Investigation Of Brain GABA Levels In Patients With Alzheimer’s Disease:a Proton Magnetic Resonance Spectroscopy Study

With the development of aging, dementia gets worldwide concern. Dementia is defined as a collection of symptoms ranging from memory impairment, a loss of communication skills and a gradual deterioration in the person’s ability to carry out daily tasks and activities of living, which causes significant impairment in social and occupational functioning of the dementia people. There are several forms of dementia, Alzheimer’s disease (AD), vascular dementia (VaD), Dementia with Lewy bodies (DLB), frontotemporal dementia (FTD) et al. AD is considered to be the most common dementia, characterizing by three hallmark pathological Lesions (amyloid plaques, neurofibrillary tangles and synaptic loss). VaD is the second most common form dementia, which is described as the memory and cognitive loss symptoms due to vascular diseases within the brain. Although the exact biochemical processes of synaptic failure in dementia remain poorly understand, it is likely to include a gross dysfunction of neurotransmitter systems, including the excitatory pathways and inhibitory pathways. The disruption of excitatory pathways has been broadly accepted. γ-aminobutyric acid (GABA), the predominant inhibitory neurotransmitter in human brain, is also considered to be associated with cognitive function.Magnetic resonance spectroscopy (MRS) enables the non-invasive in vivo measurement of major neurometabolite levels and has been successfully used to evaluate brain regional N-acetylaspartate (NAA), creatine (Cr), choline (Cho) and myo-inositol (mI). Recent developments involving1H-MRS editing PRESS-based techniques, such as MEGA-Point Resolved Spectroscopy (PRESS), offers the noninvasive technique to evaluate GABA levels in the human brain, and has been successfully applied to measure GABA levels in normal brain and a number of neurologic, psychiatric diseases. While, paucity MRS study has assessed the brain level of GABA in AD patients.In this study, the edited MRS technique, MEGA-PRESS, was used to explore in vivo whether brain GABA levels change in AD and VaD patients by comparing to those of age-and gender-matched healthy control subjects. The assessment of relationship between GABA levels and the cognitive impairments in AD, VaD patients were also performed.This study contains three parts: Part1.Volume of Interest placement study for edited magnetic resonance spectroscopyObjective:MEGA-PRESS sequence is aneditedmagnetic resonance spectroscopy (MRS) technique, and has been successfullyused to evaluate the brain regional GABA levels.The accuracy of volume of interest (VOI) placementmay affect the reliability of GABA levels quantification. In this part of study we use registration tools to evaluate the MRS voxels overlapping rate within and between subjects.Material and Methods:13healthy subjects were recurited in this experiment. For each participant,T1-weighted3D TFE images were acquired for localization and GABA edited MRS data were acquired using MEGA-PRESS sequence. VOIs were firstly placed in the right sensorimotor region (SM) and in the occipital region (OCC). The second placement was done with the same protocol to log voxel location parameters, following with the MEGA-PRESS sequence to get the GABA data again.The voxel data of subjects acquired from the two scans were then registered to their Tl-TFE images of the first scan, and all the voxels from the first scan were regsitrated to the standard space. Theregistration process went as follows:Ⅰ.The voxels (both OCC and SM voxels) of each subject from the two scans were registered to their T1W-images of the first scans using Svmask tool, generating the voxel mask images, which were used forvoxel overlapping rate calculationwithin subjcect; Ⅱ. The voxels and whole brain.PAR files were converted to Nifti format with the use of Matlab toolkit; Ⅲ. In order to exclude the effect of non-brain tissue, the3D Nifti format brain images were extracted using Brain Extraction Tool (BET) from FSL software package;Ⅳ. All the voxels and extracted brain images from the first scanwere registered to the reference standard space using FMRIB’s Linear Image Registration Tool (Flirt) from FSL software package, and the registrated images were used to calculate thevoxel overlapping rate between subjects. After images registration, the Dice overlap coefficient (DOC) was used to calculate the MRS voxels overlap within and betweensubjects. Finally, the GABA levels (GABA+/Cr) of all subjects were calculated using Gannet tool in Matlab.Results:The DOC results