Introduction

γ-aminobutyric acid (GABA) and the glutamate-glutamine complex (Glx) are the primary inhibitory and excitatory neurotransmitters in the central nervous system, respectively, and their imbalance is implicated in neurodegenerative and neuropsychiatric disorders¹. In vivo measurements of GABA and Glx are prone to be affected by many factors, therefore, to investigate their alterations in the progress of diseases, it is essential to validate the reproducibility of GABA and Glx measurements in healthy populations. In the current study, aimed to evaluate the reproducibility of GABA and Glx measurements, we take advantage of proton magnetic resonance spectroscopy (1H-MRS) with spectral editing technique MEscher-Garwood Point RESolved Spectroscopy (MEGA-PRESS)² to access the concentration of GABA and Glx in the basal ganglia and midbrain regions in the same group of healthy individuals in different period.

Methods

This study was approved by the local ethics committee (No: KY2022-266) and written informed consent was obtained from all participants. 23 young healthy individuals were enrolled and underwent the first scan, among them 16 subjects underwent the second scan after 2 weeks.
All scan was conducted on a 3.0 T scanner (Elition X, Philips Healthcare) with a 32-channel head coil. A transverse T1-weighted sequence was scanned to produce the anatomical reference images. Detailed imaging parameters were: field-of-view = 256×192×150 mm³; acquired voxel size = 0.67×1.00×1.34 mm³; reconstructed voxel size = 0.67×0.67×1.34 mm³; repetition time/echo time = 27/4.7 ms; pixel bandwidth = 362 Hz. The acquired images were reconstructed to coronal and sagittal planes for MRS localization. Two volumes of interest (VOIs) were placed on the basal ganglia and midbrain regions to acquire the spectral-edited data, respectively. Detailed parameters were VOI = 30 ×30 ×30 mm³ (basal ganglia) or 25×35×30 mm³ (midbrain); TR/TE = 2000/130 ms; spectral bandwidth = 2000 Hz; readout duration = 512 ms; edit on/off frequency = 1.89/7.46 ppm, 9 dynamics with 16 signal average each.
Spectral-edited data were processed using the GANNET 3.1 version to get the GABA, Glx and Cr concentration³. Fit error larger than 15% was considered unreliable and excluded. Visual inspection of the remaining edited spectrum was done to confirm the correctness of the result. The final relative quantification results were expressed as the ratio of GABA/Cr and Glx/Cr.
A paired t-test was applied to compare the longitudinal changes in GABA and Glx level. P values < 0.05 were considered statistically significant. The range of agreement between two scans was visualized with Bland–Altman plots⁴. Results were given as mean ± SD. The statistical analysis was performed using IBM SPSS Statistics (version 24; IBM Corp., Armonk, NY).

Results

Based on the spectral quality inspection, 5.1% (2/39) of the MRS data from basal ganglia and 23% (9/39) of the MRS data from the midbrain were excluded. The demographic and the available MRS features of the participants are shown in Table 1.
The spectra with fitting error < 15% for the basal ganglia and midbrain are shown in Figure 1. According to the Bland-Altman plot analysis (Figure 2), all data points for GABA and Glx level in the midbrain and Glx level in the basal ganglia were located within the limit of agreement, and 2/14 data points for GABA level in the basal ganglia region fall outside the 95% confidence region (bias±1.96 SD). Although, the group levels showed no statistical differences (Table 2), the GABA and Glx levels of two scans in intra-subjects are fluctuating (Figure 3).

Discussion and Conclusion

The reproducibility of GABA and Glx measurements in vivo in the midbrain and basal ganglia regions is essential which is beneficial to the longitudinal study in neurodegenerative and neuropsychiatric disorders. The preliminary results demonstrated that applying ¹H-MRS to investigate the progression of GABA and Glx level in vivo is applicable. Fluctuations of GABA and Glx levels could be seen for each subject, this could be a result of many factors: e.g. female menstrual cycle, circadian rhythm, or insomnia⁵. Cares should be taken for the choice of scanning time. The exclusion ratio for the midbrain is higher than the basal ganglia region. Vascular pulsation, and field inhomogeneity could be the potential confounding factor for the GABA quantification⁶, which could be improved by careful positioning to avoid vessels and air-tissue interfaces.
In conclusion, the reproducibility of GABA and Glx measurements is beneficial to the longitudinal study in neurodegenerative and neuropsychiatric disorders, the preliminary results informed that applying ¹H-MRS to investigate the progression of GABA and Glx levels in vivo is applicable.

Acknowledgements

This work was supported, in part, by the National Natural Science Foundation of China (grant number: 82271954, 81971576); Chinese National Science & Technology Pillar Program (grant number: 2022YFC2009900/2022YFC2009905) and the Innovative Research Team of High-level Local Universities in Shanghai.