Brainstem cavernomas: long-term results of microsurgical resection in 52 patients

To review the natural history and the long-term results of microsurgical resection of brainstem cavernous angiomas operated on in one institution.

A retrospective analysis was conducted of the preoperative and postoperative course in 52 consecutive patients who underwent microsurgical resection of a brainstem cavernoma between 1990 and 2002. The role of sex, age, cavernoma location, size, multiple bleedings, relationships to the pial-ependymal surface, surgical approach, and preoperative magnetic resonance imaging appearance were evaluated as prognostic factors possibly influencing outcome. Discrete data were compared by use of the chi2 test and Fisher's exact test as appropriate.

The risk of hemorrhage was 3.8% per patient per year. The rebleeding rate was 34.7%. Nineteen of 29 patients who experienced new neurological deficits after surgery improved over time to their preoperative condition or better. Permanent morbidity was observed in 10 (19%) of 52 patients (follow-up: 1.5-10.5 yr; mean, 4.7 yr; median, 4.3 yr; standard deviation, 0.2 yr). The final Karnofsky Performance Scale score for these 10 patients was 90 in 2 patients, 80 in 2, 70 in 2, 60 in 2, 50 in 1, and 30 in 1. The mortality rate was 1.9%. The incidence of permanent new neurological deficits was lower in the 20 patients whose lesion could be removed through an anterolateral pontine approach (5 versus 29%; P = 0.035).

Surgical resection is recommended for superficial lesions and for lesions that can be reached through the anterolateral pontine surface. Surgery is also recommended for symptomatic cavernomas with a satellite subacute hematoma.