Abstracts

Seizure control in patients undergoing hemispherectomy varies from 50% to 80%. A critical analysis of factors associated with postoperative seizure control has never been addressed. Our objective is to identify possible factors associated with recurrent seizures after hemispherectomy.
Forty-nine patients underwent hemispherectomy procedures from November 1996 through March 2002. A total of 55 procedures were performed.
Pre and postoperative evaluation included extensive video-EEG monitoring, MR imaging, PET and neuropsycological testing. Seizure outcome was correlated with possible factors associated with persistent postoperative seizures. Risk factors included surgical technique, incomplete disconnections evidenced in post-operative images, diagnosis, duration of epilepsy syndrome, bilateral ictal and interictal EEG activity, bilateral MRI findings, bilateral SPECT and PET abnormalities. Mean follow-up was 33.2 months (minimal follow-up of 6 months).
The Chi-test and Fisher test were applied as statistical analysis. P values bellow 0.05 were considered statistical significant.
Diagnosis included Malformation of Cortical Development (MCD) in 29 patients, peri-natal infarction (13 patients), Rasmussen[apos]s encephalitis (5 patients) and Sturge-Weber Syndrome (2 patients). Thirty-nine patients (79.6%) were seizure-free, 9 patients (18%) had remarkable improvement and 1 patient (2%) showed no improvement. Complication rate was 14% (postoperative hydrocephalus and infection).
In postoperative MRIs, 35 procedures (in 29 patients) resulted in incomplete disconnection (residual connected tissue with the exception of insular cortex) and 20 procedures (in 20 patients) in complete disconnections. The most commom site of incomplete disconnection was the basal posterior frontal area (33 procedures).
In the 20 procedures with complete disconnection, 18 procedures resulted in seizure control (90%). In the 29 patients with incomplete disconnection, 21 patients were seizure free (72%). Five patients from the incomplete disconnection group were re-operated and eventually 4 (80%) were seizure free. The statistical analysis showed a significant difference between those groups with a better seizure freedom outcome resulting from complete disconnection (p[lt]0.05). MCD was also associated with persistent seizures after surgery.
In 28 patients, insular infarction was demonstrated on postoperative MRI. The presence of insular cortex did not correlate with persistent seizures after surgery. In patients with hemimegaloencephaly, functional hemispherectomy technique was associated with persistent seizures after surgery.
Incomplete disconnection and MCD were statistically associated with persistent seizures after surgery. The overall rate of seizure freedom is encouraging (80%) with a satisfactory complication rate. Attention to technique during disconnection, in paticularly of the basal frontal lobe, is mandatory.