PURPOSE: To determine the accuracy of gadolinium-enhanced three-dimens
ional magnetic resonance (MR) angiography in the evaluation of carotid
artery stenosis. MATERIALS AND METHODS: A prospective blinded compari
son of first-pass MR angiography with conventional digital subtraction
angiography (DSA) was performed in 21 patients suspected of having ca
rotid artery stenosis. MR angiography was performed on a 1.5-T magnet
with gradient overdrive equipment, with a coronal radio-frequency-spoi
led, three-dimensional, fast low-angle shot sequence after intravenous
injection of contrast material. The guidelines of the North America S
ymptomatic Carotid Endarterectomy Trial for measuring stenosis of the
internal carotid artery were applied on maximum intensity projection i
mages and conventional angiograms. RESULTS: Grading of stenoses at MR
angiography agreed with that at DSA in 92% of the 44 carotid arteries.
In the 18 carotid arteries with severe stenosis (70%-99%), agreement
was 94%. All internal carotid occlusions (n=7) or pseudo-occlusions (n
=3) were accurately detected with MR angiography. CONCLUSION: The accu
racy of gadolinium-enhanced MR angiography in characterizing the degre
e of carotid stenosis was high. Findings with MR angiography were accu
rate for differentiating between occlusion and pseudo-occlusion.