The risk of stroke with occlusion of the internal carotid artery

Reports of all cervicocephalic arteriograms (n = 1836) performed at one institution during a 10-year period were reviewed and the patients were classified into three groups according to the indication for arteriography. Group I included all patients with symptoms or findings compatible with occlusive disease of the carotid or vertebral artery (n = 806). Group II included patients with cerebrovascular symptoms unrelated to carotid or vertebral disease (e.g., patients with subarachnoid hemorrhage) (n = 367). Group III consisted of patients with no evidence of cerebrovascular disease (e.g., patients with primary and metastatic brain tumors) (n = 663). One hundred ten atherosclerotic occlusions of the internal carotid artery (ICA) were found in 106 patients in group I. Fifty-one percent of these patients had a history of stroke before arteriography, 24% had transient ischemic attacks (TIAs) or amaurosis fugax (AF), and 12% had nonhemispheric symptoms. Only 13% (1.7% of group I patients) were without symptoms. Ninety-one percent of the strokes and 75% of the TIAs or AF were ipsilateral to the ICA occlusion. Seventy-six percent of patients with stroke and 80% with ipsilateral TIAs or AF vs only 29% of patients without symptoms had contralateral stenosis of 60% diameter reduction or greater (p less than 0.003). No occlusions of the ICA occurred in groups II or III. Three hundred forty-six patients in groups II and III were more than 60 years of age. Assuming either Poisson or binomial distributions, the incidence of silent ICA occlusion in the population at large older than 60 years was estimated at less than 1% (p less than 0.03).