CT angiography of stented carotid arteries: comparison with Doppler ultrasonography.

PURPOSE: To determine whether computed tomographic angiography (CTA) is a feasible modality for assessing stented carotid arteries and whether in-stent restenosis based on CTA concurs with ultrasonography (US). METHODS: A retrospective review was conducted of 37 follow-up CTA and US images from 27 patients (23 men; median age 70 years, range 56-77) who received 34 nitinol carotid stents. CTA and US images were compared with respect to assessability and percent stenosis. Both visual estimation (>or=50% or not) and the NASCET method were used to determine percent stenosis in CTA images. For US, a determination of >or=50% stenosis was based on peak systolic velocity (>or=200 cm/s) and an internal carotid artery to common carotid artery ratio >or=2.5. Percent stenosis values by CTA were also compared to values (n=7, 21%) determined by catheter angiography. RESULTS: CTA and US images were "totally assessable" in 27 (73%) and 15 (41%), "totally non-assessable" in 0 (0%) and 3 (8%), and "partially assessable" in 10 (27%) and 19 (51%), respectively. Assessability of CTA images was equal to or better than that of US images in 33 (89%). The percent stenoses by CTA and US were comparable in 20 cases. CTA found >or=50% stenosis using the NASCET method in 4 of 20 stents; none of these showed >or=50% stenosis by visual estimation of CTA or by spectral Doppler US. Compared with catheter angiography, CTA overestimated percent stenosis from 34% to 66% (mean 53%). US confirmed 2 angiographically proven restenoses, but CTA identified only 1. CONCLUSION: CTA provides better image quality for stented carotid arteries than US, but it might be inferior to US in determining restenosis in assessable cases. Therefore, CTA is likely to be an alternative to US in cases of non-assessability. A large-scale study including more restenosis cases is warranted to reveal which modality is more reliable for diagnosis of restenosis.