Association of CTA-derived carotid plaque imaging features with 1-year ischemic stroke risk: An exploratory analysis

Objective This investigation was designed to explore the relationship between the imaging characteristics of carotid plaques and the likelihood of experiencing ischemic stroke during 1-year follow-up, utilizing computed tomography angiography as the primary imaging modality. Methods Computed tomography angiography (CTA) was employed to acquire cervical arterial images. Plaques were classified by computed tomography (CT) attenuation into a single prespecified four-category scheme: lipid-rich (<60 Hounsfield units [HU]), fibrous (60–130 HU), calcified (>130 HU), and mixed (containing two or more components). Surface morphology was assessed separately as regular or irregular. These attenuation-defined categories are semi-quantitative imaging phenotypes and are not histopathologically validated. To evaluate the associations between plaque composition, morphological features, volumetric measurements, the extent of luminal narrowing, and the subsequent risk of stroke during 1-year follow-up, multivariate logistic regression modeling was applied. The primary endpoint for the primary analysis was recurrent ischemic stroke in the observation group; first-ever ischemic stroke in the control group was described separately. Results A total of 200 participants were analyzed, including 100 with baseline ischemic stroke in the observation group and 100 without baseline ischemic stroke in control group. During 1-year follow-up, 29 patients in the observation group had recurrent ischemic stroke, and 5 participants in the control group had first-ever ischemic stroke. The primary analysis was restricted to recurrent events in the observation group; recurrent and incident events were not pooled. Severe stenosis or complete occlusion of the carotid artery was observed predominantly in the observation group, whereas only two instances of mild narrowing were identified in the control group. A total of 118 carotid plaques were detected: 103 in the observation group and 15 in the control group. Using the single prespecified four-category scheme, the 118 plaques comprised 27 lipid-rich plaques (22.88%), 54 fibrous plaques (45.76%), 19 calcified plaques (16.10%), and 18 mixed plaques (15.25%). In the observation group, the 103 plaques were classified as lipid-rich (n = 26), fibrous (n = 46), calcified (n = 15), and mixed (n = 16). In the control group, the 15 plaques were classified as lipid-rich (n = 1), fibrous (n = 8), calcified (n = 4), and mixed (n = 2). The marked intergroup disparity was observed in the prevalence of carotid stenosis, which reached statistical significance ( P < 0.05). Adjusted associations were observed for irregular morphology (OR 3.648, 95% CI 1.049–12.69), plaque volume (OR 2.832, 95% CI 1.282–6.261), and lipid-rich plaque (OR 1.453, 95% CI 1.164–1.814). Irregular morphology, greater plaque volume, and lipid-rich composition were associated with recurrent ischemic stroke in the observation group during 1-year follow-up. Specifically, increases in both plaque volume and lipid content were associated with a heightened risk of recurrent ischemic stroke in the observation group. Conclusion CTA provides clear visualization of key carotid plaque features. Lipid-rich plaques, irregular surface contours, and larger plaque volumes were associated with an increased risk of recurrent ischemic stroke in the observation group during the 1-year follow-up. CTA-derived plaque morphology, volume, and lipid-rich composition may be considered candidate markers for 1-year ischemic stroke risk stratification, pending prospective external validation.