Cerebral infarction: shortcomings of angiography in the evaluation of intracranial cerebrovascular disease in 25 cases.

We studied the utility and limitations of conventional cerebral angiography in 25 patients with cerebral infarction unassociated with extracranial cerebrovascular disease during a 7-year period. In only one-third of cases was the angiogram diagnostic, and in a single case it altered the pre-angiogram diagnosis by revealing a previously unsuspected embolus. Among the cases clinically diagnosed as cerebral emboli, the 2 confirmatory angiograms were performed early (within 48 hours), and demonstrated medium-large or large vessel filling defects. Two-thirds of the negative angiograms in the embolic clinical category were delayed, but there was no statistically relevant predilection for specific vessel size involvement. The category, primary cerebral vasculopathy, comprised the largest group, 10 in all, and one-half had angiographic confirmation despite time delays. Angiographic recognition was dependent on a characteristic picture of vascular involvement, and not on timing or vessel size predilection. Mitral valve prolapse figured prominently in the clinical cases of vasculopathy of uncertain etiology, which contained a total of 4 cases. The 3 cases with nondiagnostic angiograms were all delayed and demonstrated nonspecific radiographic changes. Clinically, these cases demonstrated signs or symptoms of autoimmune dysfunction, raising the specter of primary cerebral vasculopathy as a cause of cerebral infarction, in contrast to recurrent cerebral emboli.