15th European Molecular Imaging Meeting
supported by:

To search for a specific ID please enter the hash sign followed by the ID number (e.g. #123).
*.ics

Imaging Metabolism

Session chair: Yannick Cremillieux (Bordeaux, France); Mor Mishkovsky (Lausanne, Switzerland)
 
Shortcut: PS 08
Date: Wednesday, 26 August, 2020, 12:00 p.m. - 1:30 p.m.
Session type: Parallel Session

Contents

Abstract/Video opens by clicking at the talk title.

12:00 p.m. PS 08-01

Introductory Lecture

Mor Mishkowsky1

1 EPFL SB IPHYS LIFMET, Lausanne, Switzerland

 
12:18 p.m. PS 08-02

[64Cu]Cu2+ a simple tool for PET imaging of Brown adipose tissue and its activation

Claudia A. Castro Jaramillo1, Marco F. Taddio1, Ioannis Kritikos2, Federica Petruzzelli1, Peter Runge2, Nicholas van der Meulen4, Roger Schibli3, 1, Stefanie D. Krämer1

1 Center for Radiopharmaceutical Sciences ETH, PSI and USZ, Department of Chemistry and Applied Biosciences of ETH Zurich, Zürich, Switzerland
2 Pharmaceutical Immunology, Department of Chemistry and Applied Biosciences, Zürich, Switzerland
3 Laboratory of Radiochemistry, Paul Scherrer Institute, Zürich, Switzerland
4 Center for Radiopharmaceutical Sciences ETH, PSI and USZ, Paul Scherrer Institute, Zürich, Switzerland

Introduction

The function of brown adipose tissue (BAT) in metabolic diseases is of growing interest. BAT activation can increase whole-body energy expenditure and is, therefore, under investigation as a therapeutic strategy against metabolic disorders. Beta-3 adrenergic receptor stimulation is one approach to activate BAT and induce the "browning" of white adipose tissue (WAT). Copper(II) is an enzymatic cofactor required for various biochemical processes, and it was recently linked to BAT homeostasis. We investigated whether BAT and BAT activation can be imaged by [64Cu]CuCl2 in vivo.

Methods

[64Cu]CuCl2 was produced and used directly to evaluate [64Cu]Cu2+ uptake in BAT of C57BL/6 mice. To evaluate the specificity of the uptake, [64Cu]CuCl2 was administrated in a complexed with the chelator NOTA [5uM]. To evaluate the effect of β-3 adrenergic receptor activation of BAT and browning of WAT, C57BL/6 mice were treated with CL 312,643, 1mg/kg (24 and 1h before [64Cu]Cu2+ injection). To evaluate the effect of inflammation on BAT, C57BL/6 mice were inoculated with LPS or PBS (3, 6 and 10 days). PET/CT-scans were performed 24 h p.i. with ~10 MBq [64Cu]CuCl2. All the studies were confirmed by post-mortem biodistribution. Using flow cytometry, RT-qPCR and confocal microscopy, the expression of Ctr‑1 (high affinity copper uptake protein 1) and other targets of interest was evaluated.

Results/Discussion

 [64Cu]Cu2+ uptake was significantly higher in BAT than WAT (figure 1), in agreement with the higher Ctr-1 in BAT revealed by RT-qPCR and flow cytometry. Specificity was confirmed by the co-administration of the chelator NOTA where no uptake was observed. [64Cu] Cu2+ uptake in both BAT and WAT was significantly increased by treatment with CL312,643, indicating BAT activation and WAT browning (figure 2). LPS-induced local r inflammation significantly increased both immune cells and [64Cu]Cu2+ uptake in BAT.  The  [64Cu]Cu2+ uptake in BAT correlated with the degree of inflammation evaluated by flow cytometry studies.   Upregulation of Ctr-1 in BAT but not WAT of animals with local inflammation was confirmed by flow cytometry.

Conclusions

We successfully imaged BAT using [64Cu]CuCl2 . Activation of BAT and browning of WAT correlated with and increased uptake of [64Cu]CuCl2  and an increase in the expression of Ctr-1.  [64Cu]CuCl2  can be used as a simple yet effective tool for imaging BAT and BAT activation under different types of stimuli.

