Abstracts

Rationale: There is a long-standing hypothesis that prolonged febrile seizures (PFS) cause mesial temporal sclerosis (MTS), a common pathology seen in temporal lobe epilepsy (TLE). Human and animal data suggest PFS may result in transient hippocampal changes, but whether PFS results in any long-term hippocampal changes or MTS is uncertain. In an ongoing follow-up study, we are investigating a prospectively ascertained population-based cohort of children to determine their outcomes within 10 years after an episode of convulsive status epilepticus. Here, we report preliminary quantitative hippocampal volumetry results in a subgroup that had PFS. Methods: All enrolled children were invited to have brain MR Imaging (MRI) on a Siemens 1.5T scanner. In addition to conventional sequences, T1-weighted images were acquired using a three-dimensional fast low angle shot (3D FLASH) sequence providing 1mm³ isometric voxels for volumetry. Hippocampal volumetry was performed on 3D FLASH images using FSL version 4.1.9. The images were reformatted in tilted coronal plane perpendicular to the long axis of hippocampus and each hippocampal slice was measured on each side. Hippocampal volumetry was performed blind to clinical information and using conventional hippocampal boundaries. The hippocampal volume (HV) was measured for each side and right-left asymmetry calculated using asymmetry index (AI= (HV Right- HV Left) / 0.5x (HV Right+ HV Left)). Nonparametric statistical tests were used to compare results between PFS group and controls in SPSS version 19. Results: Hippocampal volumetry was performed on 19 children with PFS (9 male, mean age 9.6 (range 7.5 to 13.1) years) and 5 controls (2 male, mean age 9.2 (range 7.6 to 11.9) years). Mean interval between PFS and MRI study was 8.15 (range 6.7 to 9.6) years. The right hippocampus was larger than the left in all controls and in 14 (74%) of the PFS group. There was no significant difference in the average HV between PFS group (median 3151 mm³, range 2255 to 3638), and controls (median 3109 mm³, range 2505 to 3176), or in their AI (median 0.069, range 0.0 to 0.28 for PFS; median 0.068, range 0.04 to 0.11 for controls). However, the AI for 6 children in the PFS group (31%) were more than the highest value in the control group. None of these children with a high AI had any reported neurological or neurocognitive impairments. Conclusions: Our data suggests that although there are no group differences in hippocampal volumes between children who had PFS and controls, a subgroup of PFS children will demonstrate hippocampal asymmetry within 10 years after PFS. Examination of factors that may identify such a group and longer term follow up are needed.