Abstracts

RATIONALE: Structural abnormalities often render localization related epilepsy medically refractory. We reviewed the findings at surgery for medically refractory epilepsy in a modern surgical series accumulated at Massachusetts General Hospital (MGH) since 1993 and compared them to historical series accumulated before the MRI era.
METHODS: Records of 122 consecutive patients who underwent surgery for epilepsy from 1993 were reviewed. Clinical and presurgical data were tabulated. Histopathological impression on review of the biopsy slides was utilized as the gold standard for diagnosis of structural abnormality. Data were correlated with post surgical clinical outcome (Engel classification).
RESULTS: Biopsy slides were available for review in 117/122 patients (63 females). Mean age at surgery was 32y +/- 10.3 (female)and 27y +/- 12.9 (male). Mean age at seizure onset was 19.4y +/-13.3 (female)and 15.5y +/- 11.9 (male) Resections were temporal in 77 (63.1%) of the patients, frontal in 33(27 %), centro-parietal in 12 patients (9.8%), and occipital in one (0.8%). Corpus callosotomy was performed in 3 patients (2.5%), functional hemispherectomy in 4 (3.3%) and stereotactic biopsy in 2 (1.6%). Pathologically, hippocampus was affected in 51 (41.7%) of the patients, while the amygdala gliosis occurred in 9 (7.4%), and tumors occurred in 33 (27%). Oligodendroglioma was the most common tumor (n=12, 9.8%). Other tumors included astrocytomas (n=8, 6.5%), ganglioglioma (n=6, 4.9%), and dysembyoplastic neuroepithelial tumor (n=6 4.9%). Malformative lesions accounted for 23% of the pathology. Focal cortical dysplasia was the most common (n=14, 11.5%). Other pathology includes cavernous angioma (n=8, 6.5%). There was no diagnostic abnormality recognized in 4.1% of the cases. 103 patients had at least 24 months follow-up data available. Of these 71.8% were in Engel Grade 1, 7.8% Grade 2, 8.7% Grade 3 and 11.7% in 4. On subgroup analysis patients with temporal lobe interventions had significantly better clinical outcome at 24 months when compared to those with extratemporal interventions: 76.8% of patients with temporal lobe intervention, 67.9% of frontal lobe patients, and 66.7% of parietal lobe patients, and the single occipital lobe patient were Engel Grade 1. By contrast, all 4 callosotomy patients, while helped, remained in Grade 4.
CONCLUSIONS: While older series reported no identifiable abnormality in up to 40% of cases, histological abnormalities were present in 96% of our patients. In particular malformations of cortical development and small tumors were seen more frequently than in historical series. At least part of this shift may reflect improved patient selection, especially related to the use of high resolution imaging.
Support: MW a Research Fellow at MGH Epilepsy Service is supported by NIH/WHO grant number F5NS10975A.