Doppler emboli signals vary according to stroke subtype.

BACKGROUND AND PURPOSE Doppler ultrasound detection of emboli signals may assist in distinguishing embolic from thrombotic stroke. Selected patient groups have a high incidence of such signals. We have examined consecutive stroke cases to identify the incidence of Doppler emboli in different etiologic subtypes of stroke. METHODS Forty-five patients presenting with first-ever acute carotid territory cerebral ischemia were studied prospectively. Transcranial Doppler examination of both middle cerebral arteries, carotid color duplex ultrasound, and transthoracic or transesophageal echocardiography were completed within 48 hours of deficit onset. Clinical and imaging data were interpreted independent of emboli data, and stroke etiology was classified according to recent multicenter trial criteria. RESULTS Middle cerebral artery signals were identified in at least one cerebral hemisphere in 41 of the 45 patients. Emboli signals were present in 29 of these 41 cases (71%). These signals were bilateral in 22, within the affected (symptomatic) cerebral hemisphere only in 5, and contralateral only in 2 cases. No emboli signals were detected in any of 8 patients with lacunar stroke. The overall difference in emboli signal counts between etiologic subgroups was significant (P = .001, Kruskal-Wallis). A significantly higher emboli signal count was found within affected cerebral hemispheres than contralaterally in the 8 patients with large artery atherosclerosis (11.3 versus 1 signals per hour, median [95% confidence interval, 3 to 40 and 0 to 3, respectively], P = .02), but this interhemisphere difference was not present for other etiologic subgroups. CONCLUSIONS Emboli signals are common in patients with acute stroke, with the notable exception of lacunar stroke. This is consistent with the small vessel etiology for the latter group and provides support for the relevance of Doppler emboli signal detection in thromboembolic cerebrovascular disease.