Abstracts

Rationale: Secondarily generalized seizures (SG) manifest with dramatic motor manifestations of generalized tonic?"clonic/tonic seizures. This study aimed at understanding the network changes associated with secondary generalization in drug resistant partial epilepsy. We used multimodal graph theory analysis of simultaneous EEG-fMRI in temporal lobe (TLE) and extratemporal lobe epilepsies (ETLE) manifesting with or without secondary generalization (SG) to understand the regional differences and susceptible brain regions that are likely to be repeatedly affected. Methods: Simultaneous EEG fMRI was recorded in 84 patients with refractory focal epilepsy (TLE: 44, ETLE: 39). 3 EPIs were obtained per subject. Hence, the number of analyzable fMRI scans was: TLE with SG (TLE-SG): 60, TLE without SG (TLE-WSG): 44, ETLE with SG (ETLE-SG): 40, ETLE without SG (TLE-WSG):28. The 31x31 EEG phase lag index matrices (HERMES toolbox) and 200x200 (Craddock's atlas) fMRI correlation matrices were combined with a mutual lexicon of Graph Theory. Clustering coefficient (CC), path length (PL), small worldness (SW), local efficiency (LE) and global efficiency (GE) were derived using Brain Connectivity Toolbox. CC was used to determine the significant brain regions involved in seizures with secondary generalization. FDR corrected P value for multiple comparison was used to determine significance level. Results: In TLE, secondarily generalized seizures were associated with reduced CC, SW and LE in fMRI and increased CC, SW but decreased LE in EEG. Left thalamus, lentiform, cerebellar tonsil, pyramis and motor cortex and varied parts of bilateral lateral cortical gyri of temporal-frontal and inferior parietal region were abnormal in TLE-SG and spatially this was matched with the F4, F8, P7 channels that showed significant differences in EEG graphs. In ETLE, secondarily generalized seizures were associated with reduced CC, SW in fMRI and decreased CC, SW, GE, LE in EEG. Bilateral thalamus, bilateral sensory-motor cortex, right insula, left claustrum, left pyramis and left anterior cingulate cortex, bilateral middle frontal gyri, bilateral occipital and left temporal regions were abnormal. F3, P7 were the channels that showed significant changes on EEG graphs. Conclusions: Multimodal graph theory analysis of simultaneous EEG fMRI revealed "cortico-thalamo-cortical circuitry" to be involved in secondary generalization. Both EEG and fMRI were in concurrence and found different paths for secondary generalization in TLE and ETLE. Secondary generalization in ETLE was associated with bilateral thalamic and bilateral frontal and parietooccipital involvement whereas TLE was associated with left thalamic and prominent neocortical temporofrontal involvement. Funding: Nil