Abstracts

Rationale: Executive functions (EF) are skills necessary for goal-directed and complex activities, such as problem-solving, planning, self-regulation and working memory. EF deficits have been documented in 1/3 of youth with epilepsy, and predict long-term neurobehavioral problems, affecting academic and social outcomes. Understanding the neural basis of these deficits may help clarify why EF deficits arise so frequently in pediatric epilepsy. We used functional MRI, focused on EF networks, to examine whether abnormalities in functional connectivity underlie EF deficits in adolescents with epilepsy. Methods: 29 adolescents with MRI non-lesional epilepsy (mean age 15.41, 7 female, n=17 generalized epilepsy, n=6 localization-related and n=6 unclassified) and 20 healthy controls (mean age 15.35, 8F) ages 13-17 participated. Participants with epilepsy were taking 1-3 anti-epileptic medications; only one participant was noted as having continued seizures. Groups did not differ in age, race, body mass index or socioeconomic status.All participants performed an N-back fMRI task (block design 2-back versus 0-back condition with single-letter stimuli), 10 minutes of resting-state fMRI, and neuropsychological measures of EF (including working memory subtests from the Wechsler Intelligence Scale for Children or Wechsler Adult Intelligence Scale, and a switching/inhibition task measured by the Trail-Making subtest from the Delis–Kaplan Executive Function System). Regions of Interest (ROIs): A general linear model approach was used to create an activation map contrasting 2-back vs 0-back for the epilepsy and control groups combined. Functionally-defined ROIs were created based on clusters of group activation (initial threshold z>5, p < .05 corrected), observed in anterior cingulate/paracingulate, in middle frontal and parietal regions and frontal operculum/insula bilaterally, as well as medial and lateral cerebellum bilaterally (Crus I and II) (Fig. 1). Connectivity Analysis: These ROIs, based on the N-back task activation, were used to define a set of nodes in an EF network, and then, during the resting state, we examined differences in pairwise ROI-to-ROI connectivity between adolescents with epilepsy and controls. Results: Adolescents with epilepsy performed significantly more poorly than controls on neuropsychological testing related to EF (mean standard scores for WISC/WAIS Working Memory: 95.83 and 113.05; mean scaled scores for DKEFS Trail-Making: 9.43 and 12.00, both p < .01).  During the resting-state, pairwise connectivity was reduced between several pairs of nodes in the EF network for adolescents with epilepsy compared to controls (p < .05, FDR corrected, see Fig. 2). Conclusions: Functional connectivity appears to be reduced in EF networks in adolescents with epilepsy; these adolescents perform more poorly than peers on neuropsychological measures of EF. Abnormalities in functional connectivity may underlie EF deficits in adolescents with epilepsy despite good seizure control. Funding: Cincinnati Children's Research Innovation Pilot Program