Abstracts

Rationale: Up to a third of children with epilepsy will be refractory to medications.  Patients with temporal lobe epilepsy tend to have the best outcomes following surgery.  We aim to assess the seizure outcome of non-lesional temproal lobe epilepsy in a pediatric population and compare this to our lesional caes.   Methods: Retrospective chart review was performed on patients with TLE who underwent temporal lobe resection at Boston Children’s Hospital from 2004-2016.  Patients were categorized into lesion and non-lesional cases.  Demographics, PET data, pathology and seizure outcome were collected for comparison.  Because the lesional cases far outnumbered non-lesion, age and sex matched lesional cases were randomly selected (n=32).  Favorable seizure outcome consisted of Engel 1 or 2 scores. Results: Out of 125 patients with TLE, 16 were considered non-lesional (2 of which did not undergo resection).  The remaining 14 patients comprised the non-lesional TLE (nl-TLE) cohort. All nl-TLE patients had concordant PET hypometabolism. Average duration of follow-up was 2.7 years for nl-TLE and 4.7 years for lesional TLE.  Favorable seizure outcome was achieved in 93% of lesional TLE cases compared to 74% of the nl-TLE (p=0.0372).   There was no difference between the two groups based on age of seizure onset, age at surgery, sex or lateralization.  Patients with nl-TLE were more likely to have isolated gliosis as the only finding on pathology (p=0.002).  Conclusions: Children with non-lesional TLE make up a small fraction of patients undergoing temporal lobe surgery at our tertiary pediatric epilepsy center. However even in children with non-lesional temporal lobe epilepsy, careful selection can lead to favorable outcomes.  In the absence of an identifiable MRI lesion, temporal lobe PET hypometabolism can aid in patient selection. Funding: No funding for this abstract