Abstracts

Rationale: Mesial temporal structures in TLE are classically characterized by low metabolism as indicated by [18F]FDG-PET. Although the mechanism underlying hypometabolism is not completely understood, changes in hippocampal glutamatergic neurotransmission might be related to the low energy budget found in the epileptogenic tissue. Post synaptic metabotropic glutamate receptor type 5 (mGLUR5) availability is reduced in the epileptogenic hippocampus (as measured by [11C]ABP688-PET). Here, we aim to test the hypothesis whether glucose metabolism determined by [18F]FDG-PET is linked to mGLUR5 availability measured by [11C]ABP688-PET. Methods: We evaluated clinical [18F]FDG scans in 10 TLE patients who underwent a [11C]ABP688 scan (7 right, 3 left, lateralized according to their ictal EEG). [18F]FDG scans consisted in a 15min standard static acquisition followed [18F]FDG IV injection (4.5 0.58mCi). [11C]ABP688 scans consisted of a 1hour dynamic acquisition followed [11C]ABP688 injection (9.6 0.53mCi). Images were reconstructed using filter-back projection, and volumes were coregistered to each patient s MRI. ABP688-BP maps were generated using a simplified reference tissue method with the cerebellum as a reference region. FDG-uptake maps were normalized by the uptake at the pons. MRIs and maps were resampled to the standard space. Volumes of interest comprising the hippocampi (total structure, head and body) were manually drawn in each patient s MRI and used for extraction of mean FDG-uptake/ABP688-BP. We used paired samples 2-tailed t-test to determine differences between epileptogenic and contralateral structures, and 2-tailed Pearson correlation for FDG-uptake and ABP688-BP comparisons. Significance level was set at p=0.01. Results: The epileptogenic hippocampi showed ABP688-BP of 79.23 5.7 (total), 76.17 7.7 (head) and 79.25 7.2 (body), whereas FDG-uptake was 0.6 0.03 (total), 0.58 0.03 (head) and 0.61 0.03 (body). At the contralateral side, hippocampal ABP688-BP was 105.62 6.54 (total), 113.35 6.5 (head) and 104.26 8.01 (body), whereas FDG-uptake was 0.64 0.02 (total), 0.63 0.03 (head) and 0.65 0.02 (body). Epileptogenic hippocampi showed lower ABP688-BP as compared to contralateral side (df=9, t=-3.24, p=0.01), with higher difference identified in the head (df=9, t=-4.49, p=0.002). ABP688-BP was also lower at the body, but not reaching significance. No differences in hippocampal FDG-uptake were observed. Hippocampal FDG-uptake was correlated with ABP688-BP in the epileptogenic side (total R=0.895/p=0.0005; head R=0.789/p=0.007; body R=0.815/p=0.004) but not in the contralateral structure. Conclusions: Our data strongly suggests a link between mGLUR5 availability and tissue metabolic processes, in the epileptogenic hippocampus only. Moreover, ABP688-BP seems to be more sensitive than FDG-uptake to identify the epileptogenic hippocampus in TLE. Our findings support a role of mGLUR5 in the mechanisms of epileptogenesis and links glutamatergic dysfunction to hypometabolism in the hyperexcitable hippocampus. Support: Savoy Foundation