(755) Seizure severity differentially impacts functional MRI connectivity in limbic and frontal regions in patients with temporal lobe epilepsy

Rationale: Frequent focal seizures carry a high risk of significant disability and cognitive impairment. Recent high-density EEG work suggests the presence of local connectivity increases in focal epilepsy networks (Boly et al., Brain 2017). Because intracranially, focal seizures propagate in a nearest-neighbor fashion, we hypothesize that they may induce increases in cortico-cortical connectivity that are especially pronounced in the short distance range in the epileptic network of patients with temporal lobe epilepsy (TLE). We also hypothesized that secondarily generalized seizures may be accompanied by connectivity changes outside of the limbic network.

Methods: We analyzed data from 51 TLE patients (age range 20-57) part of the Epilepsy Connectome Project (ECP). Resting-state fMRI data were obtained using 3T MRI Human Connectome Project (HCP) sequences and preprocessed using the HCP pipeline (Glasser et al. 2013) plus a nuisance regression approach (Satterthwaite et al. 2013). For each subject, vertex-wise matrices of correlation with neighbors at short (0-6mm) and long (105-175) distances as well as the average connectivity across all distances were computed. Using Statistical Parametric Mapping software (SPM), random effects models assessed for a correlation between the monthly frequency of focal aware/focal impaired partial seizures (FA/FI), and general tonic-clonic seizures (GTC) with short range, long range and average connectivity. Results were considered significant at p< 0.05 corrected for multiple comparisons at the cluster level using SPM and/or using Probabilistic Threshold-free Cluster Enhancement (pTFCE) as implemented in SPM. 

Results: Higher FA frequency was correlated with increased long range connectivity in the left supramarginal gyrus (surviving both SPM and pTFCE). Short range connectivity in the dorsal posterior cingulate area was also positively correlated with FA frequency (surviving SPM only).

Higher FI frequency was correlated with increased short range connectivity in the right supramarginal gyrus, right postcentral gyrus, left caudate head, and left pars orbitalis. Further, direct comparison showed that these areas were more correlated with FI frequency than with FA frequency.

GTC frequency per se was not significantly correlated with brain connectivity changes (which may be due to low power given only few patients had GTC). Direct comparison to both FA and FI however showed higher average connectivity in the left anterior frontal cortex with more frequent GTC. Average connectivity in right frontal eye field and dorsolateral prefrontal cortex was also significantly more associated with GTC than with FI frequency.

Conclusions: Our findings suggest that even the least severe seizures (FA) may lead to increases in connectivity not only within but also outside of the limbic network in patients with TLE. FI seizures were associated with greater connectivity in the parietal and frontal lobes, while GTC frequency was specifically correlated with increased connectivity in the frontal lobe.

Funding: Please list any funding that was received in support of this abstract.: Epilepsy Connectome Project and Tiny Blue Dot Foundation