Abstracts

Rationale: The objective of our research is to evaluate non-invasive Diffusion Tensor Imaging (DTI) for probing gray matter (GM) microstructure to reliably and accurately delineate malformations of cortical development (MCDs) in children with intractable seizures. Previous studies have shown that the most consistent predictor of seizure-free outcome is complete resection of aberrantly developed cortical architecture, cytoarchitecture, and myeloarchitecture underlying epileptiform activity. Thus, non-invasive procedures which reliably and accurately delineate epileptogenic MCDs for surgical resection have the potential to significantly reduce the devastating effects of chronic seizures on neurodevelopment in young children. Methods: Prior to resective surgery, a 3T GE Signa MRI scanner was used to acquire high-resolution structural images with a 3d-T1weighted sequence. In the same imaging session, a 21-direction DTI dataset was also acquired. FSL, a freely available software package, was used for conducting quantitative analyses of macro- and micro-structure throughout the cerebral cortex. Following resective surgery, neuropathology of the resected tissue was determined using standard histological procedures. Neuroimaging analyses were conducted blind to all clinical data. Results: In our preliminary case series (n=3), distinctive features of DTI metrics were observed in cortical GM regions of each research subject. Case #1 exhibited elevated MD values (~160%) ipsilateral to the seizure focus in left orbitofrontal GM (1447 x10-6 mm2/s) compared to values in right orbitofrontal GM (891 x10-6 mm2/s). FA values in the same GM regions were modestly elevated ~17% in the left relative to the right hemisphere. In Case #2, elevated FA values were observed ipsilateral to the seizure focus in frontal (LH: 0.56 +/- 0.06; RH: 0.36 +/- 0.05) and temporal (LH: 0.51 +/- 0.05; RH: 0.30 +/- 0.03) GM regions. MD values were not different between the same homologous regions. In Case #3, elevated MD values were observed ipsilateral to the seizure focus in temporal GM (RH: 1169 x10-6 mm2/s; LH: 1024 x10-6 mm2/s). FA values in the same GM regions were not different. Neuropathology reported that Chaslin's gliosis and variable cortical and white matter gliosis is present in all three brain resections; another common feature in all three is increased white matter neurons, singly dispersed. The left frontal lobe (#1) cortical dysplasia lacks balloon cells, or giant cytomegalic neurons. The left temporo-parietal resection (#2) has similar features. The right temporal lobe (#3) also shows similar cortical dysplasia; the right hippocampal resection has atypical Ammon's horn sclerosis. Conclusions: All 3 cases converged on a common neuropathological finding (cortical dysplasia). However, DTI findings varied across subjects, thus providing additional information about GM microstructure which could be useful. Additional cases will provide us with the opportunity to determine the sensitivity and specificity of DTI metrics in GM to inform us about MCDs, including their extent, boundaries, and connectivity with other GM regions.