[Infratentorial hemorrhage following supratentorial surgery].

Hemorrhage in regions remote from the site of initial intracranial operations is rare, but does occur. We report three cases of cerebellar hemorrhage that developed after supratentorial surgery, all of which had similar clinical findings and CT images. The first case was a 37-year-old man with a craniopharyngioma in the suprasellar lesion. Partial removal of the tumor was performed through frontal craniotomy and the translaminaterminals approach. A large quantity of cerebospinal fluid (CSF) was suctioned from the third ventricle during the operation, resulting in marked brain shrinkage. The second and third cases were 34- and 51-year-old women with unruptured right middle cerebral aneurysms. Clipping of the aneurysms through the pterional approach was performed in both cases. In the second case, CSF was suctioned in large quantity from the carotid and prechiasmal cistern at the operation, resulting in marked brain shrinkage. In the third case, however, only a small volume of CSF was suctioned from the carotid and prechiasmal cistern during the operation, and no marked brain shrinkage was observed. CT scan showed that the hematomas were located mainly in the subdural or the subarachnoid spaces over the cerebellar hemisphere and partially extending into the cerebellar cortex. The mechanism of cerebellar hemorrhage in these series of patients was thought to be multifactorial. The possible etiology for cerebellar hemorrhage in the three cases presented was examined, including the role of CSF suction during surgery and disturbance of venous circulation in the posterior fossa. Suction of the CSF may cause intracranial hypotension. Further reduction of intracranial pressure leads to an increased transluminal venous pressure. There was no episode of hypertension or disturbed blood coagulation during or after the operation. The preoperative angiogram also revealed no abnormality at the region of the posterior fossa. Neuroimaging of infratentorial hemorrhage after supratentorial craniotomy is obviously different from that of hypertensive cerebellar hemorrhage. From the shape or extension of the hemorrhage, the main vessels of hemorrhage are the superior vermian vein and their tributaries damaged by stretching and tearing of these vessels. These vessels are not demonstrable in the angiogram, therefore there is no evidence for this hypothesis and the etiology is still unclear. There is no doubt, however, that there was a disturbance of venous circulation in this complication. We would like to emphasize the possibility of this complication. Patients who show signs of difficulty in awaking from anesthesia or the development of new neurological deficits not attributed to direct operative procedure after supratentorial craniotomy must be evaluated early, with adequate imaging including the posterior fossa.