Abstracts

Rationale: Brain (¹⁸F-FDG) PET imaging measures the glucose metabolism to map the neuronal functional status. In contrast to static PET localizing the epileptic focus by presenting focal hypometabolism, dynamic PET catches the volumetric radioactivity of the brain in serial times. It provides more information related to glucose metabolism rate and vascular heterogeneity. In this study, we explored the dynamic metabolic alterations of hippocampus and thalamus on the ipsilateral and contralateral sides in temporal lobe epilepsy (TLE).

Methods: We enrolled TLE patients from Taichung Veterans General Hospital from Dec, 2022 to Dec, 2023, who received presurgical evaluation, including video-electroencephalograph, high resolution brain magnetic resonance imaging (MRI), neuropsychological assessment, and dynamic brain PET scan. The PET imaging was collected for 30 minutes, and then divided into 6 frames to the successive 5-minute segment images. PET images were registered to the MRI images based on the normalized mutual information method, and then partial volume correction. The average intensity of whole brain gray matter was used as the reference to normalize each voxel, generating the standard uptake values (SUVs). We calculated and compared the coefficient of variation (CV) of SUVs of bilateral hippocampi and thalami within 6 frames in both TLE. Spearman correlation is used to study metabolic relationship between hippocampus and thalamus on the same side. The relationship of SUV oscillations to the five main intelligence quotient (IQ) scores [full-scale IQ (FIQ); verbal IQ (VIQ); Performance IQ (PIQ); Working Memory Index (WMI); Processing Speed Index (PSI)] was studied using partial Spearman correlation.

Results: Totally, there were 14 left (L)-TLE and 12 right (R)-TLE enrolled. There were no statistical difference of CV values comparing SUVs of hippocampi and thalami with the other side in L-TLE (Table). R-TLE revealed more prominent metabolic oscillations of the ipsilateral than the contralateral hippocampus (p = 0.019). The metabolic alterations of hippocampus and the thalamus on the epileptic side seemed to present more correlated with each other, especially L-TLE (1^st, 4^th,and 6^th frames, p = 0.05, 0.03, 0.03,). In L-TLE, FIQ was negatively related to contralateral CV of hippocampus (p = 0.015) and thalamus (p = 0.008). VIQ was negatively associated with CV of contralateral thalamus (p = 0.015), while PIQ was negatively associated with ipsilateral thalamus (p = 0.044). The R-TLE did not show the relationship between the metabolic oscillations and the cognitive functions.

Conclusions: Dynamic PET imaging not only reveals the different metabolic oscillations of the two hubs of TLE, it also characterizes the cognitive functions in different lateralization of epilepsy focus. Hippocampus and thalamus revealed stronger metabolic connectivity on the ipsilateral side of L-TLE, and their fluctuations of SUVs were more related to cognitive functions. Ipsilateral hippocampus revealed most significantly metabolic oscillations than the left side in R-TLE.

Funding: Te study was fnancially supported of the Higher Education Sprout Project by the Ministry of Education (MOE)
in Taiwan.