AcknowledgmentThe project is funded by the Molecular Imaging Network Zurich (KFSP – MINZ).
References
[1] Berbée JF, Boon MR, Khedoe PP, Bartelt A, Schlein C, Worthmann A, Kooijman S, Hoeke G, Mol IM, John C, Jung C, Vazirpanah N, Brouwers LP, Gordts PL, Esko JD, Hiemstra PS, Havekes LM, Scheja L, Heeren J, Rensen PCN. Brown fat activation reduces hypercholesterolaemia and protects from atherosclerosis development. Nat Commun 6: 6356, 2015
[2] Cannon B, Nedergaard J. Brown adipose tissue: function and physiological significance. Physiol Rev 84: 277–359, 2004.
[3] Gifford A, Towse TF, Walker RC, Avison MJ, Welch EB. Human brown adipose tissue depots automatically segmented by positron emission tomography/computed tomography and registered magnetic resonance images. J Vis Exp: doi: 10.3791/52415, 2015.
Figure 1

PET images of 28 hours after [64Cu]CuCl2 injection. B) As A but the animals C57BL/6 mice were inoculated with LPS and monitored after 6 days. One representative out of 3 mice each.

Figure 2

Ex vivo bioditribution of [64Cu]CuCl2  injection 24 hours after injection. To activate BAT and induce WAT browing C57BL/6 mice were treated with 1mg/kg of CL312,643 (24 and 1h before [64Cu]Cu2+ injection). To activate BAT by inflammation C57BL/6 mice were inoculated with LPS (6 days before [64Cu]Cu2+ PET scan).  SUV, standardized uptake value.

Keywords: brown adipose tissue, PET, thermogenic activator, inflammation
12:30 p.m. PS 08-03

Comparison of tumor metabolic volumes using hyperpolarized 13C MR and [18F]FDG PET in endogenous T-cell lymphomas in mice

Frits van Heijster1, Christian Hundshammer1, Martin Grashei1, Jason G. Skinner1, Geoffrey Topping1, Tim Wartewig2, 3, Erik Hameister2, 3, Jürgen Ruland2, 3, 4, Franz Schilling1

1 Technical University Munich, Nuclear Medicine, Klinikum rechts der Isar, Munich, Germany
2 Technical University Munich, Institute for Clinical Chemistry and Pathobiochemistry, Klinikum rechts der Isar, Munich, Germany
3 Technical University Munich, TranslaTUM, Center for Translational Cancer Research, Munich, Germany
4 German Cancer Consortium (DKTK), Heidelberg, Germany

Introduction

T-cell non-Hodgkin lymphomas are known to be highly aggressive and often have poor clinical outcomes.1 Programmed death genes have been shown to play a key role in suppressing oncogenic T-cell signaling. High glycolytic phenotype tumors often show high levels of glucose uptake (Warburg effect).3 Differentiating between high (HGP) and low glycolytic phenotype (LGP) could benefit planning of immunotherapy treatment. For this purpose, in this study HGP and LGP tumors of T-cell lymphoma are compared to healthy animals using PET and MR.

Methods

LGP (N=6) and HGP mice (N=3), were measured using both PET and MR. Healthy mice (N=4) and healthy mice that received anti-PD antibodies (BE0146, Bioxcell) (N=2) were used as controls. After injection of [18F]FDG (~11 MBq), uptake in spleen and other organs was imaged using preclinical PET/CT. Blood glucose levels were determined and the animals were transferred to a preclinical 7T MRI with 13C/1H dual tuned volume coil (RAPID, ID=31 mm). After injection of hyperpolarized (HP) [1-13C]pyruvate (HyperSense, OX063 trityl radical), HP pyruvate and lactate signals were recorded in vivo, using a 3D bSSFP sequence with spectrally selective excitations and alternating flip angles (2° for pyr., 10° for lac.) (Fig1). T2w 1H RARE images were used for co-registration and to measure spleen sizes.

