Detection of clinically silent intracranial emboli ipsilateral to internalcarotid occlusions during cerebral angiography

OBJECTIVE. Embolic ischemic events have long been suspected to occur in the cerebral arteries distal to an ipsilateral occluded internal carotid arter y (ICA). Documentation of microemboli by transcranial Doppler sonography du ring catheter angiography in patients with ICA occlusions provides objectiv e evidence of such distal emboli. SUBJECTS AND METHODS. Seven patients undergoing carotid angiography were ev aluated with transcranial Doppler sonography. Patients were also screened f or ICA occlusions using carotid duplex sonography. In the seven patients, w e saw five right ICA occlusions and two left ICA occlusions. Real-time visu al and auditory confirmations of emboli were obtained by recognizing their specific spectral signatures and harmonic qualities. Routes of collateral f low were determined from angiography. Specific phases of the examination we re correlated with embolic occurrences. RESULTS. Overall, emboli were seen during all phases of arteriography. In t he individual patients, emboli were identified in one to four of the eight angiographic phases we defined. Most emboli occurred during catheter flushi ng and contrast injection rather than during wire and catheter manipulation . The emboli were detected in the middle cerebral artery distribution ipsil ateral to the occluded ICA in all seven patients. Collateral flow patterns included, in four patients, external carotid artery-to-ICA collateral flow; in all seven patients, patent anterior communicating arteries; and in thre e patients, patent posterior communicating arteries. CONCLUSION. Emboli seen in middle cerebral arteries ipsilateral to occluded ICAs during cerebral angiography strongly indicate that emboli can occur d istal to an occlusion. Our findings support the thought that emboli arising from sources proximal to an occluded ICA may reach the hemisphere distal t o the occlusion, resulting in parenchymal ischemia or infarction.