Abstracts

Rationale: The understanding of epilepsy as a network organization disorder introduced the idea of using the brain connectome as a prognostic indicator. The objective of this study is to assess the potential of metabolic connectivity as a predictive factor of outcome in epilepsy surgery by examining preoperative metabolic connectivity in respect to post-surgical outcomes of the patients underwent Laser interstitial Thermal Therapy (LiTT) for medically resistant temporal lobe epilepsy (TLE). Methods: We collected preoperative positron emission tomography (PET) scans of 24 medically resistant TLE patients with unilateral mesial temporal sclerosis who underwent LiTT for refractory epilepsy. Patients were followed up for at least one year after surgery and classified according to seizure recurrence as seizure free (Engel class IA) or not seizure free (non-IA class) groups. Image preprocessing was done in SPM12. Connectivity matrices were constructed based on interregional glucose metabolism values correlation. In addition, graph theory analysis was performed in Brain Analysis using Graph Theory (BRAPH) software to compare metabolic connectivity between the two groups. Additionally, we performed connectivity-based analysis by conducting a z-test on the correlation values to examine the significantly different connections between different regions of the brain in the two groups. Results: Metabolic network organization in the seizure-free group differs substantially compared with the not seizure-free group. Compared to seizure-free patients, the temporal pole and cingulate regions had higher connectivity locally with the surrounding areas in the not seizure-free group, while multiple regions including cingulate, precentral gyri, postcentral gyri and superior parietal gyrus were highly clustered with surrounding nodes indicating greater functional segregation. In the connectivity-based analysis, the matrices show stronger connections in the seizure-free group between multiple areas including frontal lobe and insula, frontal lobe and cingulate, insula and temporal lobe, cingulate and temporal lobe, regions of the parietal lobes, regions of the temporal lobe. On the other hand, there were stronger connections between frontal and temporal lobes in the not seizure-free group. Conclusions: Our study demonstrated a relationship between presurgical metabolic connectivity and post-surgical seizure outcome of the patients who had LiTT surgery and the potential role as an imaging biomarker to predict surgical outcomes in this patient cohort. Extension of the disease to extratemporal networks, specifically the limbic network, plays a role in seizure recurrence after surgery. Although MTS typically involves sclerosis of the hippocampus, we can conclude that this pathological process will involve other structures located medially within the temporal lobes of the brain as well as neuronal connections projecting to other structures involving the limbic system, such as the temporal pole and cingulate. Combined with the current tests used in clinical practice, metabolic connectivity may be used as an additional prognostic/diagnostic factor during pre-surgical evaluation for refractory TLE patients. Funding: No funding