investigation.

Still, there were weaknesses. It was retrospective, with a small number of patients, all of whom were Chinese. There were no histologic data to support the analysis of plaque characteristics. Also, there are multiple possible sources of emboli; excluding other causes is imperfect in ESUS. However, one could argue too much attention is paid to patent foramen ovale and occult atrial fibrillation, both of which are unusual causes of ischemic strokes. Finally, the long-suspected-but-not-proven theory that cerebral vessel atherosclerosis and resultant embolization of debris from damaged plaques is getting its due. 


  1. Saver JL. Cryptogenic stroke. N Engl J Med 2016;374:2065-2074.
  2. Ameriso SF, Amarenco P, Pearce LA, et al. Intracranial and systemic atherosclerosis in the NAVIGATE ESUS trial: Recurrent stroke risk and response to antithrombotic therapy. J Stroke Cerebrovasc Dis 2020;29:104936.
  3. Perera KS, Ng KKH, Nayar S, et al. Association between low-dose rivaroxaban with or without aspirin and ischemic stroke subtypes: A secondary analysis of the COMPASS trial. JAMA Neurol 2020;77:43-48.