Abstracts

Rationale: Recent studies have shown that temporal lobe epilepsy (TLE) patients demonstrate altered functional connectivity of subcortical arousal structures, such as the thalamus and brainstem, and this altered connectivity is related to neurocognitive deficits and seizure frequency. Additionally, prior animal work has shown that activity of cholinergic neurons of the basal forebrain is suppressed during consciousness impairing seizures (Motelow et al. Neuron. 2015; 85[3]:561-527). We hypothesize that recurrent ictal events incite interictal connectivity disturbances of the nucleus basalis which may relate to diffuse network problems and comorbidities of TLE. In this study, we examine whether nucleus basalis exhibits abnormal network connectivity in TLE patients as compared to healthy controls. Methods: We evaluated 40 adult unilateral TLE patients and 40 matched healthy controls using functional MRI (fMRI) to measure resting-state functional connectivity between nucleus basalis and neocortex. fMRI was corrected using RETROICOR; functional connectivity was calculated between regions of interest using partial correlation with mean white matter BOLD and motion as confounds. Further, epilepsy is increasingly recognized as a network disorder. Therefore, to understand alterations of network properties of nucleus basalis ipsilateral and contralateral to the epileptogenic temporal lobe we calculated graph theory measures (positive node strength, negative node strength, clustering coefficient). We then related nucleus basalis connectivity and graph theory network measures to disease severity. Results: Functional connectivity of the bilateral nucleus basalis with frontoparietal neocortex was lower in patients as compared to healthy controls (p<0.01, paired t-test). Patients with greater frequency of consciousness impairing seizures had more abnormal connectivity between nucleus basalis and frontoparietal neocortex (r=-0.429, p<0.01, Pearson correlation). Next, we examined the network properties of the individual nucleus basalis (ipsilateral and contralateral to the epileptogenic temporal lobe). Both the ipsilateral and contralateral nucleus basalis displayed lower positive node strength and lower clustering coefficients in TLE patients as compared to controls (p<0.05 for all, paired t-test). However, only the ipsilateral nucleus basalis displayed larger negative node strength in TLE patients as compared to controls (p<0.01, paired t-test). Additionally, of all 114 brain regions examined in the study, the ipsilateral nucleus basalis displayed the single greatest decrease in clustering coefficient as compared to controls. Interestingly, higher ipsilateral nucleus basalis clustering coefficient (i.e. closer to healthy control values) was associated with both higher verbal IQ and higher performance on tests of verbal memory (r=0.38-0.40, p<0.05 for both, Pearson correlation). Conclusions: The nucleus basalis exhibits abnormalities of functional connectivity and network properties in TLE patients as compared to healthy controls. Greater aberrations of these measures are associated with greater seizure frequency and worse verbal neurocognition. These results are the first to demonstrate abnormal connectivity of nucleus basalis in TLE which may lead to an improved understanding of the widespread long-term effects of seizures on brain networks in TLE. Funding: This work was supported by NIH R00 NS097618 (DJE), R01 NS108445 (VLM), R01 NS110130 (VLM), T32 EB021937 (HFJG), T32 GM07347 (HFJG), F31 NS106735 (HFJG), and the Vanderbilt Institute for Surgery and Engineering (VISE).