Abstracts

Rationale: Operculo-insular epilepsy (OIE) is an under-recognized condition with the potential of mimicking neocortical and mesiotemporal epilepsies. Previous studies revealed structural connectivity changes in the epileptic network of focal epilepsy. To our knowledge, however, no studies have looked at the structural connectome in patients with OIE.

Methods: Nine patients with OIE, 8 age- and sex-matched patients with temporal lobe epilepsy (TLE) and 22 age- and sex-matched healthy controls (HC) were scanned using diffusion MRI (b=1500 s/mm²) and T1w sequences. Tractograms were built using Tractoflow and surface-enhanced particle filtering tractography. Convex Optimization Modeling for Micro-structure Informed Tractography (COMMIT) was then used to filter the raw tractogram and derive COMMIT-weighted structural connectivity matrices. COMMIT weights were used as markers of ‘connectivity strength’. (Figure 1).

In addition to (i) 249 x 249 COMMIT-weighted whole-brain matrices, sub-networks consisting of (ii) 6 x 243 matrices linking the 6 subinsular regions to all 243 extra-insular regions (insula-extrainsula subnetwork) and (iii) 6 x 6 matrices linking the 6 subinsular regions to each other were computed (insular subnetwork). Matrices were compared using general linear models between (a) HCs and OIE patients, and (b) OIE and TLE patients. Binary adjacency matrices of whole-brain networks were also built and a graph theory analysis of both regional and global measures was performed to compare (a) the HC to the OIE group and (b) the OIE to the TLE group.

Results: On whole-brain analyses, significant increases in COMMIT weights were observed bilaterally in multiple bundles of OIE patients as compared to HCs. Similarly, a wider pattern of increased connectivity was detected in OIE patients as compared to TLE patients. In OIE patients, ipsilateral ‘hyperconnections’ were observed between the dorsal granular insula and the pregenual cingulate gyrus (OIE group vs HC group in insula-extrainsula subnetwork analysis) and between insular subregions (OIE group vs TLE group in the insular subnetwork analysis). Graph theoretic analyses revealed higher regional connectivity within ipsilateral insular subregions of OIE patients (OIE group vs TLE group).

Conclusions: Our results reveal, for the first time, the distribution of structural connectivity in patients with OIE. The wider pattern of increased ‘connectivity strength’ in patients with OIE and could suggest a more diffuse epileptic network than TLE. Although more work is necessary, the differential morphologic pattern of ‘connectivity strength’ could eventually constitute a complementary tool to differentiate two challengingly distinguishable epilepsies of anatomically adjacent seizure origins, namely OIE from TLE.

Funding: Please list any funding that was received in support of this abstract.: Savoy Foundation for Epilepsy; Fonds de Recherche du Québec - Santé (277581); Quebec Bio-Imaging Network (5886); Canadian Institute of Health Research (CIHR; MOP-BSC343410-97930-DLGNH); Natural Sciences and Engineering Research Council of Canada (RGPIN-2016-05216N); Université de Sherbrooke Institutional Chair in Neuroinformatics.