Carotid Plaque Characterization Using Dual-Energy Computed Tomography: Predicting Imminent Ipsilateral Ischemic Stroke in 30 Days.

Background: Despite progress in noninvasive means for visualizing and characterizing carotid plaque morphology, current clinical practice and AHA/ASA guidelines do not employ any plaque characteristics in deciding whether a patient should undergo carotid endarterectomy. Our objective is to evaluate the utility of dual-energy computed tomography imaging characteristics of the carotid plaques to differentiate between the stable and unstable plaques and to predict subsequent ipsilateral ischemic stroke within 30 days. Methods: In this retrospective case-controlled study, 1185 consecutive neck computed tomography angiograms from November 2017 to October 2019 were included. In this cohort, 14 patients who experienced acute ischemic strokes caused by carotid plaque rupture within 30 days were identified and analyzed. Forty-two patients who did not suffer an ischemic stroke but had matched risk factors were selected as control. Univariable and multivariate logistic regressions were used to develop a predictive model. Through a similar retrospective review of 483 stroke patients from November 2019 to October 2020, an independent test data set of 60 patients (15:45) was identified and used to assess the predictive performance of the model. Results: In the univariable models, the plaque thickness (odds ratio [OR],1.59; 95% CI, 1.12-2.24; P=0.009), degree of stenosis (OR, 1.05; 95% CI, 1.02-1.09; P=0.002), plaque ulceration (OR, 20.00; 95% CI, 3.42-116.80; P=0.001), and intraplaque hemorrhage (OR, 7.22; 95% CI, 1.45-35.93; P=0.016) showed significant association with a subsequent stroke. Patent luminal diameter (OR, 0.62; 95% CI, 0.45-0.85; P=0.003) and maximal external diameter (OR, 0.70; 95% CI, 0.50-0.99; P=0.046) were inversely associated with a subsequent stroke. The final predictive model derived from multiple logistic regression included patent luminal diameter, maximal external diameter, degree of stenosis, plaque ulceration, and intraplaque hemorrhage. On the independent test data set, the model showed high specificity (84%) and negative predictive value (90%) with moderate sensitivity (73%) and positive predictive value (61%). The overall accuracy was 82%. Conclusions: Dual-energy computed tomography imaging features of carotid plaque may aid in prediction of a subsequent ipsilateral ischemic stroke within 30 days following the imaging study. Such information may guide the clinical management of vulnerable plaque.