Results/Discussion

Standard uptake values (SUV) for [18F]FDG in LGP/HGP mice or healthy controls (Fig.1C) and tumor metabolic volumes (TMV) were calculated using spleen size (Fig. 2D). AUC ratios of pyruvate and lactate 13C time curves were calculated by selecting 2D ROIs covering the spleen, or 3D ROIs over the whole spleen (Fig.2F-G). Selection of ROIs is done in co-registered T2w-images. Similar to PET-TMV, a TMV is calculated using the AUC ratios (Fig.2B-C). Both PET and AUC-derived TMV increased with glycolytic phenotype. Intragroup differences became smaller when using 3D ROI’s, but relative intergroup differences became smaller. This might be explained by high proximate uptake in kidney and partial volume effects. Inhibition of PD in controls did not induce changes in SUV, AUC or TMV, showing that anti-PD alone does not induce a high-glycolytic phenotype. However, spleen sizes were smaller in anti-PD controls compared to healthy mice (Fig 2A).

Conclusions

Using fast 3D bSSFP imaging, tumor metabolic volumes could be calculated from 13C-MR-derived AUC values, similar to TMV derived from SUV (PET). Both were found to increase with glycolytic phenotype. While metabolic phenotypes could be differentiated by PET, AUC ratios showed only a positive trend for the 2D ROI analysis.

Acknowledgment

We thank Markus Mittelhäuser and Sybille Reder for performing the PET measurements and Sandra Sühnel for her help with the animal handling. This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 820374.

References
[1] Casulo, C. et al. T-cell lymphoma: recent advances in characterization and new opportunities for treatment, J. Natl Cancer Inst., 109:248, 2016
[2] Wartewig T, Kurgyis Z, Keppler S, Pechloff K, Hameister E, Öllinger R, Maresch R, Buch T, Steiger K, Winter CJN. PD-1 is a haploinsufficient suppressor of T cell lymphomagenesis, Nature, 553:238, 2017
[3] Hanahan, D., Weinberg, R. A. Hallmarks of cancer: the next generation. Cell, 144, 2011
13C and 1H MR and PET measurements

Fig.1: T2w 1H images of an HGP tumor (B) and a healthy mouse (A) with spleens indicated. 13C-pyruvate (red) and lactate (blue) signal over time are shown in the bottom, derived from ROIs indicated in co-registered 1H images. (C) [18F]FDG uptake measured by PET with spleen, heart and bladder indicated.

AUC SUV and TMV values
Fig.2: (A) Spleen volumes of low glycolytic phenotype (LGP), high glycolytic phenotype (HGP) mice, and healthy control animals (Ctrl.) with or without anti-PD antibody treatment. (B-D) Tumor metabolic volumes (TMV) calculated from PET-derived SUVmean values (E), or 13C-MRSI-derived AUC ratios (2D ROI (F), 3D ROI (G)), using Eq. 1&2. TMV increases with glycolytic phenotype. (E) Glucose concentrations in blood. (*p<0.05, **p<0.01). Not all animals measured by PET could be measured using MR.
Keywords: Hyperpolarization, 18F-FDG-PET, T-cell lymphoma
12:42 p.m. PS 08-04

Cerebral metabolism of the hyperpolarized neuroprotective agents [1-13C] lactate and [1-13C] pyruvate in a mouse model of transient ischemic stroke

Thanh Phong Lê1, 2, Lara Buscemi3, Elise Vinckenbosch1, Mario Lepore4, Lorenz Hirt3, Jean-Noël Hyacinthe1, 5, Mor Mishkovsky2

1 Geneva School of Health Sciences, HES-SO University of Applied Sciences and Arts Western Switzerland, Geneva, Switzerland
2 École polytechnique fédérale de Lausanne (EPFL), Laboratory of Functional and Metabolic Imaging, Lausanne, Switzerland
3 Lausanne University Hospital (CHUV), Department of Clinical Neurosciences, Lausanne, Switzerland
4 École polytechnique fédérale de Lausanne (EPFL), Centre d'Imagerie Biomédicale (CIBM), Lausanne, Switzerland
5 University of Geneva (UNIGE), Image Guided Intervention Laboratory, Geneva, Switzerland

Introduction

Ischemic stroke is a major cause of death and disability worldwide. Neuroprotective strategies in acute stroke could improve patient recovery. Administration of pyruvate and lactate provide neuroprotection in preclinical stroke models1,2. Hyperpolarized (HP) 13C MRI is a new way for real-time molecular imaging. The MR signal of both pyruvate and lactate can be hyperpolarized by dissolution DNP3,4.
We measured the cerebral metabolism of a therapeutic bolus of HP lactate or pyruvate in a transient mouse model of ischemic stroke to evaluate their potential as theranostic agents for stroke.

