Abstracts

Rationale: In Krynauw s original description (1950) on hemispherectomy for intractable epilepsy, he stated that the hemiparesis/hemiplegia is much more severe if the basal ganglia were also removed. Our own studies using PET scanning have shown that following large cortical resections in children with intractable epilepsy, robust functional changes occur in the striatum on the side of resection (Chugani & Jacobs, 1994; Chugani et al., 2008), presumably related to plasticity mechanisms. We have suggested that these metabolic changes may be due to increased cortico-striatal projections from the contralateral hemisphere to the ipsilateral striatum. In the present study, we used diffusion tensor imaging (DTI) and probabilistic tractography to test this hypothesis. Methods: We analyzed postsurgical DTI scans from 8 children (age: 8 4.3 years) with intractable epilepsy who had undergone left-sided cortical resection (anatomical hemispherectomy: 4; subtotal hemispherectomy: 2; fronto-temporal resection: 2) and compared these scans with those from 14 normal controls (age: 7.6 3.1 years). All 8 patients had normal pre-surgical MRI and FDG PET findings in the right hemisphere and became seizure free after surgery. In each child, regions were manually drawn on the left caudate using the structural MRI and used as a seed region for probabilistic tractography (FSL 4.1, Oxford, UK) applying the two crossing fibers per voxel model and 10,000 samples per seed voxel. The mean connectivity values were calculated for 5 contralateral cortical regions (frontal, parietal, temporal, occipital and insular cortex). Results: Repeated measures ANOVA showed a significant interaction between the 2 groups and the 5 contralateral cortical regions (F = 2.8; p = 0.05). More specifically, the mean connectivity values between the left caudate and right frontal cortex (57 /- 21 vs. 17.8 /-16; p = 0.1) and insula (1.5 /- 0.3 vs. 0.47 /- 0.28; p = 0.032) showed an apparent increase in the postsurgical group compared to controls. Whereas connectivity values between left caudate and right parietal cortex showed a decrease (0.95 /- 1 vs. 4.3 /- 1; p = 0.037), the occipital (0.44 /- 0.2 vs. 0.83 /- 0.16; p = 0.16) and temporal cortex (1.3 /- 0.4 vs. 1.7 /- 0.33; p = 0.54) showed no difference in connectivity between postsurgical and control groups. Conclusions: The specific increase in fiber connectivity between caudate ipsilateral to the resection and contralateral frontal cortex and insula is consistent with our previous findings of functional changes in striatum following resection and supports the notion that, following large cortical resections, the ipsilateral caudate participates in functional reorganization. However, since preoperative DTI scans were not available in this cohort, we cannot exclude the possibility that these connectivity changes were the consequence of the left-sided lesion itself and were already present prior to surgery. This issue will be further elaborated in future studies.