Purpose

While current approaches for glucose metabolism assessments such as positron emission tomography (PET), ¹³C magnetic resonance spectroscopy (¹³C-MRS), and direct deuterium imaging¹ are technically challenging when requiring special hardware or expensive synthesis of intravenous radioactive tracers, a recent animal study showed the incredible potential of indirect detection of deuterated compounds via proton magnetic resonance spectroscopy (¹H-MRS) to quantify glucose (Glc) metabolism in the rat brain.² The method of indirect ²H detection enables quantification of both oxidative and anaerobic Glc utilization and neurotransmitter synthesis. Thus, we aimed to establish, for the first time, non-invasive detection of deuterium enrichment of downstream Glc metabolites such as glutamate (Glu), glutamine (Gln), and gamma-aminobutyric acid (GABA) after peroral administration of 6,6’-²H₂-glucose (Deu-Glc) using state-of-the-art ¹H-MRS³/MRSI⁴ in the human brain.

Methods

Five healthy lean volunteers (30±4 y.o., four males) were scanned after overnight fasting at 7T MR system utilizing a single-channel transmit/ 32-channel receive-array Nova head coil. Each participant had 2 separate MR sessions immediately after Deu-Glc and non-deuterated D-Glc (nonDeu-Glc) (0.8 g/kg) administration. MRSI and MRS data were interleaved with Siemens Autoalign navigator to assure the stable position of the localized volume.⁵ The first MRS/MRSI block was acquired after MRS calibrations and B₀-shimming within 30 minutes after Deu-Glc/nonDeu-Glc administration. Standard 2^nd-order B₀-shimming was performed via an imaging-based approach (2 iterations) and FASTMAP⁶ for MRSI and MRS, respectively. Multivoxel FID-MRSI was performed with ultrashort TE of 1.3 ms, TR of 320 ms, 3D k-space concentric ring encoding, variable temporal interleaves, 36x36x26 matrix, and 5x5x4.8 mm³ voxel size.⁷The FOV was 180x180x110 mm to minimize acquisition time (3 min. per block) and lipid contamination. A four cm-thick slab was centered around the posterior cingulate region (PCC). Single-voxel MRS data were obtained from the PCC (22×20×20 mm) using a semi-LASER pulse sequence (TR = 7s, TE = 28 ms, AT = ~4 min.).⁸ The metabolite spectra were collected with VAPOR water suppression⁹ and outer volume suppression (number of excitations = 32). Unsuppressed water spectra were utilized to remove residual eddy currents and as a reference for quantification. Basis sets of simulated spectra for LCmodel included separated Glu₄ and Glu₂₊₃ components accounting for the prominent effect of deuteration on C4 position in Glu molecule. Separate Gln₄ and GABA₂ components allowed the stable fitting of MRS data but were not used to quantify MRSI.

Results

Significant decreases of Glu₄ and Gln₄ in the PCC (14±2%, 12±5% resp.) and 13±4% and 14±3% decrease of Glu₄ in the whole gray (GM) and white matter (WM) after Deu-Glc (first minus last point) were detected using MRS and MRSI, respectively. All metabolites quantified in LCmodel did not change significantly after nonDeu-Glc. Glu₄ was quantified with within-session CVs of ~2% as calculated in PCC and from regional averages in GM and WM, Gln₄ of ~4% (MRS), and GABA2 of ~8% (MRS, nonDeu scan), allowing robust detection of Glu₄ signal drop by both methods (Fig. 1) and non-significant results for GABA₂ in the analysis of individual subject’s data. The exponential fits (M=M0*exp-t/tau+c) of Glu₄ time-courses yielded time constants of the decay (tau) with inter-subject coefficients of variation of 25% (MRS, Fig. 2), 50% (GM), and 45 (WM, MRSI, Fig. 3). The time constants of the decay differed significantly between GM and WM and were ~18% lower in the GM. The higher Glu turnover in the GM resulted from voxel-wise fitting of single-subject data obtained with high time-resolution (Fig. 4). The first and last time-point spectra were pooled from all subjects, summed per session, and subtracted (Deu-Glc minus nonDeuGlc session, Fig. 5). The resulting difference spectrum represents the effect of deuterium incorporation into metabolite pathways and allowed quantification of GABA₂ with CRLB of 16%.

Discussion

The inter-subject variance in Glu rates of ~25-50% is slightly above the gold standard CMRGlc measures with FDG-PET, reported in the range of 19%-29% (mean – median)¹⁰ and can be mainly ascribed to high physiological variation in the resting brain metabolism. About 20% difference in the Glu turnover in GM/WM is in line with fluorodeoxyglucose ([18F]FDG) PET studies, which measured ~33% higher oxidative Glc consumption in the GM than WM.¹¹ The voxel-wise fitting with linear regression yielded a contrast between GM/WM slopes, demonstrating feasible non-invasive mapping of Glu turnover without radioactive tracers and PET (Fig. 5). The method allowed to separate Glu from Gln, which is not feasible from deuterium-spectrum even at ultra-high 16.4T fields.¹² The basis sets with separated Glu and Gln components on C4 position enabled to extract quantitative information about Glu and Gln turnover similar to technically challenging ¹³C-MRS. We expect a further boost of quantification of relatively low abundant J-coupled metabolites (i.e., GABA and Gln) by optimizing the sequences for fields above 7T and spectral editing techniques.^13,14

Conclusions

We introduced a novel, affordable approach for metabolic measures that utilizes widely available ¹H MR scanner equipment to indirectly detect deuterated-compounds and can, thus, be easily used in clinical applications. The negligible risk associated with the deuterium administration compared to radioactive FDG predetermines the method for studies with a multi-session design. In contrast to direct deuterium-MRS, ¹H-MRS allows quantification of a vastly extended neurochemical profile, including non-deuterated compounds.

Acknowledgements

Acknowledgments: Authors thank Drs. Patrick Bolan and Chris Rogers for providing a tool to store and apply 7T B0-shims at 7T MR scanner. The authors acknowledge helpful discussions with Dr. Vladimir Mlynarik. PB was supported by the European Union’s Horizon 2020 research and innovation program under a Marie Skłodowska-Curie grant agreement, no. 846793, and by a NARSAD Young Investigator Grant from the Brain and Behavior Research Foundation, no. 27238. AS has received funding from the European Union’s Horizon 2020 research and innovation program under a Marie Skłodowska-Curie grant agreement, no. 794986. WB and GH acknowledge support by the Austrian Science Fund (FWF) grants P 30701, KLI 718 (WB), and KLI 646 (GH), respectively. TS was supported by FWF grant KLI 782.

References

1.De Feyter. Sci Adv 2018; 4: eaat7314.2.Rich.Nat Biomed Eng 2020; 4: 335–342.3.Bednarik.J Cereb Blood Flow Metab 2015; 35: 601–610.4.Bednarik.ISMRM 2020.5.Dou. Magn Reson Mater Physics, Biol Med 2015. doi:10.1007/s10334-014-0469-9.6.Gruetter. Magn Reson Med 2000; 43: 319–323.7.Hingerl. Invest Radiol 2020; 55: 239–248.8.Oz. Magn Reson Med 2011; 65: 901–910.9.Tkac. Magn Reson Med 1999; 41: 649–656.10.Sundar. J Cereb Blood Flow Metab 2019; 39: 1516–1530.11.Yu. J Cereb Blood Flow Metab 2018. doi:10.1177/0271678X17708691.12.Lu.J Cereb Blood Flow Metab 2017; 37: 3518–3530.13.Moser.Neuroimage 2019; 184: 475–489.14.Magnusson. Magn Reson Med 2019. doi:10.1002/mrm.27450.