Effect of Ketamine on Human Neurochemistry in Posterior Cingulate Cortex: A Pilot Magnetic Resonance Spectroscopy Study at 3 Tesla

Abstract

Ketamine is a powerful glutamatergic long-lasting antidepressant, efficient in intractable major depression. Whereas ketamine's immediate psychomimetic side-effects were linked to glutamate changes, proton MRS (¹H-MRS) showed an association between the ratio of glutamate and glutamine and delayed antidepressant effect emerging ∼2 h after ketamine administration. While most ¹H-MRS studies focused on anterior cingulate, recent functional MRI connectivity studies revealed an association between ketamine's antidepressant effect and disturbed connectivity patterns to the posterior cingulate cortex (PCC), and related PCC dysfunction to rumination and memory impairment involved in depressive pathophysiology. The current study utilized the state-of-the-art single-voxel 3T sLASER ¹H-MRS methodology optimized for reproducible measurements. Ketamine's effects on neurochemicals were assessed before and ∼3 h after intravenous ketamine challenge in PCC. Concentrations of 11 neurochemicals, including glutamate (CRLB ∼ 4%) and glutamine (CRLB ∼ 13%), were reliably quantified with the LCModel in 12 healthy young men with between-session coefficients of variation (SD/mean) <8%. Also, ratios of glutamate/glutamine and glutamate/aspartate were assessed as markers of synaptic function and activated glucose metabolism, respectively. Pairwise comparison of metabolite profiles at baseline and 193 ± 4 min after ketamine challenge yielded no differences. Minimal detectable concentration differences estimated with post hoc power analysis (power = 80%, alpha = 0.05) were below 0.5 μmol/g, namely 0.39 μmol/g (∼4%) for glutamate, 0.28 μmol/g (∼10%) for Gln, ∼14% for glutamate/glutamine and ∼8% for glutamate/aspartate. Despite the high sensitivity to detect between-session differences in glutamate and glutamine concentrations, our study did not detect delayed glutamatergic responses to subanesthetic ketamine doses in PCC.

Keywords: depression; glutamate; glutamine; ketamine; ketamine metabolites; magnetic resonance spectroscopy; neurotransmitters; posterior cingulate cortex.

FIGURE 1

Voxel position and sample MR spectra. An example of spectra acquired before (MRI1) and after ketamine administration (MRI2) is shown along with their fits and residuals of the fits resulting from LCModel quantification. The insets with T1-weighted MPRAGE images depict the typical MRS-voxel position in the posterior cingulate cortex in MRI1 (red) and MRI2 (blue), and their mutual overlap (in purple).

FIGURE 2

Coefficients of variation. Mean between-session (within-subject) coefficients of variation (CV, SD/mean) illustrate the variance between baseline and post-ketamine scan. Bars are means; error bars represent between-subject standard deviations.

FIGURE 3

Metabolite quantification in LCModel. Concentrations of metabolites were assessed separately or as sums and are shown in absolute units. Ratios of metabolites are presented in relative units. Data acquired at baseline and 193 ± 4 min. after ketamine administration (N = 12) were compared with the standard paired t-test, which revealed no differences between pre- and post-ketamine sessions.

FIGURE 4

Measured differences and estimated effect size. Plot displays absolute values of measured average differences between sessions (baseline minus post-ketamine, N = 12) and minimal detectable differences estimated with power of 0.8 and alpha = 0.05.