Abstract TP160: The Non-stenosing Carotid Artery Plaque In Embolic Stroke Of Undetermined Source: A Retrospective Study

Introduction: Atherosclerotic cervical ICA disease is one of the major causes of ischemic stroke. The risk of stroke from mild to moderate stenoses (i.e., <50% stenosis) might be underestimated. Further investigation is mandated to describe the association between high-risk plaque features and ESUS. Methods: This was a retrospective observational study. Using the stroke registry of our hospital's system between June 20th 2016 and June 20th 2021, we reviewed data for patients diagnosed with ESUS according to previously published definition criteria. Using CTA, we analyzed laterality of high-risk plaque features in relation to the stroke side, and then we identified the incidence of recurrent stroke events. Results: Out of 1779 patients with cryptogenic ischemic stroke, only 152 met the inclusion criteria for ESUS. We Compared high-risk plaque features ipsilateral to stroke side as to contralaterally. There were significantly more ulcerations defined as >1 mm depression (19.08% vs 5.26%, p<.0001), plaque thickness >3 mm (19.08% vs 7.24%, p=0.001), and plaque length >1cm (13.16% vs 5.92%, p=0.0218). Also, there was a significant difference in stenosis of ipsilateral to stroke when compared contralaterally, especially for stenoses of 10-30% and 31-49% (17.76% vs 10.53% and 5.26% vs 2.63%, respectively. p=0.0327). There was also a significant difference in plaque component; both components (soft and calcified) and only soft plaque (42.76% vs 23.68% and 17.76% vs 9.21%, respectively. p<.0001) were more prevalent ipsilaterally. In total, 17 patients were found to have a recurrent stroke event, 8 patients had an ipsilateral stroke to the index event, 7 had a bilateral and 2 had a contralateral event. Conclusion: ESUS is more commonly found ipsilateral to high-risk plaque features. The small number of our sample is definitely a limitation. Further large and multicenter studies aiming to form precise prediction models and scoring systems are needed to help guide treatment.