Introduction

Cognitive impairment can occur in patients with early phase relapsing-remitting multiple sclerosis (RRMS). Gamma-aminobutyric acid (GABA) is associated with numerous aspects of cognitive function^[1], and previous studies have demonstrated that there is a dysfunctional GABAergic neurotransmission in the central nervous system in animal models of MS^[2,3]. However, to date there has been a paucity of studies exploring correlation between GABA levels and cognitive function in patients with MS. In this study, therefore, J-difference edited MRS was used to investigate brain GABA levels in patients with RRMS and healthy controls, and their relationship to cognitive function.

Methods

31 RRMS patients and 26 age-, gender- and education-matched healthy controls were recruited in this study. All subjects were scanned on a 3 T scanner (Philips “Achieva” TX, Best, The Netherlands) using an eight-channel phased-array head coil for receive. T1-weighted three-dimensional TFE images were used as a localizer. The MEGA-PRESS sequence was performed to detect GABA signal from three brain regions: (1) medial prefrontal cortex (mPFC) (2×3×3 cm³); (2) posterior cingulated cortex (PCC) (3×3×2 cm³); (3) left hippocampus (LHC) (4×2×2 cm³), as shown in Fig. 1. The MEGA-PRESS data were analyzed using 'Gannet' (GABA-MRS Analysis Tool) in Matlab with Gaussian curve fitting to the GABA+ peaks. All subjects were assessed using a set of standardized comprehensive cognitive tests. Information processing speed was assessed using the Symbol-Digit Modalities Test (SDMT). Executive function was measured using the Trail-making test (TMT). Verbal memory was assessed using the Rey Auditory Verbal Learning Test (RAVLT) for immediate and delayed recall. Visuospatial memory was assessed using the Rey-Osterrich complex figure test. Attention was assessed using the Stroop colour word interference test. The raw scores were standardized (z-transformed) based on the mean and SDs of the healthy control subjects. Group comparisons used analysis of covariance (ANCOVA). Between-group GABA levels comparison controlled for GM/(GM+WM) ratios and all between-group neuropsychological comparisons controlled for the influence of education. Correlations between GABA+ levels and neuropsychological scores in patients with RRMS were tested using the Pearson correlation coefficients.

Results

Patients showed reduced GABA+ in the PCC (1.047±0.172 IU (patients) vs 1.138±0.128 IU (controls), p=0.03) and LHC (1.184±0.223 IU (patients) vs 1.331±0.263 IU (controls), p=0.024) compared with controls. No statistical difference in GABA+ in the mPFC was seen between groups (p=0.064). Significant lower cognitive scores (RAVLT, SDMT, TMT) were found in patients compared with controls (p<0.05). In patients with RRMS, TMT scores were associated with GABA+ levels in PCC (r= -0.431，p=0.016); significant correlations between RAVLT scores and GABA+ were found in LHC (r=0.372，p=0.039), as shown in Fig. 2; no significant correlations between cognitive scores and GABA+ were found in mPFC.

Discussion

The main results of this study are that significantly decreased GABA+ levels were present in the PCC and LHC of patients with RRMS, and GABA+ levels correlated with cognitive scores. This is the first in vivo demonstration of reduced GABA+ concentrations in patients with RRMS, and the first in vivo demonstration of a relationship between GABA levels and cognitive function in patients with RRMS. The results might demonstrate a mechanistic connection between the GABAergic neurotransmission and cognitive function. Reduced GABA levels may be due to loss of inhibitory neurons, dysfunctional GABA synthesis or to dysfunction in enzymes involved in the glutamate/GABA-glutaminecycle. The observed reduced GABA+ levels may reflect deficits of GABA-mediated synaptic transmission, leading to altered inhibitory signaling within cortical circuits, this may ultimately result in the loss or deterioration in cognitive function^[4]. Therefore, these results suggest that the neurochemical characteristics of PCC and hippocampus statistically explain part of cognitive impairment in RRMS.

Conclusion

This study demonstrates that abnormalities of the GABAergic neurotransmission may be present in the pathogenesis of MS and provides a novel imaging marker for the diagnosis and treatment of cognitive impairment in MS.

Acknowledgements

This project was supported by the Shandong Provincial Natural Science Foundation of China (grant nos. BS2015YY003) and Shandong Provincial Medical and Healthy Technology Development Program of China (grant nos. 2015WS0176).