Methods

30’ transient middle cerebral artery occlusion (MCAO) was induced in C57BL6/J male mice (6-10 weeks) to model stroke.
A therapeutic dose of HP [1-13C]lactate (1.01±0.17umol/g, polarization: P=36%) or [1-13C]pyruvate (1.15±0.12umol/g, P=60%), was injected either 1h or 2h post-reperfusion in MCAO mice and 1h post-surgery in sham. Dynamic 13C MR spectra were acquired every 3s with 30° BIR-4 adiabatic pulses at 9.4T using a surface coil.
Kinetic modelling was performed on the time course of the metabolites. Metabolite ratios were computed on the summed signal of the first 120s post-injection.
Ratios and rate constants are scaled to the actual HP substrate dose and reported as mean ± standard deviation. Significance was tested by one-way ANOVA with Tukey-Kramer post-test.

Results/Discussion

HP lactate and pyruvate metabolism are observed after stroke (Fig.1b, f). While both substrates report on the same metabolic pathways and depicted similar trends of stroke-induced metabolic alterations, only changes with HP lactate were significant, showing lower pyruvate labeling after stroke (Figs.1c, 2b), faster pyruvate to bicarbonate conversion in MCAO 2h compared to sham (Fig.2c) and lower alanine labelling in MCAO 2h compared to MCAO 1h and sham (Fig.1e).
Unlike HP pyruvate, HP lactate allows to distinguish between transport and metabolism as the pyruvate labelling relates to lactate transport across the blood-brain barrier4, while its subsequent conversion into alanine or bicarbonate depict intracellular metabolic activity. Furthermore, at the given doses, lactate therapy is more physiological as endogenous lactate is higher than pyruvate5. These differences could contribute to the better distinction between stroke and sham animals when injecting HP lactate compared to pyruvate.

Conclusions

MR of the neuroprotective agent HP [1-13C]lactate highlights significant alteration of cerebral metabolism after transient cerebral ischemia, whereas moderate changes were depicted with HP pyruvate. Our next step will be to differentiate between damaged and healthy tissues metabolism to further characterize the properties of these potential HP theranostic probes.

Acknowledgment

The authors gratefully thank Prof. Rolf Gruetter for supporting this collaboration, Drs. Analina Da Silva and Stefan Mitrea for their assistance in the animal preparation, as well as Drs. Hongxia Lei and Bernard Lanz for fruitful discussions. This study is supported by the Swiss National Science Foundation (310030_170155), the Centre d’Imagerie Biomédicale of the University of Lausanne, École polytechnique fédérale de Lausanne, University of Geneva, Geneva University Hospitals, Lausanne University Hospital, and the Leenaards and Louis-Jeantet Foundations.

References
[1] Berthet, C. et al. Neuroprotective role of lactate after cerebral ischemia. J. Cereb. Blood Flow Metab. 29, 1780–1789 (2009).
[2] Yi, J. S., Kim, T. Y., Kyu Kim, D. & Koh, J. Y. Systemic pyruvate administration markedly reduces infarcts and motor deficits in rat models of transient and permanent focal cerebral ischemia. Neurobiol. Dis. 26, 94–104 (2007).
[3] Kurhanewicz, J. et al. Hyperpolarized 13 C MRI: Path to Clinical Translation in Oncology. Neoplasia (United States) 21, 1–16 (2019).
[4] Takado, Y. et al. Hyperpolarized 13 C Magnetic Resonance Spectroscopy Reveals the Rate-Limiting Role of the Blood-Brain Barrier in the Cerebral Uptake and Metabolism of l -Lactate in Vivo. ACS Chem. Neurosci. 9, 2554–2562 (2018).
[5] Soto, M. et al. Pyruvate induces torpor in obese mice. Proc. Natl. Acad. Sci. U. S. A. 115, 810–815 (2018).
Fig.1: Striatal lesion after 30’ MCAO, 13C MRS and metabolite ratios after HP infusion

