Fifty-eight children who underwent anatomical, functional, or modified anatomical hemispherectomy for intractable seizures from 1986 to 1995 were evaluated for seizure control, motor function, and complications. Age at surgery ranged from 0.3 to 17.3 years (median 2.8 years). Twenty-seven anatomical, 27 functional, and 4 modified anatomical hemispherectomies were performed. Seizure control and motor function in the 50 patients with more than 1 year follow-up revealed a 90% or better reduction in seizure frequency in 44/50 (88%) overall: 19/22 (86%) anatomical, 23/26 (89%) functional, and 2/2 modified anatomical. Motor function of the preoperatively hemiparetic extremities was improved or unchanged postoperatively in 38/50 (76%) of the patients. Complications included one intraoperative death, one late death from shunt obstruction managed elsewhere, late postoperative seizure breakthrough requiring reoperation and further disconnection in 5/27 functional hemispherectomy patients, mild cerebrospinal fluid infections in 3/27 anatomical hemispherectomy patients, and hydrocephalus requiring shunting in 3/27 functional hemispherectomy patients. A review of the literature and comparison of techniques is presented.

The metabolic changes that occur in the neonatal brain as a result of hydrocephalus, and the response to ventriculoperitoneal shunting, vary with the maturational stage of the brain. In this study, local glucose utilization (LCMRglu) and oxidative metabolic capacity were estimated using 2-deoxyglucose autoradiography and cytochrome oxidase histochemistry, respectively. Hydrocephalus was induced in rabbit pups via intracisternal kaolin injections at 4-6 days of age. Shunting occurred at 19–26 days of age and the animals were sacrificed at ages ranging from 33 to 331 days. In normal animals there was a high glucose demand early in life which showed a decrease at about 60 days of age. In rabbits sacrificed prior to 60 days of age the controls showed the highest LCMRglu with significant decreases in both the hydrocephalic and shunted animals. After 60 days of age the shunted animals had higher LCMRglu than both the hydrocephalic and control subjects. Oxidative metabolic capacity peaked before 50 days of age in normal animals. At the youngest age, both the hydrocephalic and shunted animals showed higher cytochrome oxidase density rates than the control rabbits. In the older group, the hydrocephalic animals remained high while the shunted animals approximated the control densities. Neither the changes seen in the LCMRglu nor the oxidative metabolic capacity were correlated with changes in cell packing density or increased intracranial pressure. These data suggest that when the brain is compromised by hydrocephalus, there is an initial compensatory increase in oxidative metabolic capacity. The development of the glycolytic pathway appears to be retarded by hydrocephalus, but with shunting and the passage of time, the LCMRglu rebounds to levels above that of controls.

No abstract