Cardiac sources of embolism in patients with pial artery infarcts and lacunar lesions.

Abstract

On the assumption that the majority of lacunes are caused by small-vessel diseases and that pial artery infarcts arise from cardio-embolic or large-vessel diseases, 194 patients from the Berlin Cerebral Ischemia Data Bank with either lacunar or pial artery infarcts were analyzed for the frequency of cardiac sources of embolism. The primary hypothesis was that the frequency of cardiac sources of embolism is higher among pial artery infarct subjects. The presence of cardiac sources of embolism was estimated by electrocardiographic and transthoracic and transesophageal echocardiographic studies. Cranial computed tomography scans were evaluated by two masked observers. The overall rate of cardiac sources of embolism did not differ significantly between the lacunar and the pial artery infarct group (66% versus 71%; odds ratio, 0.80; confidence interval, 0.43 to 1.50). Echocardiographic evidence of cardiac thrombi was positively associated with pial artery infarcts (odds ratio, 0.18; confidence interval, 0.04 to 0.80); atrial fibrillation and all other cardiac sources were not. Left cardiac thrombi are significantly associated with pial artery infarcts. Other presumed cardiac sources of embolism, including atrial fibrillation, may often represent coincidental findings or have a less strong tendency to result in pial artery infarcts.