(a) Evolution of striatal lesion after transient 30’ MCAO. 13C MRS after infusion of (b) lactate or (f) pyruvate in MCAO 1h. The sum (red) is scaled to 1/50 and 1/3 of the lactate and pyruvate peaks. (c) After injecting HP lactate, the pyruvate-to-lactate ratio (cPLR) decreases in 1h MCAO compared to sham (-54%), while (e) the alanine-to-lactate ratio (cALR) is lower in 2h MCAO compared to both 1h MCAO and sham (-45% and -54%), distinguishing between both MCAO timepoints. With HP pyruvate, (g) the lactate-to-pyruvate (cLPR) and (i) alanine-to-pyruvate ratios (cAPR) tend to decrease after MCAO.

Fig.2: Kinetic models and kinetic rate constants of HP lactate and pyruvate metabolism

Mathematical models for kinetic modelling of (a) [1-13C]lactate and (d) [1-13C]pyruvate metabolism. Steps are modelled as first-order reactions. (b) With HP lactate infusion, the rate of lactate-to-pyruvate turnover ckLP is lower in MCAO 1h and 2h compared to sham (-39% and -43%), while (c) the rate of pyruvate-to-bicarbonate conversion kPB is higher in MCAO 2h compared to sham (+108%). (e) After injection of [1-13C] pyruvate, the pyruvate-to-lactate rate (ckPL) tends to be lower after stroke while (f) the pyruvate-to-bicarbonate rate seems higher in MCAO 2h compared to sham (ckPB).

Keywords: Brain, DNP, Stroke, Kinetic modeling, hyperpolarization
12:54 p.m. PS 08-05

Imaging extracellular lactate produced by tumours in vivo using CEST-MRI and a paramagnetic shift reagent

Lei Zhang2, Sabrina H. L. Hoffmann1, Veronica Clavijo-Jordan4, 3, Alexander Funk3, Dean Sherry3, 2, Andre F. Martins1, 3, 2

1 University Hospital of Tuebingen, Werner Siemens Imaging Center, Tuebingen, Germany
2 University of Texas at Dallas, Chemistry and Biochemistry, Richardson, United States of America
3 UT Southwestern Medical Center, Advanced Imaging Research Center, Dallas, United States of America
4 Harvard Medical School, Athinoula A. Martinos Center for Biomedical Imaging, Boston, United States of America

Introduction

Glucose taken up by cancer cells is converted largely to lactate even in the presence of abundant oxygen although the amount of pyruvate diverted into the mitochondria is largely unknown.  Hence, a method for imaging actual extracellular lactate production by tumors is of critical importance. We report here the design of shift reagents (SR) that form complexes with lactate and shift the lactate –OH CEST signal to a different frequency far away from water.  Given that SR’s are confined to extracellular space, the resulting lactate –OH CEST signal becomes a specific cancer biomarker.

Methods

Several tris-amide derivatives of the common macrocyclic ligand, DO3A, were prepared and characterized.  The respevtive Yb3+ and Eu3+ complexes were confirmed by 1D and 2D NMR, X-ray crystallography and density functional theory (DFT) calculations. The CEST profiles were recorded and the pH and temperature dependencies of the proton exchange rate in each complex (kex) were determined using the Omega fitting method.  T1-weighted and CEST images of mice having xenograft tumors growing on a lower flank were recorded in vivo after injection of a lactate-specific SR. Healthy controls were also performed for the same regimens.

