Abstracts

Rationale: One third of localization-related epilepsy patients have normal-appearing clinical MRIs. These patients are 2-3 times more likely to have worse outcomes after surgery compared to those with lesions identified on MRI, perhaps due to an inability to accurately identify the epileptogenic focus. We previously demonstrated that the noninvasive, functional, single-slice Glutamate Chemical Exchange Saturation Transfer (GluCEST) imaging technique can lateralize and localize epileptogenic regions in the hippocampus on 7T MRI in MRI-negative patients, indicating increased GluCEST signal in the ipsilateral total hippocampus and ipsilateral hippocampal head. Here, we present confirmatory findings using the novel CEST imaging technique capable of volumetric imaging (Krishnamoorthy et al., 2016). Methods: Imaging was performed on a Siemens 7T MRI scanner equipped with a volume transmit/32-channel receiver array head coil. The MRI protocol included: 1) A localizer scan, 2) T1-w MPRAGE scan (TR/TI/TE = 2800/1500/4.4 ms, FA=70, GRAPPA=2,  170 sagittal slices, voxel size 0.8 mm3), 3) T2w-MRI for subfield segmentation (TR/TE = 3000/388 ms; Matrix = 448×428; in-plane resolution = 0.4 mm x 0.4 mm; slice thickness = 1.0 mm; 224 oblique coronal slices perpendicular to the hippocampal long axis), 4) B0 field-map , 5) B1 field map and 5) 3D GluCEST scan (TR/TE = 5.9/2.83 ms; Matrix = 240x192; in-plane resolution = 1 mm2; slice thickness = 1.0 mm; GRAPPA=2; 60 axial slices). Raw CEST images were acquired by varying saturation offset frequencies from ± 1.8 to ± 4.2 p.p.m. with a step size of 0.3 p.p.m and saturation B1rms values of 3 mT, 1.5 mT and 0.75 mT. B0 inhomogeneity was less than 1.2 p.p.m with global shimming while relative B1 inhomogeneity varied from 0.5 to 1.4. To acquire whole brain B0 and B1 corrected GluCEST, a segmented elliptical center encoding strategy was used for the phase encode(ky) – slice encode(kz) plane, with ASHS segmentation to measure hippocampal GluCEST. The total scan session took about 1 hr. Results: In 4 nonlesional (MRI-negative) left-sided TLE patients, we found increased GluCEST signal in the ipsilateral total hippocampus relative to the contralateral total hippocampus with statistical significance (p=0.048, 1-tailed two-sample t-test) using multi-slice GluCEST. Although parameters are still being optimized for high fidelity acquisitions, these results are consistent with our previous study of single-slice GluCEST in MRI-negative TLE patients.  Conclusions: As we recruit more patients, we will further investigate GluCEST signal in hippocampal subfields and correlate these findings to electrophysiological and clinical outcomes, eventually leading to a superior understanding of epilepsy excitatory networks and improvement of surgical resection outcomes. Funding: Research reported in this publication was supported by the National Institute of Health (NIH) National Institute of Neurological Disorders and Stroke (NINDS) under grant number K23-NS073801 (to K.A.D), National Institute of Biomedical Imaging and Bioengineering (NIBIB) supported Biomedical Technology Research Center (P41-EB015893), R01-NS087516, and the National Center for Advancing Translational Sciences of the NIH under award number TL1TR001880.