Inaccuracy of angiography in the diagnosis of carotid ulceration.

Abstract

It is generally stated that ulceration of the carotid bifurcation plaque may give rise to embolization and symptoms of cerebral ischemia. It has been suggested that prophylactic carotid endarterectomy is indicated in asymptomatic patients if the angiogram is interpreted as showing ulceration. We therefore determined the accuracy of the radiologic diagnosis of ulceration by comparing the radiologic and morphologic findings in 155 carotid endarterectomies. Ulceration was diagnosed in 54% of the angiograms and 47% of the surgical specimens and was not related to the degree of stenosis. Angiography had a sensitivity of 73%, a specificity of 62% and an overall accuracy of 67%. For the non-flow-restrictive lesions these figures were 65%, 66%, and 66%. If surgery were based on the angiographic diagnosis of ulceration in plaques with less than 50% stenosis, 16 out of 42 operations (38%) would have been done unnecessarily. Variability of the angiographic diagnosis of ulceration is also shown by a high interobserver variability; two sets of two authors read the angiograms independently and disagreed in 24% of the cases. The decision to operate should not depend entirely upon the angiographic diagnosis of ulceration.