Results/Discussion

We demonstrate here that one can shift the –OH resonance of L-D-lactate far downfield from tissue water protons using Eu3+ and Yb3+-based paramagnetic shift reagent (SR) and then use the “on” versus “off” CEST response to quantify lactate produced by tumor cells growing in tissue culture (Figure 1). Depending upon which paramagnetic complex is used, the shifts can be quite large (50-130 ppm), moving the CEST signal of lactate well away from other confounding endogenous signals from tissues. Interestingly, the intensity of the lactate CEST signal was enhanced under slightly acidic conditions (similar to that produced by tumors) and was more intense at 37°C than at 25°C.  The potential of the SR was demonstrated by imaging excess lactate excreted into the bladder of a wild-type mouse and also by detecting extracellular lactate produced in subcutaneous cancer mouse models (Figure 1).

Conclusions

In conclusion, we have demonstrated for the first time that CEST MRI can detect extracellular lactate produced in subcutaneous tumour mouse models in vivo (Figure 1). These results provide the framework to develop and optimize new SRs to image lactate production in tumors by MRI. It also opens new venues towards understanding the tumour metabolism and the microenvironment in vivo.

AcknowledgmentFinancial support from the Junior Academy der Medizinischen Fakultät Tübingen, National Institutes of Health (DK-095416 and EB-015908), and the Robert A. Welch Foundation (AT-584) is gratefully acknowledged.
References
[1] Viswanathan S, Kovacs Z, Green KN, Ratnakar SJ, Sherry AD. Alternatives to Gadolinium-based MRI Metal Chelates. Chem Rev. 2010,110, 2960–3018.
[2] de Bari L, Moro L, Passarella S. Prostate cancer cells metabolize d-lactate inside mitochondria via a d-lactate dehydrogenase which is more active and highly expressed than in normal cells. FEBS Lett. 2013, 587, 467–473.
[3] Zhang L, Martins AF, Mai Y, Zhao P, Funk AM, Clavijo Jordan MV, Zhang S, Chen W, Wu Y, Sherry AD. Imaging Extracellular Lactate In Vitro and In Vivo Using CEST MRI and a Paramagnetic Shift Reagent. Chem – Eur J. 2017, 23,1752–1756.
[4] Zhang L, Martins AF, Zhao P, Tieu M, Esteban-Gomez D, McCandless GT, Platas-Iglesias C, Sherry AD. Enantiomeric recognition of D- and L-lactate by CEST with the aid of a paramagnetic shift reagent. J Am Chem Soc. 2017, 139, 48, 17431-17437
In vivo ShiftCEST imaging
Figure 1. (A) Axial view T1-weighted imaging and the respective (B and C) in vivo CEST MR imaging of mice bearing a small cell lung cancer (SCLC) tumor at the lower flank after an i.v. injection of SR.
Keywords: extracellular lactate, tumour microenvironemnt, Shift Reagent, CEST, Metabolism
1:06 p.m. PS 08-06

Imaging intra-tumour heterogeneity of glioblastoma metabolism

Anastasia Tsyben1, Jyotsna Rao1, Gregory Hamm4, Andreas Dannhorn5, Richard Mair1, 3, Stephanie Ling4, Alan Wright1, Maria Fala1, Richard Goodwin4, Kevin Brindle1, 2

1 University of Cambridge, 1. Cancer Research UK- Cambridge Institute, Cambridge, United Kingdom
2 University of Cambridge, 2. Department of Biochemistry, Cambridge, United Kingdom
3 University of Cambridge, 3. Division of Neurosurgery, Department of Clinical Neurosciences, Cambridge, United Kingdom
4 AstraZeneca, 4. Pathology, Drug Safety and Metabolism, IMED Biotech Unit, Cambridge, United Kingdom
5 Imperial College London, Department of Metabolism, Digestion and Reproduction, Faculty of Medicine, London, United Kingdom

On behalf of The Grand Challenge Collaboration, CRUK

Introduction

Glioblastoma multiforme (GBM) is an inherently heterogenous and invasive primary brain tumour, which spreads into the normal brain parenchyma, often beyond what can be captured with current clinical imaging1. To understand the intra- and inter-tumour heterogeneity of GBM, we performed metabolic tracing experiments in rats harbouring patient-derived orthotopic xenografts (PDOX). In addition, we analysed patient GBM samples following intra-operative multi-regional sampling. For both PDX and patient samples the sections were analysed using mass spectroscopy imaging (MSI).

