Publication details
Metabolic connectivity and memory in mesial temporal lobe epilepsy
| Authors | |
|---|---|
| Year of publication | 2026 |
| Type | Peer-reviewed scientific article |
| Magazine / Source | BRAIN COMMUNICATIONS |
| MU Faculty or unit | |
| Citation | |
| web | https://academic.oup.com/braincomms/article/8/5/fcag365/8916464 |
| Doi | https://doi.org/10.1093/braincomms/fcag365 |
| Keywords | mesial temporal lobe epilepsy; positron emission tomography; metabolic connectivity; memory |
| Attached files | |
| Description | Memory decline after anterior mesial temporal resection in mesial temporal lobe epilepsy with hippocampal sclerosis (MTLE-HS) remains incompletely explained by structural and clinical factors, suggesting that networks may contribute to cognitive vulnerability. We investigated whether presurgical metabolic connectivity (MC), derived from fluorodeoxyglucose positron emission tomography (FDG-PET) and reflecting metabolic network organization, is associated with memory performance and postsurgical cognitive outcome in MTLE-HS. We studied 83 patients with MTLE-HS who underwent anterior mesial temporal resection. All had interictal FDG-PET and pre- and postsurgical neuropsychological testing. MC was computed using a normative connectivity framework and examined in relation to memory performance, analysed separately for left and right TLE. The effect of resection extent on cognitive outcome was also assessed. Presurgical analyses revealed distinct lateralized patterns. In right temporal lobe epilepsy (RTLE), immediate visual memory correlated positively with MC across widespread fronto-cingulate-temporal networks, frequently including the lesional hippocampus. In left temporal lobe epilepsy (LTLE), immediate verbal memory showed only negative correlations with extratemporal MC. In RTLE, presurgical MC also correlated with postsurgical change in immediate visual memory. Predominantly negative associations involved ipsilateral hippocampal-cingulate and fronto-temporal pathways, whereas positive correlations were confined to contralateral posterior cingulate connections with frontal and temporal regions. In LTLE, no significant correlations with postsurgical memory change were identified. Extent of resection did not correlate with cognitive outcome. MC was significantly associated with memory performance in MTLE-HS, showing clear lateralized differences. In RTLE, presurgical MC was associated with postsurgical outcome: ipsilateral hippocampal-cingulate connectivity was linked to decline, whereas contralateral posterior cingulate pathways supported preservation. No such associations were found in LTLE, consistent with more constrained network reorganization within left-lateralized verbal memory systems. These effects were independent of resection extent, indicating that memory prognosis after AMTR in RTLE may depend on network-level mechanisms and that FDG-PET-derived MC could serve as a biomarker for cognitive risk assessment. |
| Related projects: |