Received: 15 October 2024 Revised: 13 March 2025 Accepted: 14 March 2025 DOI: 10.1002/mrm.30519 R E S E A R C H A R T I C L E A pathway toward clinical translation of hyperpolarized [1,4-13C2,2,3-d2]fumarate as an imaging biomarker for early cellular necrosis in vivo Jonathan R. Birchall1 Pascal Wodtke1,2 Ashley Grimmer1,2 Esben S. S. Hansen3 Lotte B. Bertelsen3 Nikolaj Bøgh3 Marta Wylot1 Maria J. Zamora-Morales1 Otso Arponen1,4 Ines Horvat-Menih1 Elizabeth C. Latimer1 Fung Tan5 Evita Pappa6 Johann Graggaber7 Joseph Cheriyan6,7 Kelly Holmes2 Matthew J. Locke1 Helen Sladen2 Joan Boren8 Mikko I. Kettunen9,10 Anita Chhabra5 Ian B. Wilkinson6,7 Christoffer Laustsen3 Kevin Brindle2,10 Mary A. McLean1,2,10 Ferdia A. Gallagher1,2 Correspondence Ferdia A. Gallagher, Department of Radiology, University of Cambridge, Cambridge CB2 0QQ, UK. Email: fag1000@cam.ac.uk Funding information NIHR Cambridge Biomedical Research Centre, Grant/Award Number: BRC-1215-20014; Gates Cambridge Trust, Grant/Award Number: OPP1144; National Cancer Imaging Translational Accelerator (NCITA); The Mark Foundation Institute for Integrative Cancer Medicine; Wellcome Trust; Cancer Research UK Cambridge Institute, University of Cambridge; Cambridge Experimental Cancer Medicine Centre Abstract Purpose: The detection of hyperpolarized carbon-13 (HP 13C)-fumarate conver- sion to 13C-malate using 13C-MRSI is a biomarker for early detection of cellular necrosis. Here, we describe the translation of HP 13C-fumarate as a novel human imaging agent, including the evaluation of biocompatibility and scaling up of the hyperpolarization methods for clinical use. Methods: Preclinical biological validation was undertaken in fumarate hydratase-deficient murine tumor models and controls. Safety and biocompat- ibility of 13C-fumarate was assessed in healthy rats (N = 18) and in healthy human volunteers (N = 9). The dissolution dynamic nuclear polarization pro- cess for human doses of HP 13C-fumarate was optimized in phantoms. Finally, 2D 13C-MRSI following injection of HP 13C-fumarate was performed in an ischemia–reperfusion porcine kidney model (N = 6). Results: Fumarate-to-malate conversion was reduced by 42%–71% in the knock- down murine tumor model compared to wildtype tumors. Twice-daily injection of 13C-fumarate in healthy rats at the maximum evaluated dose (120 mg/kg/day) showed no significant persistent blood or tissue effects. Healthy human volun- teers injected at the maximum dose (3.84 mg/kg) and injection rate (5 mL/s) showed no statistically significant changes in vital signs or blood measurements Jonathan R. Birchall and Pascal Wodtke contributed equally to co-first authorship. Mary A. McLean and Ferdia A. Gallagher contributed equally to co-last authorship. For affiliations refer to page 12 This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2025 The Author(s). Magnetic Resonance in Medicine published by Wiley Periodicals LLC on behalf of International Society for Magnetic Resonance in Medicine. Magn Reson Med. 2025;1–16. wileyonlinelibrary.com/journal/mrm 1 https://orcid.org/0000-0003-3920-4038 https://orcid.org/0000-0002-6109-4261 https://orcid.org/0000-0001-6013-5271 https://orcid.org/0000-0001-5512-9870 https://orcid.org/0000-0002-2491-3184 https://orcid.org/0000-0002-0321-3269 https://orcid.org/0000-0002-1373-6944 https://orcid.org/0000-0002-2004-660X https://orcid.org/0000-0002-0317-2911 https://orcid.org/0000-0003-3883-6287 https://orcid.org/0000-0002-3752-0179 https://orcid.org/0000-0003-4784-5230 http://creativecommons.org/licenses/by/4.0/ http://wileyonlinelibrary.com/journal/MRM 2 BIRCHALL et al. 1 h post-injection. Spectroscopic evidence of fumarate-to-malate conversion was observed in the ischemic porcine kidney (0.96 mg/kg). Conclusion: HP 13C-fumarate has shown promise as a novel and safe hyperpo- larized agent for monitoring cellular necrosis. This work provides the basis for future imaging of HP 13C-fumarate metabolism in humans. K E Y W O R D S clinical translation, hyperpolarization, HP 13C-MRI, imaging biomarker, necrosis, preclinical 1 INTRODUCTION There is an unmet clinical need for the early detection of necrosis in diseases such as cancer, tissue ischemia, or infarction, and in metabolic disorders where early detec- tion of cell death may aid the identification of successful response to therapy. For example, clinical assessment of tumor shrinkage may take several weeks or months to become detectable, but imaging metabolic changes may highlight this much earlier.1,2 Hyperpolarized carbon-13 MRI (HP 13C-MRI) can facilitate the detection of tissue metabolism by increasing net nuclear spin magnetization by four to five orders of magnitude,3 enabling high SNR spectroscopic measurements and image contrast using a wide variety of probe molecules. Translating HP 13C-MRI into the clinical setting could improve patient stratifica- tion, enable early detection of treatment response, and therefore reduce the burden on healthcare services in the future. However, the roadmap for biomarker transla- tion from preclinical research to approval as a safe and reliable clinical decision-making tool is challenging, par- ticularly for intravenous contrast agents injected at high concentrations.4 The majority of clinical hyperpolarization studies have focused on [1-13C]pyruvate for assessing glycolytic metabolism.5 However, because the labeled carbon at the C1 position does not enter the tricarboxylic acid cycle, as is the case for [2-13C]pyruvate, this approach does not provide direct