Deep and brainstem cavernomas: a consecutive 8-year series.

OBJECT: The goal of this study was to provide epidemiological and clinical data on the management of cavernomas of the basal ganglia and brainstem from a long-term series at one institution.
METHODS: All 68 patients who were referred to the authors' department between 1992 and 2000 for deep cavernomas were evaluated by clinic examinations, review of neuroimaging examinations, and review of charts and operative notes. Twenty-nine patients underwent microsurgical procedures, which carried a 69% risk of transitory neurological deterioration. Radical excision was achieved in 25 of these patients, as determined by a review of neuroimages; the remaining four patients all experienced new hemorrhages that led to increased morbidity or even to mortality. Surgical results were better if surgery was performed early, within 1 month posthemorrhage, than if operations were postponed. In selected patients, deep lesions not reaching a pial surface could be safely removed from the thalamus, basal ganglia, or medulla oblongata. Of five patients who underwent gamma knife surgery, two experienced hemorrhages, one at 2 and the other at 5 years following treatment. Patients who did not undergo surgery had a yearly incidence of hemorrhage that was 2% in cases of incidental cavernomas and 7% in symptomatic ones.
CONCLUSIONS: Over the long term, outcomes were worse following conservative treatment or shunt insertion surgery than after microsurgery of symptomatic cavernomas. Incidental cavernomas carried a low risk of neurological deterioration. Surgery should follow generally accepted indications, but only with the confidence that total removal can be safely achieved. Surgery that is performed within 10 to 30 days following ictus may be preferable to delayed surgery.