Abstracts

Rationale: Medically-refractory epilepsy in children can be associated with widespread epileptogenic foci involving one hemisphere. “Hemispheric cortical dysplasia” (HCD) was previously reported as an underlying etiology for such widespread epileptogenic foci (see references below). In the present study, we reviewed the clinical data of patients with HCD seen in our center. Methods: We retrospectively reviewed the medical charts of a consecutive series of 39 patients (age: 3 months to 18 years; 17 girls) who underwent anatomical hemispherectomy (N=27) or subtotal hemispherectomy sparing the primary sensorimotor cortex (N=12) in Children’s Hospital of Michigan between January 2002 and May 2008. Results: Clinico-pathological assessment showed HCD in five children (13%). Ulegyria-porencephaly associated with a vascular insult was noted in seven patients; hemimegalencephaly in five; Rasmussen encephalitis in four; schizencephaly in two; tuberous sclerosis complex in two; and congenital perisylvian syndrome in one patient. No definite abnormalities other than gliosis were noted in the remaining nine patients. Clinical features of the five children with HCD included: developmental delay in four cases, cognitive impairment in one, and mild hemiparesis in three cases. The age of seizure-onset ranged from 1 to 24 months. History of epileptic spasms, complex partial seizures and secondary generalized tonic-clonic seizures were noted in 3, 4 and 3 patients, respectively. All had daily refractory seizures. Scalp EEG showed interictal epileptiform discharges in the affected hemisphere in all patients. Generalized interictal spike-waves and hypsarrhythmia were noted in one child. Brain MRI was initially read as normal in two patients and hemispheric abnormalities were documented in the other three patients. All FDG-PET studies (performed in 4 patients) showed widespread cortical hypometabolism in the epileptogenic hemisphere and normal metabolism pattern in the contralateral hemisphere. Four patients have been seizure-free following hemispherectomy (N=2) or subtotal hemispherectomy (N=2) (duration of follow up 14 to 66 months) ; one child died due to DIC associated with pseudomonas sepsis two days after surgery; the parents later reported that they had been supplementing the child with “immune-augmenting medications” obtained overseas. Conclusions: HCD occurs later in gestation than hemimegalencephaly and may be missed on MRI when seizures first present. FDG-PET will indicate the extent of hemispheric involvement and sparing of sensorimotor cortex, thus guiding surgery to either hemispherectomy or a subtotal procedure sparing sensorimotor cortex.