1. Correlation of Multi-modality Imaging Evaluation of Seizures in Children Control#: 2745 ASNR 2016 Annual Meeting Children’s Healthcare Of Atlanta And Emory University School Of Medicine Department Of Radiology And Imaging Sciences A ALAZRAKI, B SOARES, N DESAI, S PALASIS

2. Nothing To Disclose

3. Background About 1% of children (0-17 y) in the United States have epilepsy 2/3-3/4 of children with epilepsy will become seizure free with anti-convulsant medication The remaining will become intractable

4. Background Epilepsy can be: – Non-Lesional: A structural abnormality may not exist A lesion exists, but not detected by standard anatomic imaging – Lesional If a structural lesion can be identified, the odds of becoming seizure-free are at least 2.5 times higher than non-lesional epilpesy Normal MRI DNET

5. STRUCTURAL/MRI

6. MRI High resolution MRI at 3.0T using 32 channel and 64 channel head coils provides exquisite anatomic detail Double inversion recovery (DIR) provides increased conspicuity of focal cortical dysplasia, type II, in this child with frontal lobe epilepsy due to complementary suppression of both cerebrospinal fluid and normal white matter signal. DIR FLAIR

7. MRI High resolution oblique coronal T1 inversion recovery (IR) provide anatomic images of the hippocampal architecture In this child with intractable epilepsy, gliosis in Sommer’s sector (CA1) is clearly demonstrated on MRI; this area is the most vulnerable to seizure induced hippocampal injury T1 IR COR thedoctorweighsin.com https://quizlet.com

8. FUNCTIONAL/NUCLEAR IMAGING

9. Nuclear Medicine Techniques Functional/nuclear medicine localization techniques can confirm a suspected lesion identified on MRI or identify a seizure focus in non-lesional epilepsy These include: FDG-PET, SPECT, SISCOM and ISAS

10. 18F-FDG PET/CT Fluorodeoxyglucose (FDG): glucose analog assesses metabolism of tissue; Glucose is the major source of energy for neurons PET is acquired with CT for attenuation correction and localization Hypometabolism in the epileptogenic focus is presumed to represent synaptic changes that occur from a seizure When MRI is nonlocalizing or discordant with EEG, PET may show a discrete area of hypometabolism that identifies the epileptogenic focus

11. 18F-FDG PET/CT FDG image is automatically fused to the attenuation correction (low dose) CT images for localization

12. PET fused to MRI The FDG PET data can then be fused to MRI for more accurate localization in 3 planes

13. The influence of brain maturation in PET interpretation All regions show an increase in SUVmax with age, but the rate of increase is lowest in the parietal and occipital lobes At age 5, the parietal uptake is higher than both frontal and thalami At age 15, frontal uptake is the highest and parietal uptake is now less than thalami London, et al. Eur J Nuc Med Mol Imaging, 2014 Regional SUVmax by age:

14. Perfusion imaging: 99m-Tc-HMPAO Ictal SPECT excellent sensitivity to localize Cerebral blood flow (CBF) increases during seizure at the site of onset CBF decreases both during and following seizure at multiple locations (“postictal switch”) The goal of ictal perfusion imaging is to capture the seizure-induced hyperperfusion prior to the ictal switch

15. Perfusion imaging Injection time is critical (≤40 sec) Late injection- false localization/ lateralization Requires inpatient admission and NM technologist to sit bedside with EEG tech during continuous video EEG monitoring because of the narrow window of opportunity to capture the seizure Interpretation requires comparison to the patient’s own interictal scan as a baseline Epileptologie 2004; 21: 105–108

16. Interictal and Ictal SPECT Identification of the seizure focus by visual analysis can be difficult Interictal SPECTIctal SPECT

17. Subtraction imaging of ictal-interictal SPECT co-registered to MRI (SISCOM) This technique allows easy identification of maximal hyperperfusion by subtracting the ictal from the interictal SPECT images It allows quantitative extrapolation of areas of hyperperfusion Subtraction map is laid on top of volumetric MRI data set for localization

18. SISCOM Easier identification of the seizure focus Superimposing the subtracted data set on the MRI allows more accurate localization SISCOM Interictal SPECTIctal SPECT

19. Ictal-Interictal SPECT analysis by statistical parametric mapping (ISAS) Generates hyperperfusion and hypoperfusion maps comparing the patients perfusion to a database of control patients Generates statistical data that identify the most significant foci of perfusion

20. ISAS ISAS images are displayed in composite layout in 3 planes The most significant cluster values are identified and mapped to the corresponding sites of perfusion The most significant cluster corresponds to hyperperfusion (red/yellow) in the left temporal lobe, which is concordant with the PET imaging

21. Purpose The purpose of this study was to evaluate these techniques through our comprehensive epilepsy program We present several interesting cases utilizing multimodality approach to seizure localization We highlight the added value of a collaborative approach to the patient

22. Approach/Methods Prospective IRB approval to perform advanced processing analysis on clinically performed diagnostic studies Chart review, image review, post-processing interpretation was performed

23. Results 239 Brain PET CT examinations were performed between April 2008 and April 2016 Of these, 209 were performed for seizure localization 29 also underwent ictal-interictal SPECT perfusion imaging

