Table 3.
Neuroimaging methods of reviewed studies.
| Glutamatergic compound | References | Imaging sample size (MDD, HC) | Imaging technique | ROIs for activity/cerebral blood flow/entropy/receptor binding/glucose metabolism measurement | Methods for connectivity measurement | Timepoint(s) of MDD imaging relative to treatment |
|---|---|---|---|---|---|---|
| Ketamine | Abdallah et al. (131) | 18, 25 | fMRI | N/A | Average correlation between BOLD time series of a voxel and all other gray matter voxels in brain (GBCr) | Pre-treatment and 24 h post-treatment |
| Abdallah et al. (132) | 22 (Cohort A only), 29 HC in Cohort A and 18 HC in Cohort B | fMRI | N/A | Average correlation between BOLD time series of a voxel and all other gray matter voxels in brain (GBCr) | • Cohort A: pre-treatment and 24 h post-treatment • Cohort B: pre-treatment and repeated during each treatment session |
|
| Abdallah et al. (133) | 56, 0 | fMRI | N/A | Average correlation between BOLD time series of a voxel and all other gray matter voxels in brain (GBCr) | Pre-treatment, during infusion, and 24 h post-treatment | |
| Carlson et al. (134) | 20, 0 | 18F-FDG-PET | • Whole-brain CMRGlu • ROIs in AMYG, sgACC, anterior HPC, and medial TH |
N/A | Pre-treatment and ~120 min post-treatment | |
| Chen et al. (135) | 48, 0 | fMRI | N/A | • Six ROIs in prefrontal regions (dACC, DLPFC, and mPFC) • Seeds defined as structures with a 4 mm radius around the coordinates • FC map of ROIs identified by correlating low-frequency fMRI fluctuations with the seeds |
Pre-treatment and on third day after treatment (~48 h post-treatment) | |
| Chen et al. (136) | 48, 48 | fMRI | N/A | • Four striatal seeds (limbic, executive, rostral-motor, caudal motor) • FC map of striatum identified by correlating low-frequency fMRI fluctuations with the ROIs |
Pre-treatment | |
| Downey et al. (137) | 56, 0 | fMRI | ROI in sgACC (x = 0, y = 26, z = −8 mm) | N/A | Pre-treatment and 1 h scanning session which treatment was administered during | |
| Evans et al. (138) | 33, 25 | fMRI | N/A | • DMN defined using seed-based correlation method (3dTcorr) • Average time course from 6 mm radius sphere at PCC (3dROIstats) at the MNI template coordinates of (0, 252, 27) correlated with all other brain voxels |
Pre-treatment and ~2 and 10 days after both infusions | |
| Gärtner et al. (139) | 24, 0 | fMRI | N/A | • ROIs (MNI): sgACC (sphere at 2, 28, −5 with 5 mm radius), the left and right DLPFC (sphere at ±40, 36, 32 with 10 mm radius), the left and right PCC (sphere at ±6, −50, 24 with 7 mm radius), and the left and right HPC (anatomical seeds from the FSL Harvard-Oxford Atlas) • Seed-to-voxel correlation maps calculated by extracting residual BOLD time course from seed and computing correlation between that time course and time course of all other voxels; GLM |
Pre-treatment and 24 h post-treatment | |
| Gilbert et al. (140) | 29, 0 | MEG | ROIs in left AI (Talairach coordinates: −32, 13, −8) and ACC (Talairach coordinates: −2, 43, 2) | N/A | Pre-treatment and 6–9 h post-treatment | |
| Kraus et al. (141) | 28, 22 | fMRI | N/A | • GBC map transformed to MNI; maps resampled to final voxel size of 3.5 mm isotropic (voxel volume: 42.875 mm3) • Intra-PFC GBC calculated using same procedures, but GBC calculations restricted to the PFC with a PFC GM mask |
Pre-treatment and day 2 or 3 after ketamine or placebo | |
| Lally et al. (142) | 20, 0 | 18F-FDG-PET | Whole-brain CMRGlu and ROIs in ventral striatum and OFC | N/A | Pre-treatment and 2 h post-treatment | |
| Li et al. (143) | 48, 0 | 18F-FDG-PET | • ROIs in AMYG and PFC • Whole-brain voxel-wise analyses |
N/A | Pre-treatment and immediately after treatment | |
| Loureiro et al. (144) | 44, 31 | fMRI | Whole-brain analysis and ROI in AMYG | N/A | Pre-treatment and 24–72 h after last infusion | |
