Improving Measurement Accuracy of the Degree of Carotid In-Stent Stenosis Using CT Angiography: A Comparison With DSA

Computed tomography angiography (CTA) tends to overestimate restenosis after cervical carotid stenting because of the substantial blooming artifacts from metallic stent materials. Hence, we propose a new measurement method to address this issue and assess the accuracy and reliability of evaluating the degree of in-stent stenosis (ISS) compared with digital subtraction angiography (DSA) as a post-hoc analysis of the CAS-CARE (Carotid Artery Stenting with the Cilostazol Addition for Restenosis) trial.

Among the image data registered for the trial, 197 paired data sets of poststenting images obtained using both CTA and DSA were used. We measured the ISS degree according to the NASCET (North American Symptomatic Carotid Endarterectomy Trial) criteria with the conventional and new methods using the inner edge and midportion of the stent materials as a boundary to the lumen and compared the stenotic severities with those obtained from DSA.

The ISS degrees measured using CTA with the conventional method were remarkably higher than those measured using CTA with the new method and DSA (median, 49.5% versus 11.3% and 9.8%; P<0.001). The agreements of the ISS degree between CTA and DSA were significantly improved when using the new measurement method (intraclass correlation coefficient = 0.67 [95% CI, 0.59-0.74]) when compared with the conventional method (0.24 [95% CI, -0.08 to 0.57]). In addition, there were no significant systemic biases against DSA in the new method, whereas there were significant biases (mean difference, 39.3%; P<0.001; slope, 0.14; P = 0.034) in the conventional method, although correlations with DSA were comparable between the 2 methods (r = 0.68 [95% CI, 0.59-0.75] and 0.65 [95% CI, 0.56-0.72], respectively).

Measurement using CTA with the midportion of the stent materials significantly improved the accuracy of the ISS degree, which was comparable to that using DSA.