Methods

Athymic nude rats were implanted orthotopically with A11 and S2 primary patient-derived GBM cell lines (n=4 per line). The presence of tumour was confirmed by T2-weighted MRI. Three animals per line were infused with [U-13C]glucose (0.4 mg/g bolus, 0.012 mg/g/min at 300 µL/h infusion)2. One animal from each cohort acted as a control and was infused with [12C]glucose.
Three patients with histologically confirmed GBM, underwent multi-regional sampling using Stealth navigation and 5-ALA fluorescence. Samples were immediately frozen in liquid nitrogen. All samples were run in negative ion mode DESI (35 µm). Analysis was performed using SCiLS Lab software (2019c, Bruker Germany).

Results/Discussion

Unsupervised spatial clustering of the MSI data segmented normal brain from tumour in the PDOX models. Both A11 and S2 tumours had high levels of 13C lactate and low levels of 13Cglucose, indicating rapid metabolic turnover of glucose via glycolysis. The margin around the lactate rich core had higher levels of TCA cycle intermediates. This was particularly the case for the S2 model, which demonstrated higher levels of fumarate and succinate. Bisecting k-means analysis showed significant metabolic differences between A11 and S2 tumours, with S2 tumours demonstrating higher levels of GSH and serine in the tumour core (Figure 1).
Human GBM core and margin sections also showed significant intra-tumoural heterogeneity. The TCA cycle metabolites, succinate and malate, showed differential signal within the sections. In addition, there was increased asparagine and hypotaurine in the tumour core, especially in areas of high proliferation (positive Ki67 staining, Figure 2).

Conclusions

MSI of human and PDOX sections demonstrate remarkable metabolic heterogeneity, particularly in the levels of glycolytic, TCA cycle, GSH and amino acid synthesis intermediates. The spatial extent of these metabolic territories is sufficiently that there is the potential for in vivo metabolic imaging, which has a lower spatial resolution than MSI. In addition, these data expose potential metabolic vulnerabilities for therapeutic development.

AcknowledgmentWe would like to thank the lab technicians at CRUK for ongoing animal care and support and Jodi Miller (Histocore, CRUK )for assisting with IHC. 
References
[1] Price, S. J. & Gillard, J. H. Imaging biomarkers of brain tumour margin and tumour
invasion. Br J Radiol 84 Spec No 2, S159-167, doi:10.1259/bjr/26838774 (2011).
[2] Maher, E. A. et al. Metabolism of [U-13 C]glucose in human brain tumors in vivo. NMR Biomed 25, 1234-1244, doi:10.1002/nbm.2794 (2012).
Figure 1
Coronal sections from [U-13C]glucose-infused rats with S2 and A11 tumors. 
Figure 2

Human GBM section from the core and margin of the tumour. Ki67 immunohistochemistry is shown for reference.

Keywords: Mass spectrometry imaging, DESI, GBM, Metabolism
1:18 p.m. PS 08-07

Combining 1H MRS and deuterium labeled glucose for mapping of neural metabolism in humans: pilot study

Laurie Rich1, Puneet Bagga1, Deepa Thakuri1, Ravi P. Nanga1, Abigail Cember1, Mark Elliot1, Mohammad Haris2, 3, John A. Detre4, Ravinder Reddy1

1 University of Pennyslvania, Radiology, Philadelphia, United States of America
2 Sidra Medicine, Doha, Qatar
3 LARC, Qatar University, Doha, Qatar
4 University of Pennyslvania, Neurology, Philadelphia, United States of America

Introduction

We developed a novel strategy named quantitative exchange label turnover (QELT) MRS (qMRS) that increases the range of applications for magnetic resonance based metabolic mapping without the requirement for specialized hardware. Similar to recently developed deuterium metabolic imaging (DMI)1, this technique relies on the administration of deuterium labeled glucose resulting in accumulation of downstream 2H labeled metabolites (Fig 1a). Since the 2H label is invisible on 1H MRS, accumulation of labeled metabolites leads to an overall reduction in 1H MRS signal of the corresponding metabolites.

