Abstracts

Rationale: We employed a novel approach to examine the distribution of hippocampal injury in TLE, involving measurements of sodium concentrations in vivo using Na23 MRI. The brain maintains an exquisite balance between intracellular and extracellular sodium concentration and there are dynamic shifts of sodium ions from the extracellular to intracellular space during epileptiform discharges, seizure propagation and seizure termination.Methods: We built and optimized a dual-tuned H1/Na23 MRI birdcage RF coil to allow simultaneous imaging of Na23 and Proton and customized it for a clinical 3T MR scanner. Four calibration tubes with different sodium concentrations (100, 75, 50 and 25 mM/L) in 4% agar gel were placed in the field of view to quantifify of the sodium concentration (Fig. 1). The T2 constants of the calibration tubes were calculated with TEs = 0.2, 6.0, 11.8, and 17.6 ms. Since the phantom’s sodium concentrations are known, we used 25 and 100 mM phantoms for calibration and 50 and 75 mM phantoms as “unknowns” to determine the measurement accuracy of our technique. To demonstrate the clinical feasibility of this system, we scanned a control subject (age 44) and a patient with right temporal lobe epilepsy with hippocampal sclerosis (age 39). Scanning parameters: 3 T system with multi-nuclear option (ACHIEVA, PHILIPS), 3D radial gradient echo [repetition time (TR) = 50 ms; TEs = 0.2, 6.0, 11.8, 17.6 ms; flip angle = 90; 16 averages; field of view = 240x240 mm2; resolution = 4x4x20mm3; acquisition time = 16 min.Results: The phantom calibration showed that the calculation of total sodium concentration has an error < 5% (Actual [Na]=50mM, Calculated [Na]=51.2mM, error=2.4%; Actual [Na]=75mM, Calculated [Na]=78.5mM, error=4.7%). Figure 2 showed brain axial sodium images co-registered to T2-weighted proton images obtained in a healthy control (A) and an individual with right TLE (B). Note that we obtained sodium and T2-weighted images with the same head coil, thus minimizing alignment errors between the two scans. The sodium color maps demonstrate higher total sodium concentration (TSC, warmer color) in the bilateral hippocampus (HP) and adjacent anterior temporal regions (yellow regions) in the right TLE subject. We further quantified TSC in the amygdala, anterior and posterior hippocampus in these two individuals using the co-registered anatomical landmarks while correcting for brain volume. As seen in Fig 2C, the individual with right TLE demonstrated greater TSC in the amygdala and anterior hippocampus, most evident ipsilateral to the side of seizure onset.Conclusions: Our study demonstrated for the first time preliminary evidence that: 1) TSC is selectively increased in the epileptogenic zone and surrounding regions, and 2) TSC varies along the anterior-posterior longitudinal axis of the hippocampus in TLE. It also showed obtaining sodium MRI in a clinical setting. We will validate our findings with additional TLE patients and relate alterations in TSC with clinical variables and surgical outcome.