Abstracts

Rationale: Cross-sectional brain MRI studies have identified morphometric abnormalities in children with chronic epilepsy. However, the timing and progression of these abnormalities remains uncertain. This prospective cohort investigation of children with new/recent onset epilepsy and healthy controls (age 8-18) compared patterns of brain development over a 2 year interval. Methods: Structural MRIs of 44 children with newly diagnosed epilepsy (20 IGE, 24 LRE) and 34 healthy controls (age 8-18 years) were obtained at study entry and two years later. We applied pair-wise deformation-based morphometry to baseline and follow-up scans of each subject to quantify annual change in local tissue volume. In particular, for each MR image, after the removal of non-brain tissue, remaining image volume was corrected for intensity inhomogeneities. Then, for each subject, the follow-up brain image volume was fluid-flow warped to the subject’s baseline image volume. The annual change in local tissue volume was quantified in terms of the logarithmic determinate of Jacobian matrix --- log(|Jac|) --- of the fluid-flow warping field at each image voxel. Using an automated surface-based morphometry analysis framework, each subject’s annual gray matter volume loss estimate was first locally mapped to the subject’s cortical surface representation and then mapped to a reference cortical surface for statistical group comparison. These cortical surface maps of annual change in local gray matter volume of each individual were then analyzed using a general linear model. The healthy and epilepsy groups were compared with a categorical variable with sex and age as covariates. Because the statistics were computed independently at each cortical surface point, the data were FDR corrected. Results: The overall comparison of controls versus epilepsy groups reveals regions of brain where children with epilepsy demonstrate statistically significant differences in local gray matter volume reduction including left frontal, right medial occipital, left lateral occipital, and bilateral temporal cortices. In the comparison of controls to epilepsy syndromes, the differences in temporal and medial occipital cortices were associated with idiopathic generalized epilepsy (IGE) (i.e., children with CAE or JME), while the differences in frontal cortices were mostly associated with localization-related epilepsy (LRE) (i.e., children with CPS). Differences in the log(|Jac|) distribution between groups indicate generally greater gray matter volume loss, except for a lesser degree of gray matter volume loss in the left lateral occipital region in IGE and in the parieto-temporal region in LRE. Conclusions: This longitudinal (2 year) analysis of brain development in children with epilepsy compared to healthy controls demonstrates areas of accelerated as well as slowed cortical change. The findings suggest altered patterns of brain development in children with new/recent onset epilepsy.