Abstracts

Rationale: Behavior disorders are common comorbidities of pediatric epilepsy with enduring adverse effects on life performances, but their neuroanatomical underpinning is unclear. Subcortical structural abnormalities have been associated with childhood onset psychiatric disorders, suggesting that these brain regions play a role in the pathophysiology of behavior disturbances in children. We, therefore, hypothesized that subcortical volumes will be reduced in children with epilepsy, compared to healthy controls and the degree of atrophy will be linked to measures of behavior problems. Methods: T1 weighted SPGR MR images (1.5 Telsa GE Signa scanner) were obtained in 20 children with complex partial seizures (CPS, age 11.0±2.7 yrs), 21 children with childhood absence epilepsy (CAE, age 9.6± 2.1 yrs), and 27 healthy controls (HC, age 10.9±2.6 yrs). The groups were not significantly different in age, gender, and socioeconomic status. Subcortical volumes (caudate, putamen, pallidum, thalamus, nucleus accumbens, hippocampus and amygdala) were obtained from automated segmentation using FIRST (part of FSL, http://www.fmrib.ox.ac.uk/fsl)). Mixed models, with lateralization (left, right) as within-subject factor and group (HC, CPS, CAE) as between-subject factor, were used to compare subcortical volumes, controlling for age, gender and total intracranial volumes. Specific subcortical volumes in children with epilepsy that were significantly different from controls were correlated with Childhood Behavioral Checklist (CBCL) using Spearman correlation coefficients. Asymmetries were compared using ANCOVAs. Results: The left thalamic volumes were significantly different among the three groups (p=.03). Post-hoc pair-wise analyses showed left thalamic volumes were reduced in CPS (8036.9+802.6,p=0.01) and CAE (8146.2+899.7,p=0.05) compared to HC(8419.6+911.2). There were no significant group differences in other subcortical regions. Although a left greater than right volumetric asymmetry was found in all three groups (all p<.05), the degree of asymmetry was significantly smaller in CPS and CAE, compared to HC (p<.05). In CPS and CAE, the left thalamic volume was significantly correlated with social problems (CPS, r=-0.52, p=0.04; CAE, r=-0.46, p=0.05). In CPS, left thalamus volume was also associated with attention (r=-0.57, p=0.02) and school problems (r=.56, p=0.05). There were no correlation between left thalamus volume and CBCL scores in HC. Conclusions: The left thalamus was selectively atrophied and correlated with social problems in both CPS and CAE, suggesting that this subcortical region has a shared role in the pathophysiology of commonly found social problems in childhood onset epilepsy. However, the association of left thalamus atrophy with attention problems and school problems in CPS but not in CAE implies a syndrome specific contribution. These findings provide important new insights into the anatomical correlates of behavior problems in childhood onset epilepsy.