Abstracts

Rationale: High resolution (3-4 mm voxel) single-photon emission computed tomography (SPECT) has been recently developed using a 72-detector focused collimator ring scanner system (inSpira, Samsung-NeuroLogica Corp). The goal of this study was to determine if the extent to which transient hyperperfusion-related ictogenic grey matter visualized in presurgical candidates with refractory extratemporal and temporal lobe epilepsies correlated with long-term post-resective outcomes over time. All subjects were scanned following the same partial-onset seizure with both the 72-detector ring scanner system and a conventional 2-headed detector SPECT scanner (Siemens). Methods: Fifteen subjects with medically-refractory extratemporal and temporal ictogenic sources were enrolled in this investigator-initiated IRB-approved study prior to resective surgery. Injection of SPECT radiotracer, using either ^99mTc-ECD (Neurolite) or ^99mTc-HMPAO (Ceretec), was successfully completed at the ictal onset. A baseline SPECT was acquired for each subject following an 18-24 hour seizure-free period. Ratio ictal SPECT co-registered to MRI (RISCOM) was completed for each subject acquired with the inSpira scanner system and Siemens 2-detector SPECT scanner. The data were normalized and processed using custom software utilities. These data were co-registered to each subject's gapless SPGR MRI study using the ITK co-registration algorithm and visualized (MIM Software). Subtraction ictal SPECT co-registered to MRI (SISCOM) was also performed for each subject and compared with RISCOM datasets using Analyze v10 (AnalyzeDirect). For each subject, RISCOM and SISCOM analyses of the extratemporal and temporal ictogenic onset regions were compared across SPECT scanners. These data were validated utilizing chronic electrocorticography capturing stereotypic seizures prior to resective surgery. Resolution of SPECT systems were also visually compared with interictal PET image sets. Results: RISCOM datasets using the 72-detector system demonstrated a concordant overlap of transient regions of hyperperfusion seen with the SISCOM data obtained with the 2-detector SPECT scanner. However, the InSpira RISCOM data demonstrated regions of hyperperfusion not seen with the SISCOM datasets. For 13 of the 15 patients, these new regions of hyperperfusion were concordant with chronic electrocorticography and included in the volume of resection. All 15 patients demonstrated an Engel's class I-II outcome after a mean of 12-48 months post-resective surgery. Conclusions: RISCOM image processing using the 72-detector ring collimator SPECT scanner facilitated identification of the ictal onset in extratemporal and temporal epileptogenic networks not visualized using conventional 2-detector SPECT scanner technology. Application of this novel SPECT technology can facilitate better visualizing the maximal extent of ictogenic tissue necessary for strategic placement of chronic intracranial electrodes. This technology and complementary post-processing strategy are associated with improved long-term post-resective outcomes. Funding: Departmental