Abstracts

Progressive impairment of hippocampal glucose metabolism with increased epilepsy duration

Abstract number : 2.138
Submission category : 5. Neuro Imaging / 5C. Functional Imaging
Year : 2016
Submission ID : 195363
Source : www.aesnet.org
Presentation date : 12/4/2016 12:00:00 AM
Published date : Nov 21, 2016, 18:00 PM

Authors :
Shubhi Agrawal, National Institute of Neurological Disorders and Stroke; Giampiero Giovacchini, Triemli Hospital, Zurich, Switzerland; Irene Dustin, National Institute of Neurological Disorders and Stroke; John Heiss, National Institute of Neurological Di

Rationale: Several studies have looked at the relation between increasing duration of epilepsy and progressive hippocampal volume loss on magnetic resonance imaging (MRI) and hypometabolism on positron emission tomography (PET). It is unclear if the hypometabolism seen on PET is mainly due to volume loss. Methods: We used 18Fluorine 2deoxyglucose (FDG) PET and MRI to study patients with temporal lobe epilepsy as confirmed on ictal video electroencephalography (vEEG). An automated partial volume correction (PVC) algorithm was applied to the PET study based on MRI coregistration. Metabolic activity in congruent regions of interest (ROI) was used to calculate asymmetry indices as following 2*(C-I)/C+I. Regression modelling was used to compare effect of age of onset, duration of epilepsy, and post surgical outcome group (seizure free vs non seizure free) on hippocampal asymmetry index (HAI) using ANOVA. Asymmetry indices across the two post surgical groups were also compared using Mann-Whitney test Results: 38 patients (17 females) post anterior temporal lobectomy with minimum 1 year follow up were included. Mean age of onset was 14.3 years (range 1-43years) and mean duration of epilepsy prior to imaging was 19.7 years (range 2-45 years). MRI was normal in 13 patients. 23 patients had mesial temporal sclerosis, one patient had a lateral temporal lesion and one had dysgenesis of corpus callosum. PVC FDG-PET analysis showed gradually increasing asymmetry index with increased duration of epilepsy for patients who became seizure free after surgery. R2 was 0.32 and F statistic was 10.9 with a p value of .003. This trend was not seen in patients who continued to have seizures after surgery. Asymmetry indices of parahippocampal gyrus and amygdala were significantly correlated with hippocampal asymmetry index (Pearson's coefficient 0.8 and 0.6 respectively) but the trend of increasing asymmetry was not seen in the amygdala. For the overall group of patients, hippocampal asymmetry index on partial volume corrected PET showed significant correlation with epilepsy duration and post surgical seizure outcome group but not MRI findings or age of onset of epilepsy. Hippocampal asymmetry index distribution was significantly different between the two post surgical groups (p 0.003) with the seizure free group having higher values. Conclusions: This study provides proof of progressive impairment of glucose metabolism in the hippocampal cells in addition to neuronal loss with increasing duration of epilepsy. This trend was also seen in the parahippocampal gyrus but not the amygdala. Interestingly, we found that the patients with higher asymmetry indices did better after surgery. In patients with lower asymmetry index or with a negative asymmetry index, the poor outcome could be due to either an incorrect presurgical hypothesis or presence of a more diffuse pathology that lowered both the asymmetry index and odds of seizure freedom. Funding: NIH Intramural program
Neuroimaging