Management of spontaneous cerebellar hematomas: a prospective treatment protocol

To identify easily applicable guidelines for the surgical and conservative management of spontaneous cerebellar hematomas.

A treatment protocol was developed and prospectively applied for the management of 50 consecutive cases of cerebellar hematomas. The appearance of the fourth ventricle, adjacent to the hematoma, on computed tomographic scans was divided into three grades (normal, compressed, or completely effaced). The degree of fourth ventricular compression was correlated with the size and volume of the hematoma and the presenting Glasgow Coma Scale (GCS) score. The hematoma was surgically evacuated for all patients with Grade III compression and for patients with Grade II compression when the GCS score deteriorated in the absence of untreated hydrocephalus. Patients with Grade I or II compression were initially treated with only ventricular drainage in the presence of hydrocephalus and clinical deterioration.

The degree of fourth ventricular compression was classified as Grade I in 6 cases, Grade II in 26, and Grade III in 18. The degree of fourth ventricular compression was significantly correlated with the volume of the hematoma (r(s) = 0.67, P < 0.0001), hydrocephalus (r(s) = 0.44, P = 0.001), the preoperative GCS score (r(s) = 0.43, P = 0.001), the maximal diameter of the hematoma (r(s) = 0.43, P = 0.001), and a midline location of the hematoma (chi(2) = 6.84, P < 0.009). Acute deterioration in GCS scores occurred for 6 (43%) of 14 patients with Grade III ventricular compression who were conscious at presentation. Thirteen patients with Grade I or II ventricular compression and stable GCS scores of more than 13 were treated conservatively. Nine patients were treated with ventricular drainage only, and 28 underwent posterior fossa craniectomy and evacuation of the hematoma with ventricular drainage. The mortality rate at 3 months was 40%. None of the patients with Grade III fourth ventricular compression and GCS scores of less than 8 at the time of treatment experienced good outcomes. Overall, 15 (60%) of 25 patients with hematomas with maximal diameters of more than 3 cm and Grade I or II compression did not require clot evacuation.

Conscious patients with Grade III fourth ventricular compression should undergo urgent clot evacuation before deterioration. Surgical evacuation of the clot may not be required for large hematomas (>3 cm) if the fourth ventricle is not totally obliterated at the level of the clot.