Abstracts

Rationale: In temporal lobe epilepsy (TLE), focal seizures lead to widespread brain network alterations and significant neuropsychological problems. We have hypothesized that recurrent seizures lead to dysfunction of subcortical arousal centers, which in turn contributes to neocortical connectivity disturbances and neurocognitive deficits. This is supported by our recent studies in TLE patients, suggesting overall reductions in functional connectivity in the brainstem ascending reticular activating system (ARAS). However, functional connectivity patterns of individual ARAS nuclei, or structural connectivity studies of ARAS networks, have not been previously examined in TLE. Methods: In 26 adult TLE patients and 26 matched controls, we used magnetic resonance imaging (MRI) to measure structural connectivity (probabilistic diffusion tractography) and functional connectivity (functional MRI partial correlation) of the cuneiform/subcuneiform nuclei (CSC), pedunculopontine nucleus (PPN), and ventral tegmental area (VTA). Connectivity patterns were compared across modality and anatomical regions, and then related to neuropsychological test performance and disease parameters. Results: Across all subjects, structural connectivity of CSC, PPN, and VTA was highest to the thalamus, frontal lobe, and basal ganglia, while functional connectivity was strongest at thalamus, limbic structures, and para-limbic cortex. Compared to controls, TLE patients demonstrated significant reductions in both structural and functional connectivity of CSC, PPN, and VTA, most prominently in neocortical regions (p < 0.05, paired t-tests, corrected). Among epilepsy patients, while structural connectivity of CSC was positively correlated with performance IQ, functional connectivity of CSC was significantly related to verbal IQ (p < 0.05, Pearson correlations). Finally, CSC structural connectivity was significantly lower in patients with secondarily-generalized seizures compared to those without this history (p < 0.05, unpaired t-test), and a similar nonsignificant trend was observed with CSC functional connectivity (p = 0.07, unpaired t-test). No relationships were observed between connectivity measures and seizure frequency or epilepsy duration. Conclusions: Brain connectivity perturbations of brainstem arousal centers may result from recurrent seizures in TLE, and may contribute to neurocognitive problems suffered in this disorder. Differences are seen between functional vs. structural ARAS connectivity in terms of anatomic distribution and associations with neuropsychological performance. In particular, CSC connectivity appears related to certain neuropsychological and disease parameters. These studies may ultimately lead to novel neuromodulation targets to treat or prevent deleterious brain network effects of seizures in TLE. Funding: This work was supported in part by the National Institutes of Health grants K99 NS097618 (DJE) and R01 NS075260 (VLM).