Abstracts

Rationale: Temporal lobe epilepsy (TLE) is increasingly reconceptualised as a network disorder, with a growing literature suggesting changes in large-scale network organization. Most of these studies have studied topological parameters of brain graphs, in which network links were purely quantified in terms of connectivity strength, thus not taking into account physical properties of the brain. In the current work, we enriched functional connectomics with geodesic distance information, thereby generating a novel and physically-grounded measure of average anatomical distance of a brain area’s connectivity profile. We illustrate feasibility of our approach in areal stratification of healthy connectomes and applied it to unveil shifts in the connectional distance distribution in drug-resistant TLE. Methods: We studied a cohort of 30 drug-resistant unilateral TLE patients (14 males; 27.2±8.6 years) and 26 age- and sex-matched healthy controls. In each subject, we obtained high-resolution structural and resting-state functional MRI at 3T. All patients had postoperative histological confirmation of hippocampal sclerosis. Using a boundary-based registration technique, we mapped the resting-state functional time-series to each participant’s cortical surface and computed pairwise correlations between all pairs of regions to generate individualized functional connectomes. For each region within the z-transformed connectome matrices, we retained the top 10% of weighted connections and calculated the average geodesic distance to all other regions in this connectivity profile. Distance maps in patients were z-scored relative to data in controls and sorted into ipsilateral/contralateral to the focus. Using surface-based linear models, we compared patients to controls and corrected for multiple comparisons at a family-wise error rate (FWE) of p<0.05. Results: In healthy controls, we observed marked inter-regional variations in average connectivity distance across the cortical mantle, with transmodal networks showing longest distances and sensory cortices showing the shortest distances (Figure 1).  Notably, comparing TLE patients to controls, we observed marked reductions in connectivity distance in an ipsilateral network encompassing lateral temporal, insular, and ventrolateral prefrontal cortices.  Findings did not overlap with TLE-related cortical thinning, which followed a bilateral distribution affecting predominantly fronto-central areas. Furthermore, distance enriched centrality alterations persisted even after regionally correcting for cortical thinning, suggesting independence of functional network effects from regional morphological alterations (Figure 2). Conclusions: Our findings provide novel evidence for marked shifts in the connectivity distance distribution in drug-resistant TLE. The strictly ipsilateral pattern of findings and complementarity to cortical atrophy suggests clinical potential of connectomic approaches enriched with physical information.  Funding: CIHR, SickKids Foundation, NSFC, FRQS