Abstracts

Rationale: Controlled longitudinal neuroimaging studies are lacking in juvenile myoclonic epilepsy (JME) and the onset and course of structural abnormalities are unknown. Although cross-sectional studies have detected focal brain structural abnormalities, the findings have been heterogeneous; revealing both increased and decreased volumes in spatially variable regions. The aim of the current study was to determine prospectively the rate and location of brain tissue changes over 2 years in a cohort of new and recent onset JME compared to healthy controls. Further, we correlated brain tissue change to seizure severity.Methods: Two sets of T1 weighted SPGR images were obtained, 2 years apart, on a 1.5 Telsa GE Signa scanner in 16 individuals with new and recent onset JME (age = 15.4 2.8 years; epilepsy duration = 8.5 3.7 months) and 42 healthy controls (HC, age = 12.9 3.1 years). We computed percent brain volume change (PBVC) for each participant, using SIENA (Structural Image Evaluation, using Normalization, of Atrophy, part of FSL, http://www.fmrib.ox.ac.uk/fsl). First, brain and skull images are extracted and the time 1 and time 2 images are registered to each other. Second, tissue segmentation is carried out to determine the boundary between brain and non-brain edge in both time 1 and time 2 scans. Third, for every edge point in time 1 scan, a search is performed to find the closest matching edge point in time 2 scan, based on gradient intensity. Once the edge points are matched, a value is derived based on the magnitude of edge points displacement between the two scans. Finally, PBVC is calculated based on the brain volume change divided by total brain volume. Analysis of variance was performed, with age as a covariate, to compare PBVC between JME and control groups. Voxel-wise analysis with correction for multiple comparisons was performed to determine the location of volumetric changes. PBVC in JME was correlated with seizure severity, using Spearman rho.Results: Over 2 years, the JME group exhibited 1.08% greater brain volume loss compared to controls (JME=-3.05 -2.34%, PBVC standard deviation, controls=-1.97 -1.22, p=0.012). Voxel-wise analysis revealed that the most significant atrophy regions were located in the left lateral frontal, insular and cerebellar regions. Greater atrophy rate was significantly correlated with greater seizure severity (Seizure Severity Scale for Children, Spearman rho = -0.56, p=0.028).Conclusions: This is the first longitudinal neuroimaging investigation of new onset JME. The results indicate abnormal brain development characterized by accelerated brain atrophy rate compared to controls, even early in course of the epilepsy. Voxel-wise analysis demonstrated that atrophy is located in networks previously implicated in JME and concordant with cross-sectional imaging studies. The accelerated atrophy rate is clinical important, as it relates to seizure severity.