Association of Cortical Lesions With Regional Glutamate, GABA, N-Acetylaspartate, and Myoinositol Levels in Patients With Multiple Sclerosis

Abstract

Background and objectives: Cortical lesions contribute to disability in multiple sclerosis (MS), but their impact on regional neurotransmitter levels remains to be clarified. We tested the hypothesis that cortical lesions are associated with regional glutamate and gamma-aminobutyric acid (GABA) concentrations within the affected cortical region.

Methods: In this cross-sectional study, we used structural 7T MRI to segment cortical lesions and 7T proton MR-spectroscopy of the bilateral sensorimotor hand areas to quantify regional GABA, glutamate, N-acetylaspartate, and myoinositol concentrations in patients with MS (inclusion criteria: diagnosis of relapsing-remitting [RR] or secondary progressive MS [SPMS]; age 18-80 years) and age and sex-matched healthy controls. Data were collected at a single center between August 2018 and September 2020. Linear mixed-effects models were used to test for associations between metabolite concentrations and cortical lesion volumes within the same MR-spectroscopy voxel.

Results: Forty-seven patients with MS (34 RRMS, 13 SPMS; 45.1 ± 12.5 years; 31 women) and 23 healthy controls (44.4 ± 13 years, 15 women) were studied. In patients, higher regional glutamate and lower regional GABA concentrations were associated with larger cortical lesion volume within the MR-spectroscopy voxel [glutamate: 0.61 (95% CI 0.19-1.03) log(mm³), p = 0.005, GABA: -0.71 (-1.24 to -0.18) log(mm³), p = 0.01]. In addition, lower N-acetylaspartate levels [-0.37 (-0.67 to -0.07) log(mm³), p = 0.016] and higher myoinositol levels [0.48 (0.03-0.93) log(mm³), p = 0.037] were associated with a larger regional cortical lesion volume. Furthermore, glutamate concentrations were reduced in patients with SPMS compared with healthy participants [-0.75 (-1.3 to -0.19) mM, p = 0.005] and patients with RRMS [-0.55 (-1.07 to -0.02) mM, p = 0.04]. N-acetylaspartate levels were lower in both patients with RRMS [-0.81 (-1.39 to -0.24) mM, p = 0.003] and SPMS [-1.31 (-2.07 to -0.54) mM, p < 0.001] when compared with healthy controls. Creatine-normalized N-acetylaspartate levels were associated with performance in the 9-hole peg test of the contralateral hand [-0.004 (-0.007 to -0.002) log(s), p = 0.002], and reduced mean creatine-normalized glutamate was associated with increased Expanded Disability Status Scale (R = -0.39, p = 0.02).

Discussion: Cortical lesions are associated with local increases in glutamate and a reduction in GABA concentration within the lesional or perilesional tissue. Further studies are needed to investigate the causal relationship between cortical lesions and changes in neurotransmitter concentrations.

Figure 1. Voxel Placement and Mean Spectra

(A) Example MRS voxel placement (red) in the sensorimotor hand knob, corresponding to the primary sensorimotor hand area (SM1-HAND). Segmented cortical lesions are shown in blue and white matter lesions in white. (B) Example of a leukocortical lesion located inside the MRS voxel field of view. The lesion is shown in the sagittal plane on FLAIR (top), MPRAGE (middle), and T₂-weighted (bottom) sequences. (C) Average spectra (black) obtained with ¹H-MRS at 7T. The gray areas represent ± SD for participant groups. Average LCModel fit results (red) and average residuals (black, bottom) are also shown. HCs = healthy controls; ppm = parts per million; RRMS = relapsing-remitting multiple sclerosis; SPMS = secondary progressive multiple sclerosis.

Figure 2. Flowchart of Study Participants and Number of Data Sets Included in the Final Analysis

HC = healthy control.

Figure 3. ¹H-MRS Metabolite Concentrations in SM1-HAND

Box, point, and density plots of glutamate (A), GABA (B), N-acetylaspartate (C), and myo-inositol (D) between healthy control participants, patients with RRMS, and patients with SPMS. Box plots include median and interquartile range as horizontal lines and the mean as a black dot. Whiskers indicate the 5th and 95th percentiles. Individual datapoints are plotted over the box plot, with low opacity representing the dominant hand and high opacity the nondominant hand. HC = healthy control; mIns = myoinositol; NAA = N-acetylaspartate; ns = not significant; RRMS = relapsing-remitting multiple sclerosis; SPMS = secondary progressive multiple sclerosis. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 4. Associations Between Lesion Volume and Metabolite Concentrations in the ¹H-MRS Voxels Covering the Right and Left SM1-HAND

Standardized beta-coefficients from the mixed linear models with (A) cortical lesion volume, (B) white matter lesion volume, and (C) total lesion volume within the ¹H-MRS voxel as the dependent variable. MRS = magnetic resonance spectroscopy; NAA = N-acetylaspartate; ns = not significant; SPMS = secondary progressive multiple sclerosis. *p < 0.05, **p < 0.01, ***p < 0.001.

Figure 5. Association Between 9-HPT Performance and NAA/tCr Concentration of the Contralateral Hemisphere

The plot shows the marginal effects and partial residuals from the mixed linear model fit. 9-HPT = 9-hole peg test; HC = healthy control; NAA = n-acetylaspartate; RRMS = relapsing-remitting multiple sclerosis; SPMS = secondary progressive multiple sclerosis; tCr = total creatine.