Introduction: Drug-resistant epilepsy (DRE) represents a special subset of epilepsy patients who require appropriate surgical management with a goal to achieve sustained seizure freedom¹. Precise localization of the epileptogenic focus is crucial to attain adequate post-surgery seizure control, spare the non-epileptogenic brain tissue and thereby minimize postoperative neurological deficits². MRI is routinely acquired in epileptic patients to identify the epileptogenic zone and guide surgical resection³. However, the patients who do not have a focal abnormality on MRI are less likely to undergo surgery following presurgical evaluation. In such “non-lesional” refractory epilepsies, further investigation with advanced neuroimaging techniques, including metabolic and perfusion imaging, may help to identify the previously non-visualized focal brain abnormalities, though invasive EEG monitoring remains the gold standard investigation in such cases⁴. In the present study, we aim to evaluate the utility of Pseudo-continuous ASL (pCASL) perfusion-MRI in the peri-ictal period immediately after the video-EEG session as an add-on investigation to localize the seizure focus in patients with DRE.
Methods: In this study, we included patients (n=29) with drug-resistant localization related epilepsy⁵ with positive video-EEG findings undergoing presurgical evaluation (Both temporal and neocortical causes of epilepsy). MRI: MRI were obtained on 3T Philips Achieva using 32-channel head-coil. sMRI and pCASL data were acquired immediate peri-ictal period. A High-resolution 3D-TFE T1W-images were acquired (TR/TE=8.2/3.7ms, Flip-angle=8⁰, and spatial-resolution=1x1x1mm).
ASL-perfusion data was also acquired with following parameters: TR/TE=4400/14ms; dynamic-scans=40, slice-thickness=7mm, Post-label delay and labelling duration=1800sm; respectively. The video-EEG was also recorded using 10-20 channel EEG (Galileo NT, EB Neuro, Italy) machine.
PET: FDG-PET was acquired using a PET-MRI scanner (Biograph, Siemens; Erlangen, Germany). PET images were acquired in 3D-mode after intravenous injection of 5 MBq/kg of 18-fluor-FDG.
Data Analysis: Hyper-perfusion and hypo-perfusion pattern on ASL were designated based on visual analysis. Similarly, PET scans were analysed for the presence of hypo- or hyper-metabolism. To assess the concordance between ASL and other tests, the kappa coefficient was calculated. The degree of concordance was determined as follows:
(1) Discordant: Location of abnormality on ASL and the other test (e.g., video-EEG, sMRI, and PET-MR) was not matching or ASL was normal.
(2) Partially Concordant: CBF alteration on ASL was diffuse or larger as compared with any of the modality and localization overlapped partially.
(3) Concordant: Location of CBF abnormality on ASL was identical to those detected by other investigations at the sub-lobar level. Perfusion abnormalities in ASL determined after visual inspection were compared with conventional MRI findings, PET findings, and video-EEG data. The pathological area determined by the different methods were localized to various anatomical cortical regions in a blinded fashion.
Result: Of 29 patients [15M/14F; mean age=20.1±8years], 25 patients had lesions on structural MRI. ASL showed perfusion changes in 27 patients (hyper-perfusion=10, hypo-perfusion=17, and normal=2) with mean seizure to ASL scan duration of 6-hours (range 1-20hours). ASL had good concordance with structural MRI (20/29, 68.9% k= 0.638), and moderate concordance with video-EEG (k= 0.569).
MRI Negative cases: All the four patients with normal MRI, labeled as cryptogenic, showed perfusion abnormalities on ASL (3 patients showed hyper-perfusion and 1 had hypo-perfusion). The final diagnosis in these cases were established by the consensus of the epilepsy team based on clinical history, neuropsychological assessment and electrophysiological findings from video-EEG monitoring. The perfusion abnormality in ASL was concordant with the final consensus diagnosis in all the four cases. Histopathological diagnosis was present in one case (Fig. 1), which revealed left parieto-occipital focal cortical dysplasia.
PET and ASL: FDG-PET data were available for 15 patients. It showed hypo-metabolism in 14 of these patients and was normal in one case. Complete concordance between ASL and FDG-PET was found in 10 patients, partial concordance in 2 patients, and 3 cases were non-concordant. One case where PET was studied as normal, showed hyper-perfusion in ASL imaging, with epileptogenic site correlating with the Video-EEG (k= 0.661).
Discussion: In this study peri-ictal ASL showed good concordance with the localization results obtained with clinical, electrophysiological and other imaging methods. In addition, ASL was able to show focal perfusion abnormalities in 4 patients who were otherwise negative on conventional MRI and electrophysiology correlated with the focus of abnormality shown by ASL in all the cases. The perfusion changes during the inter-ictal/-ictal/immediate post-ictal stage reflect changes in the neuronal activity. These events of perfusion changes occur typically over a period of minutes to several hours6. Thus, the perfusion changes in the epileptogenic zone are a dynamic phenomenon with a possible transformation from initial hyper-perfusion to hypo-perfusion during later stages. Hyper-perfusion is noted due to an elevation of the local cerebral metabolic rate for oxygen and glucose with a consequent increase in CBF in local tissue. Similarly; post-ictal hypo-perfusion is presumed to be the result of post-ictal exhaustion/ inhibition or a steal phenomenon associated with the propagation of -ictal discharges to adjacent brain areas.
Conclusion: This study indicates that ASL perfusion could be incorporated into the presurgical evaluation of all patients with localization-related epilepsy as it may reveal seizure-induced alteration in brain perfusion and may help to identify the location and extent of the epileptogenic zone.

Acknowledgements

No acknowledgement found.