We assessed the utility of noninvasive methods for clinical decision-making in carotid stenosis. We compared the results from magnetic resonance angiography (MRA), carotid ultrasound (CUS), and conventional carotid angiography (CCA) for 39 arteries. Noninvasive studies were graded by a 5-point diameter scale. Conventional angiograms were evaluated for percent stenosis using the North American Symptomatic Carotid Endarterectomy Trial criteria and diameter of residual lumen by caliper. Agreement was 64% between MRA and CCA, 72% between CUS and CCA, and 77% between MRA and CUS. For vessels measuring greater than 60% stenosis by CCA, sensitivity was 100% for both MRA and CUS. Specificities were 65% (MRA) and 71% (CUS), reflecting the tendency for MRA and CUS to overestimate stenosis. MRA and CUS correlated with residual lumen, even in some cases where they disagreed with the percent stenosis. We conclude that patients with normal arteries or mild stenosis as determined by MRA and CUS may be excluded from further evaluation, whereas patients with highergrade lesions should be evaluated with CCA prior to consideration for surgery. Close correlation between the results of MRA and CUS, as well as their apparent correlation with residual lumen, suggests that MRA and CUS may more accurately reflect carotid flow than does calculated percent stenosis.
Copyright © 1995 National Stroke Association. Published by Elsevier Inc. All rights reserved.