Abstract Background and aims To assess whether vessel wall MRI (vwMRI) high-risk-plaque features improve differentiation between large-artery-atherosclerotic (LAA) and cardioembolic (CE) stroke beyond CTA, to improve embolic stroke of undetermined source (ESUS) reclassification. Methods We retrospectively reviewed consecutive acute ischemic stroke patients (Jun’22–Apr’24). Patients undergoing head/neck CTA and intra/extracranial vwMRI with single-territory infarcts classified as LAA or CE by TOAST criteria were included. Three radiologists assessed imaging markers, including non-calcified stenosis30% on CTA and plaque enhancement and intraplaque hemorrhage (IPH) on vwMRI. Clinical variables were collected-table1-. Associations between imaging features and stroke mechanism were evaluated using logistic regression, with model discrimination assessed by the area under the curve (AUC). Results Among 210 patients (median age 69years; 60%male), 104 had LAA and 106 CE strokes. LAA had a higher percentage of upstream segments with non-calcified stenosis 30% on CTA than CE (OR=3.0/SD;p0.001). On vwMRI, LAA was associated with a greater percentage of upstream enhancing intracranial plaques (OR=2.3/SD;p0.001) and a higher likelihood of upstream extracranial IPH (OR=4.4;p=0.005). In a CTA+vwMRI multivariable model, upstream intracranial plaque enhancement burden (aOR=2.4/SD;p0.001) and upstream extracranial IPH (aOR=6.3;p=0.001) remained significantly associated with LAA. The AUC of the CTA+vwMRI model was improved compared to CTA-based (AUC=0.83vs0.63,p0.001)-Figure1-. Conclusions Upstream atherosclerotic markers assessed on CTA and vwMRI, absent from current diagnostic criteria, were associated with LAA vs CE, with vwMRI offering incremental diagnostic value beyond CTA alone, suggesting potential utility for future multimodal models for ESUS reclassification. Conflict of interest Cristina Sanchez-Vizcaino: nothing to disclose Table 1 - belongs to Methods Figure 1 - belongs to Results
Building similarity graph...
Analyzing shared references across papers
Loading...
Cristina Sánchez-Vizcaíno
University of Washington
Daniel Hippe
Cape Town HVTN Immunology Laboratory / Hutchinson Centre Research Institute of South Africa
Ahmed Safwat
University of Washington
European Stroke Journal
University of Washington
Fred Hutch Cancer Center
Seattle University
Building similarity graph...
Analyzing shared references across papers
Loading...
Sánchez-Vizcaíno et al. (Fri,) studied this question.
synapsesocial.com/papers/69fd7e00bfa21ec5bbf0629f — DOI: https://doi.org/10.1093/esj/aakag023.702