24. CASE EXAMPLES Discussion

25. Case 1: MRI EEG localizes seizures to left frontotemporal High resolution MRI reveals a left temporal lesion Axial T2Cor 3D FLAIRSag 3D FLAIR

26. Case 1: SPECT MRI reported dysembryoplastic neuroepithelial tumor (DNET) in left temporal lobe Assessment prior to surgical planning Ictal SPECTInterictal SPECTSISCOM

27. Case 1: PET Interictal FDG PET localizes to left temporal lobe Interictal PETPET fused to MRI

28. Case 1:ISAS ISAS quantifies the most significant areas of hyperperfusion (red/yellow) and hypoperfusion (blue) The crosshairs can be placed to identify these significant clusters

29. Case 1: Epileptogenic Zone Note the discrepancy between the anatomic lesion on MRI and the more extensive epileptogenic zone demonstrated on functional nuclear medicine imaging Axial T2PET fused to MRI SISCOM ISAS

30. Case 2: Cortical Dysplasia COR T2Axial T2 Intractable epilepsy; left frontal semiology MRI shows cortical dysplasia in left insula

31. Case 2: SPECT Visual analysis shows an area of asymmetric ictal increased perfusion in the left frontal region Injection time 10:35 Injection 28 s after clinical onset, 1 sec after onset Ictal SPECTInterictal SPECT

32. Case 2: PET Interictal FDG PET confirms left frontal hypometabolism Interictal PETPET fused to CT

33. Case 2: SISCOM and PET Concordance of imaging findings is tantamount to providing strong surgical recommendations SISCOMPET fused to MRI

34. Case 3: Temporal Lobe Seizures Intractable epilepsy; EEG localizes to left temporal, with non-lesional MRI previously Visual analysis suggests asymmetric ictal perfusion in left temporal lobe Interictal SPECTIctal SPECT

35. Case 3: PET Positive PET with localized hypometabolism in the left temporal lobe forced a second look at the MRI Interictal PETAxial PET fused to MRICoronal PET fused to MRI

36. SISCOM/MRI On re-examination, a tiny temporal lobe encephalocele was identified

37. Temporal lobe encephalocele Etiology is poorly understood Postulated that early embryogenic failure leads to a nonossified membranous ala Uncommonly associated with NF1 (cranioorbital-temporal neurofibromatosis) Present with dizziness, recurrent meningitis or otitis, conductive hearing loss or medically intractable epilepsy Neurosurg. Focus / Volume 25 / December 2008

38. Case 4: MRI EEG lateralizes to the right hemisphere, near the motor cortex Focal cortical dysplasia of Taylor posterior right frontal lobe Axial 3D DIRAxial 3D FLAIR

39. Case 4: Ictal PET Injection: 8:52am Sz #1: 8:54 lasted 45 seconds Sz #2: 9:00 lasted 1 min 20 seconds Both seizures manifest with left arm extension PETSag PET fused to MRI

40. Case 4: SPECT Strip grid and depth electrodes placed Linear defect on perfusion imaging from depth electrode Ictal SPECT Interictal SPECTCT with electrode

41. Case 4: SISCOM The most significant area of hyperperfusion corresponds to both the lesion and the area of PET avidity. Smaller areas of spurious perfusion are present Axial SISCOMSag SISCOM

42. Case 5: Non-lesional epilepsy Intractable epilepsy with no MRI lesion identified Ictal-Interictal SPECT was performed, visual analysis suggests area of right frontal increased perfusion Ictal SPECT Interictal SPECTAx T2 MRI

43. SISCOM/PET concordance SISCOM makes the abnormality easy to localize PET hypometabolism corresponds to the SPECT Interictal PETSISCOM

44. Summary/Conclusion Advanced imaging techniques allow better and more accurate visualization of seizure foci Correlative imaging studies add confidence and present compelling data for the care team to pursue surgical approach

45. References Soares BP, et al. Utility of double inversion recovery MRI in pediatric epilepsy, Br J Radiol, 89(1057), 2016. Shandal, V, et al. Long-Term Outcome in Children With Intractable Epilepsy Showing Bilateral Diffuse Cortical Glucose Hypometabolism Pattern on Positron Emission Tomography, J of Child Neuro 27(1): 39- 45, 2012. Jayakar, P, et al. Diagnostic test utilization in evaluation for resective epilepsy surgery in children, Epilepsia 55(4):507-518, 2014. Fernandez, S, et al. PET/MRI and PET/MRI/SISCOM coregistration in the presurgical evaluation of refractory focal epilepsy, Epilepsy Res. Mar; 111:1-9, 2015 Kumar, A, et al. The Role of Radionuclide Imaging in Epilepsy, Part 1: Sporadic Temporal and Extratemporal Lobe Epilepsy, J Nucl Med. Oct; 54(10):1775-81, 2013. Suarez-Pinera, M, et al. Perfusion SPECT, SISCOM and PET 18F-FDG in the assessment of drug-refractory epilepsy patients candidates for epilepsy surgery, Rev Esp Med Nucl Imagen Mol, Nov-Dec; 34(6):350-357, 2015. Toledano, R, et al. Small Temporal pole Encephalocele: A hidden cause of “normal” MRI temporal lobe epielpsy, Epilepsia, **(*):1-11, 2016.