| Loureiro et al. (145) | 46, 32 | fMRI | • Cluster in CBM in lobule-VIIb in dorsal-attention part of CBM defined as psychophysiological-interaction seed • ROIs for CEN (lateral OFC, inferior frontal cortex, supramarginal gyrus, and DLPFC), SN (AI, dACC, and middle frontal cortex) and SMN (posterior INS, precentral and the superior paracentral) |
Psychophysiological-interaction analysis using CBM seed generated from the average NoGo-Go activation; GLM | Pre-treatment, 24 h after first infusion, and 24 or 72 h after last infusion | |
| McMillan et al. (146) | 26, 0 | fMRI and EEG | sgACC | N/A | Treatment administered 7 min into a 16 min scan | |
| Mkrtchian et al. (147) | 27, 19 | fMRI | N/A | • Seed-to-whole-brain analysis with four striatal seeds (3.5 mm radius spheres): ventral striatum (±9, 9, −8), dorsal caudate (±13, 15, 9), dorsal caudal Pu (±28, 1, 3), and ventral rostral Pu (±20, 12, −3) • Primary visual cortex used as control region for sensitivity analysis examining whether results were specific to PFC regions or due to global pattern |
2 days post-treatment | |
| Morris et al. (148) | 42, 20 | fMRI | ROI in sgACC | Computed between sgACC and HPC ROI's using Pearson's correlation | Pre-treatment and within 5 days after infusion | |
| Murrough et al. (149) | 18, 20 | fMRI | • Whole-brain, voxel-wise GLM • Single-subject whole-brain maps reflecting BOLD signals |
Investigated FC of regions that demonstrated brain activation main effects; GLM | Pre-treatment and 24 h post-treatment | |
| Nakamura et al. (150) | 15, 0 | fMRI | N/A | • Whole brain FC maps with seed regions for each hemispheric AMYG and bilateral ventral PCu in the MNI space • FC between seed and each voxel of whole brain computed as contrast of parameter estimates of the GLM |
Pre-treatment and 6–24 h after last infusion | |
| Nemati et al. (151) | 258 total (randomized participants were scanned), 0 | fMRI | N/A | GBC computed with average correlation of each voxel/vertex with all other gray matter voxels and vertices | Pre-treatment and during infusion | |
| Nugent et al. (152) | 13, 18 | MEG and 18F-FDG-PET | N/A | • Data from subjects entered into group independent components analysis to extract 25 components; linear regression used to obtain independent component maps • ROIs in left and right AMYG and sgACC |
Pre-treatment, MEG 6–7 h post-treatment, 18F-FDG-PET ~120 min post-treatment | |
| Reed et al. (153) | 33, 26 | fMRI | Whole brain analysis | N/A | Pre-treatment and 1–3 days after each infusion | |
| Reed et al. (154) | 33, 24 | fMRI | Whole brain analysis | N/A | Pre-treatment and 1–3 days after each infusion | |
| Roy et al. (155) | 11, 0 | fMRI | • 132 ROIs • Cortical and subcortical from the FSL Harvard-Oxford atlas • Cerebellar regions from the automated anatomical labeling atlas |
N/A | Pre-treatment and 1 day after last infusion | |
| Sahib et al. (156) | 22, 18 | Perfusion MRI | Regional cerebral blood flow values from ROIs that showed significant changes in whole brain analysis | N/A | Pre-treatment, 24 h after first infusion, 24–72 h after last infusion | |
| Sahib et al. (157) | 47, 32 | fMRI | • Whole brain analysis • ROIs in right and left SMA |
N/A | Pre-treatment, 24 h after first infusion, 24–72 h after last infusion | |
| Sahib et al. (158) | 61, 40 | fMRI | N/A | • Nodes generated with high-dimensional group independent components analysis • Network modeling performed using partial temporal correlation between node time series |
Pre-treatment, 24 h after first infusion, 24–72 h after last infusion | |
| Salvadore et al. (159) | 11, 11 | MEG | • False discovery rate calculated using a ROI approach encompassing the ACC/Brodmann area 24/32 • AMYG exploratory analyses |
N/A | Pre-treatment | |
| Salvadore et al. (160) | 15, 0 | MEG | • False discovery rates estimated within ROI encompassing the ACC/Brodmann area 24/32 | • Dynamic imaging of coherent sources using single seed point in the pgACC • ROI encompassing bilateral AMYG |
