CT angiography for the detection and characterization of carotid artery bifurcation disease

Abstract

Background and purpose: Computed tomographic angiography (CTA) is a relatively new and minimally invasive method of imaging intracranial and extracranial blood vessels. The main purpose of this study was to compare CTA to the current gold standard of arterial imaging, digital subtraction angiography (DSA), for the detection and quantification of carotid artery bifurcation stenosis. We also compared Doppler ultrasound (US) with these 2 techniques.

Methods: In a prospective study, 40 patients (80 carotid arteries) underwent CTA, US, and DSA. Patients chosen for inclusion were symptomatic with TIAs or stroke and had initial US screening that indicated >50% carotid stenosis on the side appropriate for the symptoms. Source axial, maximum intensity projection (MIP), and shaded-surface display (SSD) images were produced for each CTA study. The US, CTA, and DSA images were reviewed, with the degree of stenosis quantified and presence of ulcers determined; each type of imaging was reviewed by a separate investigator blinded to the results of the other 2 modalities. The results of CTA and US imaging were compared with the DSA images for degrees of carotid stenosis.

Results: CTA source axial images correlated with DSA more closely than MIP or SSD images for all degrees of stenosis. The correlation between US and DSA (0.808) was poorer than that between CTA and DSA (0.892 to 0.922). CTA performed well in the detection of mild (0% to 29%) carotid stenosis, as well as carotid occlusion, with values for sensitivity, specificity, and accuracy near 100%. In determining that a stenosis was >50% by DSA measurement, CTA was again useful, with a sensitivity, specificity, and accuracy of 89%, 91%, and 90%, respectively. While CTA was quite specific and accurate in identifying degrees of stenoses in either the 50% to 69% or the 70% to 99% ranges, in this task it was much less sensitive: 65% for 50%-69% stenosis and 73% for 70%-99% stenosis. These results did not change significantly when only the data from the most clinically relevant symptomatic arteries were analyzed. CTA was found to correlate quite well with DSA in the detection of ulcers associated with the carotid stenosis.

Conclusions: CTA was found to be an excellent examination for the detection of carotid occlusion and categorization of stenosis in either the 0%-29% or >50% ranges. However, CTA was unable to reliably distinguish between moderate (50%-69%) and severe (70%-99%) stenosis, which is an important limitation in the investigation and treatment of carotid stenosis.