Vessel Wall Imaging Features of Spontaneous Intracranial Carotid Artery Dissection.

ABSTRACT

BACKGROUND AND OBJECTIVES: Intracranial dissection is an important cause of stroke often with nonspecific angiographic features. Vessel wall imaging (VWI) can detect dissections, but intracranial applications remain unvalidated by pathologic specimens. We sought to determine the ability of VWI to identify the rarely reported spontaneous intracranial carotid dissection (sICD) guided by postmortem validation. METHODS: VWI features of sICD, validated by postmortem specimen analysis in 1 patient, included luminal enhancement within a hypoenhancing outer wall, narrowing the mid to distal ophthalmic (C6) segment, relatively sparing the communicating (C7) segment. VWI examinations were reviewed to identify patients (1) with matching imaging features, (2) no evidence of other vasculopathies (i.e., inflammatory, intracranial atherosclerotic disease [ICAD]), and (3) adequate image quality. These sICD VWI features were compared with those in patients with known ICAD causing similar narrowing of C6 and relative sparing of C7 by a Fisher exact test accounting for multiple samples. RESULTS: Among 407 VWI examinations, 8 patients were identified with 14 sICDs, all women aged 30-56 years, 6 (75%) bilateral. All patients with sICD had risk factors of dissection (e.g., recently postpartum, fibromuscular dysplasia, and hypertension) and 3 (37.5%) had intracranial dissections elsewhere. Seven (87.5%) were diagnosed as moyamoya syndrome on initial angiography. Enhancing lesions varied from thin flap-like defects (n = 6) to thick tissue along the superolateral wall of the internal carotid artery, within the hypoenhancing outer wall. Compared with 10 intracranial carotid plaques in 8 patients with ICAD, sICD demonstrated stronger (84.6% vs 20.0%, p = 0.003-0.025) and more homogeneous (61.5% vs 0.0%, p = 0.005-0.069) enhancement and less positive remodeling (0.0% vs 60.0%, p = 0.004-0.09). T1 hyperintensity was identified in 5 sICDs in 3 patients but not identified in ICAD. Three patients with serial imaging (8- to 39.8-month maximum intervals) revealed little to no changes in stenosis, wall thickening, or enhancement. DISCUSSION: sICD is distinguishable on VWI from ICAD by enhancement characteristics, less positive remodeling, and clinical parameters. These VWI features should raise suspicion especially in young women with risk factors of dissection. Temporal stability and a lack of T1 hyperintensity should not discourage diagnosing sICD.