Pre-treatment | |
| Siegel et al. (161) | 23, 27 | fMRI | N/A | Exploratory FC analysis to visualize connectivity between three targets (DMN, sgACC, limbic system) and the rest of the brain | Pre-treatment and 2 weeks post-treatment | |
| Sterpenich et al. (162) | 10, 0 | fMRI | • Whole brain analysis with anatomical masks for AMYG, INS, ACC, OFC, and ventral striatum • ROIs in medial substantia nigra/VTA |
N/A | Pre-treatment and 1 and 7 days post-treatment | |
| Thai et al. (163) | 11, 0 | fMRI | ROIs in left and right HPC, left and right AMYG, subcallosal cortex, ACC, left and right NAc, PCu, and PCC | N/A | Pre-treatment and 1 day following last infusion | |
| Tiger et al. (164) | 30, 0 | PET | ROI in CBM | N/A | Pre-treatment and 24–72 h post-treatment | |
| Vasavada et al. (165) | 44, 50 | fMRI | N/A | • DMN, CEN, and SN chosen to investigate FC with HPC (right and left) and AMYG (right and left) • Time courses for AMYG and HPC extracted ROI masks derived from the Harvard–Oxford subcortical structural atlases • Correlations calculated between time courses of the networks and seeds |
Pre-treatment, 24 h after first infusion, 24–72 h after last infusion | |
| Wang et al. (166) | 60 (MDD and bipolar without psychotic symptoms), 0 | fMRI | N/A | • Left and right habenula identified with centers of MNI coordinates (−2.8, −24.2, 2.3) and (4.8, −24.1, 2.2), respectively • Both of which have a radius of 3 mm |
Pre-treatment | |
| Memantine | Kilpatrick et al. (167) | 26, 0 | fMRI | N/A | Anterior and posterior DMN nodes (PCu and mPFCs) with dual regression to create individual DMN parameter estimate maps | Pre-treatment and 3-month follow-up |
| D-cycloserine | Cole et al. (168) | 10, 12 | EMG | M1 localized using neuronavigation and electromyographic electrodes positioned over right first dorsal interosseous muscle | N/A | 1–2 h post-treatment |
| Nitrous oxide | – | – | – | – | – | – |
| 4-chlorokynurenine | Park et al. (169) | 12, 0 | 1H-MRS and fMRI | pgACC | 30 ROIs previously identified as being associated with MDD using a 6 mm sphere | Pre-treatment and ~60–120 min post-treatment |
| Lanicemine | Abdallah et al. (133) | 56, 0 | fMRI | N/A | Average correlation between BOLD time series of a voxel and all other gray matter voxels in brain (GBCr) | Pre-treatment, during infusion, and 24 h post-treatment |
| Downey et al. (137) | 56, 0 | fMRI | ROI in sgACC (x = 0, y = 26, z = −8 mm) | N/A | Pre-treatment and 1 h scanning session which treatment was administered during |
1H-MRS, proton magnetic resonance spectroscopy; 18F-FDG, fluorodeoxyglucose; AI, anterior insula; ACC, anterior cingulate cortex; AMYG, amygdala; BOLD, blood-oxygen-level-dependent; CBM, cerebellum; CEN, central executive network; CMRGlu, cerebral metabolic rate for glucose utilization; CN, caudate nucleus; dACC, dorsal anterior cingulate cortex; DLPFC, dorsolateral prefrontal cortex; DMN, default mode network; EEG, electroencephalogram; EMG, electromyography; FC, functional connectivity; fMRI, functional magnetic resonance imaging; GBC, global brain connectivity; GBCr, global brain connectivity with global signal regression; GLM, general linear modeling; GM, gray matter; HC, healthy control subjects; HPC, hippocampus; INS, insula; MDD, major depressive disorder participants; MEG, magnetoencephalography; MNI, Montreal Neurological Institute; mPFC, medial prefrontal cortex; MRI, magnetic resonance imaging; N/A, not applicable/available; NAc, nucleus accumbens; OFC, orbitofrontal cortex; PCC, posterior cingulate cortex; PCu, precuneus; PET, positron emission tomography; PFC, prefrontal cortex; pgACC, pregenual anterior cingulate cortex; Pu, putamen, ROI, region of interest; sgACC, subgenual anterior cingulate cortex; SMA, supplementary motor area; SN, salience network; TH, thalamus; VTA, ventral tegmental area.