Abstracts

Rationale: Hemispheric disconnection for some forms of pharmacoresistant epilepsy has been employed since the first half of the 20th century. The earliest of such procedures, the anatomic hemispherectomy, involved complete removal of the hemisphere. Advancements led to the functional hemispherectomy and hemispherotomies, in which, through limited or no cortical resection, there is hemispheric disconnection. Numerous studies have endorsed these procedures as effective therapy for medically refractory focal onset epilepsy with diffuse pathology and ictal onset from one hemisphere. This report describes the hemispherectomy experience of a single pediatric institution between 2009 and 2016.  Methods: The epilepsy database was reviewed retrospectively for allpatients who had undergone hemispheric disconnections between 2009 and 2016. Exclusion consisted of diagnosis of generalized or non-epileptic seizures, bilateral or non-localizing seizure onset, incomplete hemispheric disconnection, and follow-up less than 3 months. Primary outcomes included seizure reduction as measured by Engels score and post-operative neurocognitive function. Secondary outcomes examined duration of stay, rates of major complications, and reduction in number of anti-epileptic agents at long-term follow-up. Stratification was based on presumed etiology of seizures, surgical method performed, age at surgery, age of seizure onset, and pre-operative neuropsychological assessment. Results: The database identified 27 cases, of which 22 were included after the aforementioned criteria. Ages ranged from 6 weeks to 25 years with mean 8.0 + 7.4 years.  Mean time of follow-up was 2.9 years (5 months to 7 years). A majority of patients (68%) achieved seizure-free outcomes, while all showed significant reduction in seizures, with preservation or improvement in neurocognitive function. AEDs were reduced in 71% of patients with 28.5% off all AEDs, and AED number remained unchanged in 23%. AED numbers increased in one individual. The median post-operative stay was 8.1 + 4.7 days. The main complications encountered were delayed shunting (22%) and aseptic meningitis aborted with steroids (22%). Rates of each are consistent with other published results. No patients required acute transfusion of blood products, the most commonly identified complication of hemispherotomy with a reported prevalence of around 30% of cases. Possible explanations for this finding include tertiary hospital specialization, surgeon experience, older mean age of patients, and preference for hemispherotomy over hemispherectomy.  Conclusions: This study supports the utility of hemispheric disconnection in pharmacoresistant epilepsy with predominant involvement of one hemisphere. It provides further proof of long-term seizure and AED reduction, and improved neurocognitive function in most patients. Future consideration would attempt to elucidate outcomes and complications as affected by duration of seizures and age at the time of procedure, for which this sample size considered insufficient for robust analysis.  Funding: None