Abstracts

Rationale: Surgical procedures for the treatment of intractable epilepsy include focal cortical resections, multilobar resections, hemispherectomy, corpus callosotomy and multiple subpial transections (MSTs). Resections tend to be larger in children, particularly infants, compared to adults due to underlying pathology, greater potential for plasticity of the developing brain and the goal of promoting a normal developmental trajectory by removing as much nociferous cortex as possible. Some children show unilateral multifocal seizure onsets involving much of the hemisphere. If there is significant hemiparesis already present, cerebral hemispherectomy would be the procedure of choice. However, if there is only a mild or no hemiparesis, it is preferable to spare the primary sensory/motor cortex. We now report the results of this ‘subtotal hemispherectomy' in 23 children. Methods: The 23 children (12 males, 11 females) were operated on between April 2001 and March 2013 at the Children's Hospital of Michigan (Detroit). Age at time of surgery ranged from 1 year 4 months to 14 years 2 months (mean: 5 years 8 months). Patients were evaluated with scalp interictal and ictal EEG, MRI and FDG-PET scans, and neuropsychological assessments when applicable. Some children also received PET with ¹¹C-flumazenil or ¹¹C-alphamethyl-tryptophan under research protocols. Subsequently, the patients were discussed in a multidisciplinary Epilepsy Surgery Conference, and a consensus was reached pertaining to candidacy for surgery and optimum surgical approach, with the understanding that the actual extent of resection would be based on results from intracranial EEG monitoring. The surgical outcome was reported based on the ILAE classification (Class 1-6). Results: Among the 23 patients, 11 had infantile or epileptic spasms as their major seizure type; the spasms occurred as the only seizure type in 8 patients, or were associated with complex partial seizures with (n=2) or without (n=1) secondary generalization. MRI was normal in 11 children and showed focal abnormalities in the remainder. FDG-PET was normal in only one subject. All except two children underwent chronic subdural EEG recordings. MSTs were performed over the sensory/motor cortex in 3 subjects. On histopathology, various malformations were seen in 9 subjects (including two with tuberous sclerosis); the remainder showed gliosis alone (n=12), porencephaly (n=1), and gliosis with microglial activation (n=1). Follow-up ranged from 13 to 157 months (mean=65 months). Outcome consisted of Class 1 (n=17, 74%), Class 2 (n=2), Class 3 (n=1), Class 4 (n=1) and Class 5 (n=2). The only child with a normal PET (and normal MRI) scan had a Class 5 outcome. Conclusions: The present study demonstrates that extensive unilateral resections sparing only sensory/motor cortex can be performed with excellent results in seizure control. Even with the presence of widespread unilateral epileptogenicity or anatomical/functional imaging abnormalities, complete hemispherectomy can often be avoided, particularly when there is little or no hemiparesis.