Abstracts

Rationale: Glutamate plays a critical role in the initiation and spread of seizure activity and has been implicated in the occurrence of seizures in stroke models. We hypothesize that the development of post-stroke epilepsy may be related to acute/subacute changes in cerebral glutamate level during stroke. This study aimed to establish the feasibility of measuring glutamate levels in the infarcted and non-infarcted brain tissues by glutamate chemical exchange saturation transfer (GluCEST) and magnetic resonance spectroscopy (MRS) using 7T MRI. Methods: 7T MRI (Siemens Magnetom) was performed within 14 days of stroke onset. T1-weighted anatomical image and diffusion weighted imaging (DWI) was performed. Single voxel (STEAM) MRS (TE = 6 ms) was performed within the region of diffusion restriction in the axial slice with the largest DWI lesion and a mirror reference voxel was placed in the contralateral hemisphere. The same axial slice was used for the GluCEST acquisition slab. The GluCEST images were coregistered with DWI and a region of interest drawn around the area of infarction, from which GluCEST contrast was calculated. Water-normalised metabolite concentrations and GluCEST contrast were compared between the region of infarction and the contralateral hemisphere across the study population. Patients are followed up for occurrence of post-stroke seizures. Results: Fourteen adults with cortical ischemic stroke were recruited (Table 1). 6 patients had MRS only, 4 had GluCEST only, and 4 had both MRS and GluCEST, at a median of 4 days (range 48 hours to 14 days) from stroke onset. On MRS, glutamate, myo-inositol, glutathione and N-acetyl aspartate concentrations were lower and lactate concentration higher in the region of infarction compared with the contralateral hemisphere (p = 0.001 for all 3 comparisons). There were no statistically significant differences in the concentrations of glutamine and gamma-aminobutyric acid.GluCEST contrast also trended lower in the region of infarction than the contralateral hemisphere, although the difference was not statistically significant across the population (p = 0.22). Of the 8 patients, compared to the contralateral side, ipsilateral CEST contrast was numerically lower in 5 patients, higher in 1, and similar in 3. For the 4 patients who had both MRS and GluCEST there was a strong correlation between the MRS glutamate concentration (ipsilateral divided by contralateral) and the GluCEST ratio (R2 = 0.81) although this wasn’t statistically significant (p = 0.10). Of the patients with bilaterally similar CEST contrast, all scans were acquired more than one week after stroke onset. The patient with increased CEST contrast had the most severe stroke clinically (NIHSS 19 at onset) with a large final infarct volume on DWI (see Figure 1). Overall however the GluCEST ratio did not appear to correlate with timing of MRI in relation to stroke onset. No patient has thus far had seizures during the follow-up period (median follow-up 14 months, range 10-19 months). Conclusions: This study demonstrates the feasibility of performing 7T MRI for quantification of cerebral glutamate levels in the setting of acute stroke. Compared with the contralateral hemisphere, glutamate level appeared lower in the region of infarction when measured by MRS, while GluCEST contrast ranged from decreased to increased in the ipsilateral hemisphere. 7T MRI may be employed in future studies to determine whether cerebral glutamate level might be a biomarker of post-stroke epilepsy. Funding: This research was funded by a grant from the Brain Foundation, Australia. John-Paul Nicolo is supported by an Australian Government Research Training Program Scholarship.