Abstracts

Rationale: Single photon emission computed tomography (SPECT) is used to localize the ictal onset zone in the presurgical evaluation of patients with refractory focal-onset epilepsy. This technology commonly employs a single 2- or 3- headed detector (7-10 mm voxel resolution). However, a high resolution (3-4 mm voxel) SPECT recently has been developed using a 72-detector focused collimator ring scanner system (InSpira^TM, Samsung-NeuroLogica Corp). Our study intended to investigate whether this new 72-detector scanner and a complementary post-processing technique provide a more accurate localization of temporal and extratemporal ictal onset zones, and, therefore, improved post-resective clinical outcomes. Methods: Thirteen patients with refractory focal-onset epilepsy were enrolled in this study. These patients were considered suitable candidates for resective surgery. SPECT radiotracer ^99mTc-HMPAO (Ceretec^TM) was injected successfully with a mean time of 33 seconds (Range: 9-127 seconds; SD: 34.6 seconds) from clinical or electrographic seizure onset. After about a 24 hour seizure-free period, the baseline SPECT was acquired. Algdev software (Samsung-NeuroLogica Corp) was used for performing attenuation correction and processing of the raw SPECT data to generate novel ratio ictal SPECT datasets. The resulting data were converted to DICOM files by Dicomgen^TM. The processed Ratio Ictal SPECT data were Co-registered to a patient-specific gapless SPGR MRI (RISCOM) using MIM^TM Software. Well-established Subtraction Ictal SPECT Co-registered to MRI (SISCOM) was also performed for each patient acquired with the inSpira^TM scanner system and a Siemens 2-detector SPECT scanner using Analyze^TM v10. A comparison between RISCOM and SISCOM for each patient was performed focusing on the hyper-perfused areas detected by the 72-detector scanner (InSpira^TM) and 2-detector scanner (Siemens^TM), respectively. These data were validated using chronic electrocorticography (ECoG). Results: Regions of transient hyper-perfusion seen by SISCOM obtained with the 2-detector SPECT scanner overlapped with the RISCOM datasets using the 72-detector system. However, in 11 out of 13 patients, the InSpira^TM RISCOM data demonstrated regions of hyper-perfusion not seen with the SISCOM datasets. In 3 out of 13 patients, these new regions of hyper-perfusion were concordant with ECoG and influenced the surgical plan. At the time of this submission, 11 out of 13 patients were in Engel's class I-II after a mean follow up period of 23 months (Range: 4 months to 34 months SD: 10.2 months) after resection. Conclusions: A greater extent of the ictal onset zone in active epileptic circuits was visualized by RISCOM using the 72-detector ring collimator SPECT scanner compared to our Siemens^TM conventional 2-detector SPECT scanner. When compared with SISCOM, this new SPECT technology and complementary RISCOM technique can facilitate strategically placing chronic intracranial electrodes for improving post-resective outcomes.