Angiographic detection of carotid plaque ulceration. Comparison with surgical observations in a multicenter study. North American Symptomatic Carotid Endarterectomy Trial.

Abstract

Carotid plaque ulceration is used as one of the determinants in deciding which patients should be submitted to carotid endarterectomy. Uncertainties about its importance persist. Its detection by angiography is an important consideration. The detection of ulceration by angiography was compared with observations during endarterectomy in the first 500 patients recruited into the North American Symptomatic Carotid Endarterectomy Trial. This represents the first multicenter compilation of data on this subject and the largest series of patients with both arteriographic and direct surgical observation. Sensitivity and specificity of detecting ulcerated plaques were 45.9% and 74.1%, respectively. The positive predictive value of identifying an ulcer was 71.8%. These results remained unchanged with differing degrees of carotid stenosis and were confirmed by analyses based on receiver operating characteristic (ROC) methodology. The area under the ROC curve (Az) was estimated to be 0.61 (95% confidence interval, 0.55 to 0.67). These observations from a multicenter study confirm that little agreement exists between angiography and surgical observation in detecting carotid plaque ulceration.