Introduction

Aberrant metabolism is a principle hallmark of many cancer types, characterized by an increased uptake in glucose and the formation of lactate in the presence of oxygen¹. Detection of elevated lactate, which is now understood to play a key role in the survival, progression, and metastatic potential of cancer cells^2,3, would permit assessment of tumour response following therapy.
Deuterium (²H) magnetic resonance spectroscopy (MRS) is a novel, non-invasive metabolic imaging method which tracks the dynamic conversion of orally administered ²H-labeled glucose into lactate and other downstream metabolites^4,5. Recent studies have demonstrated the feasibility of ²H MRS in differentiating metabolism between normal and tumour brain tissue⁴, but the sensitivity of this method to metabolic changes before and after chemotherapy has not yet been investigated. In this abstract, we study the feasibility of using ²H MRS to detect metabolic changes occurring in acute myeloid leukemia cells after treatment with cisplatin.

Methods

Cell preparation: Acute myeloid leukemia (AML-5) cells were grown in suspension in 8 flasks, each containing 150 mL α medium with 5% fetal bovine serum, and 1% penicillin and streptomycin. Flasks were maintained at 37°C and 5% CO₂ until they reached confluence (~10⁶ cells/mL). To induce apoptosis, cells in each of four flasks were treated with 10 μg mL^-1 cisplatin 48 hours prior to imaging, while four flasks were left untreated as controls. Cisplatin is a cancer chemotherapy drug that impairs DNA replication and repair, and its apoptotic effects on AML-5 cells are established in literature⁶.

Samples were prepared by combining the four flask volumes (~6 x 10⁸ cells total), which were centrifuged and resuspended in phosphate-buffered saline (PBS) with a total volume of 600 μL.10 minutes prior to the ²H scan, 100 uL of 140 mM [6,6’-²H₂]glucose solution was added to each tube for a final ²H-labeled glucose concentration of 20 mM (and final volume=700 uL). The experiment was repeated with 12 flasks, divided equally into an untreated group, and groups treated 24 and 48 hours prior to scanning.

MR Acquisition: MR experiments were performed at 7T (Bruker Biospec). A home-built ²H transmit/receive surface coil (46.1 MHz, diameter 20 mm) with proton blocking was used for acquiring the ²H signal. Three tubes (length 50 mm, diameter 6 mm) were placed side-by-side in the center of the surface coil and scanned in each experiment. 2D projection dynamic ²H MRS images were acquired starting at 10 minutes after glucose addition with an in-plane resolution of 10x3 mm2 and temporal resolution of 2 min (FOV=8x3 cm², spectral width 3 kHz, 768 spectral points, flip angle=90°, TR=300ms, 400 averages, scan duration=120min). ²H MRSI maps were overlaid on anatomical proton T1w images (TR=51 ms, TE=2 ms) acquired with a volume coil.

Analysis: The representative spectrum for each sample was chosen from the voxel in the most sensitive region of the surface coil with the highest signal. Spectra were averaged from 20 min of data acquisition. Time domain fitting was performed using AMARES implemented in the OXSA Matlab toolbox⁷. The fitted peaks were deuterium-labeled glucose (3.7 ppm), glutamate and glutamine (Glx, 2.4 ppm), lactate (1.4 ppm), and water (4.7 ppm). Concentrations of labeled glucose and lactate were quantified by normalizing the integrals of the fitted peaks to that of the baseline water signal, which has a natural ²H abundance of 12.8 mM. This value is based on the 55.5 M concentration of water and 0.0115% natural abundance of deuterium. As described by de Graaf et al.⁴, the final concentrations were normalized to the average number of deuterons per molecule.

Results and Discussion

Figure 1 demonstrates the apoptotic changes in the AML-5 cells following cisplatin treatment. The ²H spectra acquired two hours after the injection of [6,6’-²H₂]glucose are presented in Figure 2a. Signals from the injected glucose and downstream lactate product were observed, but no Glx was detected in either the treated or untreated group. This result is consistent with observations of increased lactate and reduced Glx formation in tumours compared to healthy tissue⁴. Less lactate was detected in the treated cells relative to the untreated cells, although both samples demonstrated glucose utilization.

Figure 3 shows the concentrations of the deuterium-labelled metabolites over time. At two hours following addition of ²H-glucose, the amount of ²H-lactate detected in the 48 hour cisplatin-treated sample was 3.3 fold less than that of the untreated sample. The ²H-lactate to ²H-glucose ratio was 4.0 fold less respectively, following lactate/glucose trends seen in a previous study⁸. This suggests reduced glucose metabolism in the treated sample, which may be because majority of these cells have undergone apoptosis. Reduced lactate production was also observed in the repeated experiment (summarized in Figure 5). Relative to the untreated cells, the ²H-lactate to ²H-glucose ratios were 3.5 and 102.2 fold less for the 24 and 48 hour treated samples, respectively. Future in vitro assays are required to correlate ²H-metabolite concentrations with the degree of apoptosis in the treated and untreated groups.

Conclusions

At 7T, it was shown that ²H MRS is sensitive to metabolic changes induced by cisplatin in AML cells, specifically, the decrease in lactate production. We anticipate that this method will be useful in monitoring tumour response after chemotherapy.

Acknowledgements

We gratefully acknowledge funding from NSERC (RGPIN-2017-06596).