Early alterations of thalamo- and hippocampo-cortical functional connectivity are biomarkers of epileptogenesis after traumatic brain injury

The Epilepsy Bioinformatics Study for Antiepileptogenic Therapy (EpiBioS4Rx) study is a prospective multicenter clinical observational study to identify early biomarkers of epileptogenesis after moderate-to-severe traumatic brain injury (TBI). In this preliminary analysis of 37 patients, using a seed-based approach applied to acute (i.e., [&le;] 14 days) functional magnetic resonance (MRI) imaging data, we directly test the hypothesis that the epileptogenic process following brain trauma is associated with functional changes within hippocampal and thalamo-cortical networks. Additionally, we hypothesize that the network connectivity involving thalamic and hippocampal circuits underlying early and late-onset epileptogenesis would differ. The three groups did not differ by sex distribution ({chi}2(2) = 1.8, p = .407), age (H(2) = 4.227, p = .121), admission Glasgow Coma Scale (H(2) = 3.850, p = .146) or postinjury day of the MRI session (H(2) = .695, p = .706). The primary finding is that patients with early seizures, a sign of early epileptogenesis, exhibited pattern 1, namely, an increased positive connectivity in thalamic and hippocampal networks, as compared to patients who had no epileptogenesis, or late epileptogenesis (p < .05, FWE-corrected at the cluster level). In contrast, this finding was absent in those patients who exhibited late seizures, with the latter group displayed pattern 2, namely, a lower positive and higher negative connectivity in the hippocampal network, as compared to patients who had no signs of epileptogenesis (p < .05, FWE-corrected at the cluster level). Patients with either pattern 1 or pattern 2 connectivity profiles in thalamic and hippocampal networks were significantly predictive of late (i.e., between 7 days and 2 years) epileptogenesis following brain trauma. A Receiver Operating Characteristic (ROC) Curve analysis model that included thalamic and hippocampal functional connectivity values presented an Area Under the Curve (AUC) 87.7, specificity 86.7, and sensitivity 84.6. Our results indicate that dysfunction in hippocampal and thalamo-cortical networks are potential biomarkers for early and late epileptogenesis following a TBI.