show that the overlapping rates of occipital and sensorimotor voxels withinsubject were87%±5%and86%±5%,respectively. The MRS voxels overlapping rates betweensubjects were75%±10%in OCC voxel and78%±7%in SM voxel. No statistical differences of GABA+/Cr ratios were found between the two scans (OCC:t=-1.301, p=0.206;SM:t＝0.260, p=0.797).Conclusion:The overlapping rate of voxel placement withinsubject and betweensubjects in this study show over75%, with no significant differences of GABA+/Cr ratios between the two scans, suggesting thatVOI placement of MEGA-PRESS sequence has high repeatability andMEGA-PRESS sequence can be used in the study on the measurement of brain regional GABA levels in vivo.Part2. Evaluation of GABA levels in patients with Alzheimer’s diseaseObjectives:To determine whether there are in vivo differences of y-aminobutyric acid (GABA) levels in frontal and parietal regions of Alzheimer’s disease (AD) patients, compared with matched healthy controls using proton magnetic resonance spectroscopy (1H-MRS), and whether brain GABA levels correlate with cognitive impairment.Material and Methods:Fifteen AD patients and fifteen age-and gender-matched healthy controls underwent1H-MRS of the frontal and parietal lobes using the ’MEGA-Point Resolved Spectroscopy Sequence’(MEGA-PRESS) technique, and cognitive levels of all subjects were evaluated using Mini-Mental State Examination (MMSE) tests. MRS data were processed using the Gannet program, which includes contributions from GABA, macromolecules and homocarnosine (and therefor the signal detected is defined as GABA+). Differences of GABA+to creatine (GABA+/Cr) ratios between AD patients and controls were tested using a two-tailed t-test, and Spearman correlation analysis was used to evaluate the relationship between GABA+/Cr and MMSE scores.Results:Significant lower GABA+/Cr ratios were found in the parietal region of AD patients compared to controls (t=-2.561, p=0.016). Although the mean GABA+/Cr ratios in the frontal region of AD (0.098±0.008) were lower than those of healthy controls (0.104±0.022), this finding was not statistically significant (t=-1.012, p=0.325). In the AD patients, no significant correlations between GABA+/Cr and MMSE scores were found in either the frontal (r=-0.164, p=0.558) or parietal regions (r=0.025, p=0.929).Conclusion:Decreased GABA+/Cr levels were present in the parietal region of patients with AD in vivo, suggesting that abnormalities of the GABAergic system may be present in the pathogenesis of AD. Part3.The evaluation of GABA levels in patients with subcortical ischemic vascular dementia and Alzheimer’s diseaseObjective:In this part, we use edited MRS technique to evaluate the brain GABA levels of frontal, parietal regions in patients with Subcortical ischemic vascular dementia (SVID) and to assess whether there are brain GABA levels differences among SVID, AD patients and age-, gender-matched healthy controls.Material and Methods:13clinically diagnosed SVID patients underwent T1-FFE sequence and edited MRS scan (using MEGE-PRESS sequence), and MMSE test were performed to evaluate the cognitive impairment of the SVID patients. The MRI and MRS data of the15patients with AD and age-gender-matched healthy controls collected in part-two study were also recruited this study. The MRS data were processed using Gannet tool just as part one study and the GABA levels were described as GABA+/Cr ratios. The differences of GABA+/Cr levels among SVID, AD patients and healthy controls were analyzed using one way ANOVA. The relationship between GABA+/Cr levels and MMSE scores were also performed using spearman correlation analysis.Results:Decreased mean GABA+/Cr levels were detected in both frontal (0.086±0.010) and parietal regions (0.086±0.014) of patients with SVID, compared with age-and gender-matched healthy controls (frontal:0.104±0.022; parietal:0.099±0.012). However, only in parietal region, significant lower GABA+/Cr levels were found in this study (F=4.954, p=0.012). With regard to the GABA+/Cr levels differences between SVID and AD patients, no significant differences were found in parietal regions (p=0.979). In additionally, no significant correlation between GABA+/Cr levels and MMSE scores were detected in either frontal (r=-0.301, p=0.318) or parietal regions (r=0.205, p=0.502).Conclusion:Decreased GABA+/Cr levels were found in parietal region of patients with SVID, which offered evidence that GAB Anergic system involves in the process of SVID. But no GABA+/Cr levels differences were found between SVID and AD patients, GABA+/Cr could notyet be a biomarkerto differentiate SVID from AD.