Publication:

Association between spontaneous internal carotid artery dissection and perivascular adipose tissue attenuation on computed tomography angiography

Background: Spontaneous cervical artery dissection (sCAD) is a leading cause of ischemic stroke in young patients. Studies using high-resolution magnetic resonance imaging and positron emission tomography have suggested vessel wall inflammation to be a pathogenic factor in sCAD. Computed tomography (CT) attenuation of perivascular adipose tissue (PVAT) is an established non-invasive imaging biomarker of inflammation in coronary arteries, with higher attenuation values reflecting a greater degree of vascular inflammation. Objectives: We evaluate the CT attenuation of PVAT surrounding the internal carotid artery (PVATcarotid) with and without spontaneous dissection. Methods: Single-centre prospective observational study of 56 consecutive patients with CT-verified spontaneous dissection of the internal carotid artery (ICA) admitted between 2011 and 2018. Of these patients, 6 underwent follow-up CTA. 22 patients who underwent CTA for acute neurological symptoms but did not have dissection formed the control group. Using semiautomated research software, PVATcarotid was measured as the mean Hounsfield Unit (HU) attenuation of adipose tissue within a defined volume of interest surrounding the ICA. Results: PVATcarotid was significantly higher around dissected ICA compared with non-dissected contralateral ICA in the same patients (-58.7±10.2 vs. -68.9±8.1 HU, P<0.0001) and ICA of patients without dissection (-58.7±10.2 vs. -69.3±9.3 HU, P<0.0001). After a median follow-up of 89 days, there was a significant reduction in PVATcarotid around dissected ICA (from -57.5±13.4 to -74.3±10.5 HU, P<0.05); while no change was observed around non-dissected contralateral ICA (from -71.0±4.4 to -74.1±4.1 HU, P=0.19). ICA dissection was an independent predictor of PVATcarotid following multivariable adjustment for age and the presence of ICA occlusion.