Supracardiac Atherosclerosis in Embolic Stroke of Undetermined Source

The Underestimated Source

George Ntaios; Max Wintermark; Patrik Michel


Eur Heart J. 2021;42(18):1789-1796. 

In This Article

Carotid Plaques and High-risk Imaging Features

The term non-stenotic carotid plaque has been frequently used in the literature to describe patients with carotid plaques with mild stenosis,[7–11] although it is apparently a misnomer as some degree of stenosis was indeed present.

There is firm evidence showing that the higher the degree of carotid stenosis, the higher the thromboembolic risk is.[12,13] The TOAST classification, which is the most widely used aetiological classification of stroke, used an arbitrary threshold of 50% atherosclerotic stenosis to assign an aetiological role to a plaque in patients with ischaemic stroke, whereas patients with lower degree of stenosis were categorized as of undetermined cause.[1] Still, some patients with carotid plaques will have an ischaemic stroke due to their atherosclerotic lesion. In this context, the proportion of large artery atherosclerotic strokes seems to be underestimated.[14] Subsequent aetiological classifications of stroke tried to address this limitation.[15,16] In particular, in the ASCOD phenotyping of ischaemic stroke, the presence of an ipsilateral atherosclerotic stenosis <50% in an intra- or extracranial artery with an endoluminal thrombus supplying the ischaemic field is considered as potentially causal, whereas in the Causative Classification System, large artery atherosclerosis is considered as an evident cause when there is <50% diameter reduction with plaque ulceration or thrombosis in the clinically relevant extracranial or intracranial arteries.[14,15]

In addition, the availability of sophisticated imaging methods over the recent years allows the characterization of plaque vulnerability and its potential for higher embolic risk. This might be of high importance in patients with ESUS as it could imply causality. Intraplaque haemorrhage, lipid-rich necrotic core, plaque neovascularization, plaque thickness, plaque volume, and surface morphology including thin cap and ulcerations could be used towards this direction (Take-home figure).[17–21] Such features can be assessed by different non-invasive methods including magnetic resonance imaging (MRI), computed tomography (CT), and Doppler/Duplex imaging. With the latter method, detection of clinically silent emboli in patients with atherosclerotic carotid stenosis seems to be a reliable marker for stroke risk.[22] Magnetic resonance (MR)-based imaging characteristics of intracranial plaque vulnerability are now also emerging.[23]

Take home figure.

Prevalence of supracardiac atherosclerosis in patients with embolic stroke of undetermined source and imaging modalities for its assessment. The prevalence of non-stenosing supracardiac atherosclerosis is based on previous reports for intracranial,24 extracranial carotid25 and vertebrobasilar,26 and aortic arch atherosclerosis.27 The left image depicts the arterial bed supplying the brain. The right image depicts computed tomography angiography (A) and MRI T1 fat sat (B) images showing a soft atherosclerotic plaque with speckles of calcifications in the right internal carotid artery, likely haemorrhagic (bright T1 signal on MRI). This plaque causes ~40% of stenosis according to the NASCET criteria. Because of its volume and its haemorrhagic component, this plaque is likely at higher risk of causing an ischaemic stroke despite the moderate degree of stenosis. (C) In vivo computed tomography angiography image of the internal carotid artery using an automated classification computer algorithm-derived overlay, which shows a plaque with a large haemorrhage (red) that corresponds to a VIb lesion according to the American Heart Association classification. (D) The matching ex vivo section from the histologic examination. (C and D) Republished from Saba et al. 18 with permission.