Abstract
OBJECTIVES: The indication for carotid endarterectomy (CEA) mainly relies on the degree of stenosis and neurological symptoms. Plaque vulnerability has been associated with stroke risk, but identification on single-energy computed tomography (CT) has yielded heterogeneous results and is not routinely applied to clinical diagnostics. Hence, we intended to analyse CEA specimens for vulnerability features using dual-source CT and correlate these features with the presence of preprocedural symptoms. METHODS: CT was performed on 187 carotid plaque specimens using ultra-high-resolution and dual-energy imaging on a dual-source scanner. Plaques were separated into calcified versus non-calcified volumes and analysed concerning HU-density, calcifications and volumetric dual-energy indices (DEIs). Comparative statistical analysis of plaque characteristics was performed with respect to the presence of neurological symptoms. RESULTS: The degree of stenosis of symptomatic and asymptomatic plaques was indifferent (69.2 ± 12.3% vs 66.3 ± 13.7%). The highest diagnostic accuracies were obtained by the % calcified volume (AUC 0.63 (0.54-0.71)), average whole plaque HU (AUC 0.71 (0.64-0.79)), profound calcification (AUC 0.74 (0.66-0.81)), calcification spots <1 mm (AUC 0.71 (0.63-0.79)) and spotty calcification (AUC 0.74 (0.66-0.82)). The diagnostic accuracy for symptomatic plaques was insignificant using average non-calcified plaque HU (AUC 0.59 (0.48-0.65)), but significant using average non-calcified plaque DEI (AUC 0.66 (0.58-0.74)). CONCLUSIONS: Symptomatic plaques were identified best by measuring density of the whole, calcified or non-calcified plaque and via spotty, profoundly localized and less dense calcification. A volumetric DEI identifies symptomatic plaques with non-calcified plaque characteristics more accurately than single-energy CT. Future clinical studies are necessary to confirm these findings in patients.