Methods

The protocol was approved by the Institutional Review Board at the University of Pennsylvania. Written informed consent was obtained prior to the scans. The was performed on a 35 year old healthy male. MR experiments were performed on a 7T Siemens scanners using a 32-channel receive head coil. The healthy volunteer underwent overnight fasting before the MRSI studies. Approximately two hours after drinking ~58 grams of [6,6′-2H2]glucose in water,the subject was placed in the scanner and axial T1-weighted FLASH images were obtained to enable localization of the cortex. Following localization, MRSI datasets were acquired using a water suppressed semi-localization by adiabatic selective refocusing (semi-LASER) pulse sequence, in addition to SVS. Analysis was performed using LCModel and MATLAB.

Results/Discussion

This work builds on encouraging qMRS studies performed in animal models (in press)2. To assess the feasibility of qMRS based metabolic mapping, we acquired MRSI datasets in a human volunteer after drinking [6,6′-2H2]glucose in water. Fig 1b shows Glu to NAA metabolite ratio maps in the  male subject that received [6,6′-2H2]glucose. A central gray matter region (white outline) is observed on the anatomical region which, as expected, also corresponded to a high Glu/NAA ratio. In contrast, white matter regions on either side of this central gray matter region show considerably lower Glu/NAA ratio. Repeated baseline measurements acquired on separate days show good reproducibility. Interestingly, a large decrease in the Glu/NAA ratio for both grey and white matter regions was observed at two hours after drinking. Analysis of SVS data confirmed these findings (Fig 2), with a 1.4 mM (20%) decrease measured for the central grey matter region, and a 0.4 mM (11%) decrease in the white matter.

Conclusions

Together, these findings demonstrate the feasibility of performing 1H MRSI in conjunction with deuterium labeling to resolve spatial changes in metabolite concentrations over time. In addition to Glu, qMRS enables tracking of other important metabolites including, Gln, GABA, and Lac. This first-in-human study paves the way for continued develoment of qMRS to study metabolic derangments in brain pathologies, including neurodegeneration and cancer.

Acknowledgment

This work was carried out at a US National Institutes of Health–supported resource, with funding from: the NIBIB under Grant No. P41 EB015893, National Institute of Neurological Disorders and Stroke Award Number R01NS087516 and the training grant T32EB020087-02.

References
[1] De Feyter HM, Behar KL, Corbin ZA, Fulbright RK, Brown PB, McIntyre S, Nixon TW, Rothman DL, de Graaf RA. (2018) Deuterium metabolic imaging (DMI) for MRI-based 3D mapping of metabolism in vivo. Science advances. 4(8), eaat7314
[2] Rich LJ, Bagga P, Wilson NE, Schnall MD, Detra JA, Haris M, Reddy, ‘1H magnetic resonance spectroscopy of 2H-to-1H exchange quantifies the dynamics of cellular metabolism in vivo’, Nature Biomedical Engineering (In Press)
Figure 1

Fundamentals of qMRS. (a) Schematic shows pathway of exchange of deuterium from [6,6′- 2H2]glucose to downstream metabolites that can be detected with 1H MRS. One of the key metabolites labeled is glutamate (Glu, red box). (b) Anatomical reference image from the cortex in a human volunteer showing grey matter (GM) and white matter (WM) regions. Glu to n-acetyl aspartate (NAA) ratio maps are at two separate baseline measurements and two hours after drinking [6,6′- 2H2]glucose. A clear reduction in Glu/NAA levels is observed after drinking, indicating Glu labeling.

Figure 2
LCModel based fitting of single voxel spectroscopy data acquired in grey and white matter for the human volunteer before and two hours after drinking [6,6′- 2H2]glucose. A clear reduction in the 2.35 ppm Glu peak (blue line and arrow) is observed after drinking, with a more noticeable change observed for grey matter. Cre creatine, Glu glutamate, NAA N-acetyl aspartate
Keywords: Spectroscopy, Neuroimaging, Metabolism, Deuterium, Magnetic Resonance Spectroscopy