Objective: Hyperintensity of intracranial atherosclerotic plaque on T2 weighted vessel wall imaging (VWI) may indicate certain components such as fibrous caps, juxtaluminal thrombus, or intraplaque hemorrhage. We aim to explore if the plaque signal of middle cerebral artery (MCA) stenosis on T2 weighted VWI is associated with ischemic stroke.
Methods: Patients with recent stroke (<2 weeks) were reviewed in our institutional and prospectively-collected VWI database. Patients with cardioembolism, carotid artery stenosis, or non-atherosclerotic MCA stenosis were excluded. Culprit MCAs and contralateral asymptomatic MCAs that exceeded 50% stenosis were analyzed. Stenosis degree and remodeling ratio were measured on the maximal narrowing site. Plaque area was manually delineated on T2 weighted VWI. The plaque signal was normalized (divided by masseter or buccinator muscle signal, Fig1). The thresholds of normalized plaque signal (≥1.0 to 1.8 with an increment of 0.1) were tested. Logistic regression and Receiver operating characteristic (ROC) analysis were performed to determine the optimal signal threshold of predicting culprit MCA.
Results: Eighty-two MCAs were analyzed (from 66 patients, 51 male, mean age 58±16 years, mean stenosis degree 68±14%, mean remodeling ratio 1.10±0.21), including 60 culprit MCAs and 22 contralateral asymptomatic MCAs. Logistic regression analysis indicated the model of pixel number with normalized plaque signal ≥1.4 (OR 1.08 per 1-pixel increase, 95% CI 1.01-1.15) and remodeling ratio (OR 1.90 per 0.1 increase, 95% CI 1.25-2.89) was of the highest diagnostic value (AUC 0.861, 95% CI 0.779-0.943). Stenosis degree was not independently associated with culprit MCA. Other tested thresholds from 1.0 to 1.8 also had good diagnostic value with AUC 0.814-0.858.
Conclusions: The bright signals of MCA plaque on T2 weighted VWI are associated with ischemic stroke. Further study is required to clarify the underlying pathophysiology.