Abstracts

Rationale: Dystonia and motor automatism are reported in up to 50-70% of patients with Temporal Lobe Epilepsy (TLE).These ictal symptoms help to lateralize ictal onset zone, however precise localization relies on a comprehensive pre-surgical evaluation. FDG-PET imaging is a reliable often the preferred imaging modality in MRI negative patients to identify the epileptogenic cortex and dysfunctional network based on the distribution of hypometabolism. In this study, we examined the correlation between the glucose hypometabolism pattern and ictal symptoms, dystonic posturing and motor automatism, in children with medically refractory TLE. Methods: Seventeen children (age:6-18years) were identified with TLE in two tertiary care centers. FDG-PET was obtained as part of the presurgical evaluation. Clinical features were reviewed, FDG-PET images were retrieved for analysis. PET scan: A bolus injection of approximately 10 mCi [18F] FDG was administered 30 minutes prior to scan, using a Siemens ECAT EXACT 47 scanner. Summed images were then preprocessed for statistical analysis. Statistical Analysis: Data analysis was performed using MATLAB 7.1 and SPM 2 (Wellcome Department of Imaging Neuroscience, University College London, UK). Voxel-based data is reported at a voxel extent threshold of 20 voxels and with p<0.001 (corrected). T-test was applied for statistical analysis to compare images of two groups, children with TLE and healthy control (n:8).Results: TLE was lateralized to the left in 7 and right in 9 children. One child had bilateral TLE.. Ictal dystonia was seen in 8 and motor automatism in 11. Six children had both ictal dystonia and motor automatism. Epilepsy surgery was performed in 12 and all reached Engel class IA-IB post surgery outcome. FDG-PET images revealed the temporal lobe hypometabolism in all children. SPM analysis demonstrated that in the presence of ictal dystonia, glucose hypometabolism was pronounced in the entire temporal lobe and expanded into insula (z:4.04), lentiform nucleus (z:4.24), and frontal lobe including precental (z:3.66) , postcentral (z:3.32), inferior frontal (z:3.79) and cingulated gyrii (z: 4.35) (p<0.001 corrected) . Similarly, with motor automatism, hypometabolism was noted in mesial temporal lobe (p<0.001 corrected), frontal lobe [precentral (z:3.99), postcentral (z:3.95), inferior frontal gyrus (z:4.54)], insula (z:4.13), lentiform nucleus (z:5.62), caudate (z:3.81) and thalamus (z:3.58) (p<0.001 corrected).Conclusions: Glucose hypometabolism correlates with the clinical semiology of children with TLE. Distribution of hypometabolism are similar in children with ictal dystonia and motor automatisms suggesting a network between mesial temporal , frontal lobe, insula and lentiform nucleus. The variation of hypometabolism pattern based on the ictal semiology reflects the difference in the expression of seizures and seizure propogation mechanisms and may not have an impact on postsurgery outcome.