Characteristics of culprit intracranial plaque without substantial stenosis in ischemic stroke using three-dimensional high-resolution vessel wall magnetic resonance imaging
Xia Tian,Zhang Shi,Zhen Wang,Bing Xu,Wen-Jia Peng,Xue-Feng Zhang,Qi Liu,Shi-Yue Chen,Bing Tian,Jian-Ping Lu,Cheng-Wei Shao
DOI: https://doi.org/10.3389/fnins.2023.1160018
IF: 4.3
2023-03-24
Frontiers in Neuroscience
Abstract:Background and aims: We aim to analyze the difference in quantitative features between culprit and non-culprit intracranial plaque without substantial stenosis using three-dimensional high-resolution vessel wall MRI (3D hr-vw-MRI). Methods: The patients with cerebral ischemic symptoms of the unilateral anterior circulation were recruited who had non-stenotic intracranial atherosclerosis (<50%) confirmed by computed tomographic angiographic (CTA) or magnetic resonance angiography (MRA). All patients underwent 3D hr-vw MRI within 1 month after symptom onset. 3D hr-vw-MRI characteristics, including wall thickness, plaque burden, enhancement ratio, plaque volume and intraplaque hemorrhage, and histogram features were analyzed based on T2-, precontrast T1-, and post-contrast T1-weighted images. Univariate and multivariate logistic regression analysis were used to identify key determinates differentiating culprit and non-culprit plaques and to calculate the odds ratios (ORs) with 95% confidence intervals (CIs). Results: A total of 150 plaques were identified, of which 133 plaques (97 culprit and 36 non-culprit) were in the middle cerebral artery, three plaques (all culprit) were in the anterior cerebral artery (ACA) and 14 (11 culprit and three non-culprit) were in the internal carotid artery (ICA). Of all the quantitative parameters analyzed, plaque volume, maximum wall thickness, minimum wall thickness, plaque burden, enhancement ratio, coefficient of variation of the most stenotic site, enhancement ratio of whole culprit plaque in culprit plaques were significantly higher than those in non-culprit plaques. Multivariate logistic regression analysis found that plaque volume [OR, 1.527 (95% CI, 1.231–1.894); P < 0.001] and enhancement ratio of whole plaque [OR, 1.095 (95% CI, 1.021–1.175); P = 0.011] were significantly associated with culprit plaque. The combination of the two features obtained a better diagnostic efficacy for culprit plaque with sensitivity and specificity (0.910 and 0.897, respectively) than each of the two parameters alone. Conclusion: 3D hr-vw MRI features of intracranial atherosclerotic plaques provided potential values over prediction of ischemic stroke patients with non-stenotic arteries. The plaque volume and enhancement ratio of whole plaque of stenosis site were found to be effective predictive parameters.
neurosciences