Initial and follow-up high-resolution vessel wall MRI study of spontaneous cervicocranial artery dissection

Objectives To explore the factors associated with ischemic stroke secondary to spontaneous cervicocranial artery dissection (sCCAD) and evaluate the initial imaging markers related to outcomes. Methods Initial and follow-up high-resolution vessel wall MRI (VW-MRI) in consecutive patients with sCCAD were retrospectively analyzed. The associations of clinical and imaging factors and variants of the circle of Willis (COW) with ischemic stroke were evaluated using binary logistic regression analyses. The anatomical outcomes were categorized as complete, partial, and no remodeling based on changes of the vessel wall and lumen. Ordinal logistic regression analysis was used to assess associations between initial features and outcomes. Results A total of 115 dissected arteries (79 strokes, 36 non-strokes) were detected in 103 patients. Follow-up VW-MRI was available in 46 patients (44.7%, with 51 vessels), with a median interval of 8.5 months. Pseudoaneurysm (odd ratio [OR], 0.178; 95% confidence interval [CI], 0.039-0.810; p=0.026) tended to rarely cause ischemic stroke, while intraluminal thrombus (OR, 5.558; 95% CI, 1.739-17.765; p=0.004), incomplete COW (OR, 9.309; 95% CI, 2.122-40.840; p=0.003), and partial complete COW (OR, 4.463; 95% CI, 1.211-16.453; p=0.025) were independently associated with stroke occurrence. Furthermore, the presence of double lumen (OR, 5.749; 95% CI, 1.358-24.361; p=0.018) and occlusion (OR, 12.975; 95% CI, 3.022-55.645; p=0.001) were associated with no remodeling of sCCAD. Conclusions Multiple initial factors were found to be related to stroke occurrence and anatomical outcomes of sCCAD. High-resolution VW-MRI may provide valuable insights into the pathophysiology and evolution of sCCAD.