information on downstream metabolism.6 A range of promising additional molecules have been introduced preclinically,7 enabling pH imaging,6,8–10 redox state assessment,11–13 and probing of metabolic pathways beyond the capabilities of pyruvate.14–17 Conversion of fumarate into malate via the enzyme fumarate hydratase (FH, also termed fumarase in less complex eukaryotes18) is a promising biomarker of cell death.19 Fumarate is taken up slowly from the extra- cellular space in healthy cells; however, following dis- ruption of the cell membrane during cell death, FH is released into the extravascular space and extracellular fumarate rapidly enters the cell, facilitating metabolism.17 Glucose Pyruvate LDH Lactate Acetyl CoA CO2 Citrate Isocitrate α-Keto- glutarate Succinyl CoA Fumarate Malate Oxalo- acetate Succinate CO2 CO2 H2O FH TCA Cycle NADH NAD+ NAD+ FAD NAD+ healthy cell membrane necrotic cell membrane extracellular space cytoplasm H2O H2O Fumarate Fumarate FH minimal malate production rapid malate production FH F I G U R E 1 Fumarate-to-malate conversion via the enzyme FH (dashed box) within the TCA cycle, requiring only water as a cofactor for operation. Cross-membrane leakage of fumarate and FH in necrosis is illustrated in the inset box, forming the basis of the proposed biomarker imaging. FH, fumarate hydratase; TCA, tricarboxylic acid. Importantly, the enzyme only requires water as a cofactor and therefore remains active in the context of cell death and outside the confines of the cell membrane, enabling a necrosis biomarker producing positive contrast for cell death (Figure 1). BIRCHALL et al. 3 Initial studies demonstrated increased HP 13C-malate production in murine lymphomas following drug-induced necrosis,17 with subsequent studies in hepatocellular carcinoma.20 The probe has been shown to monitor tumor shrinkage in response to chemotherapeutic agents in murine models of implanted human breast cancer21 and renal cell carcinoma,22 as well as for assessing vascular dis- rupting agents in murine lymphoma23 and xenograft mod- els of colorectal cancer.24 In addition to oncological appli- cations, the probe has been used to detect intramuscular necrosis25 and myocardial infarction26 in rats, acute kid- ney injury in mice,27,28 and hypoxia-mediated cell death in a rat model of diabetic nephropathy at an early stage prior to the onset of significant histological changes.29 This study outlines a pathway for the potential future clinical application of HP 13C-fumarate-to-malate conver- sion as a biomarker for cellular necrosis in vivo, including the doses required to provide clinically useful infor- mation and associated safety aspects. We report on the toxicological and mutagenic safety of 13C-fumarate in ani- mals and describe the first injections in healthy human volunteers. Optimization of the 13C-fumarate hyperpolar- ization process and signal acquisition on a clinical MRI system to support future human imaging studies is dis- cussed. Lastly, we present 2D spectroscopic imaging of HP 13C-fumarate-to-malate conversion in phantoms and an in vivo large animal model of ischemic reperfusion injury (IRI) to approximate human imaging. This work is of importance in the field of HP 13C-MRI because very few molecules have crossed the translational gap into clinical use to date. In addition to 13C-labeled pyruvate at the C1 and the C2 positions,30–35 13C-labeled urea has been used as a biomarker for perfusion in prostate cancer,36 and [1-13C]α-ketoglutarate has been proposed as a probe for altered glutamate production in glioma.37 13C-fumarate has also now crossed the translational gap into humans. 2 METHODS 2.1 Evaluating HP fumarate-to-malate conversion in vivo using colorectal tumors after FH knockdown FH knockdown was performed using short hairpin RNAs (shRNA) to characterize differences in HP 13C-fumarate metabolism between FH knockdown tumors and con- trols. shRNA sequences were encoded in a DNA vector and introduced into a human colorectal model via plas- mid transfection (LoVo, ATCC, Manassas, VA); these were compared to control tumors utilizing an empty vector. LoVo (ATCC) tumors have previously been shown to pro- duce increased levels of malate following anti-vascular endothelial growth factor (VEGF) treatment.24 Tumors were grown by implanting ∼5× 106 cells subcutaneously into the flanks of female C57BL/6 NOD-SCID gamma mice (N = 7). [1,4-13C2]fumaric acid was dissolved in dimethyl sul- foxide (DMSO) containing a trityl radical (AH111501, GE HealthCare, Waukesha, WI) and was prepared using a dynamic nuclear polarization (DNP) hyperpolarizer (3.35 T, Oxford Instruments, Abingdon, UK). The sample was dissolved in a buffer containing 40 mM phosphate, 50 mM sodium chloride, and 40 mM sodium hydrox- ide at pH 7.4. The final injected concentration of HP 13C-fumarate was approximately 20 mM (see Ref. 17 for detailed experimental methods). 13C-MRS was performed using a 7 T horizontal bore magnet (Agilent Technologies, Santa Clara, CA) and an actively decoupled dual-tuned proton (1H)/13C volume transmit coil with a 20 mm diameter 13C-tuned surface receive coil (both Rapid Biomedical GmbH, Rimpar, Germany) placed over the tumor. Localization was deter- mined using 1H spin-echo imaging (see Ref. 38 for detailed experimental methods). For dynamic 13C-MRS, a 6 mm oblique coronal slice through the tumor was chosen. Fol- lowing this, 0.2 mL of HP 13C-fumarate was injected intra- venously into a tail vein over a period of 6 s, and the animal was immediately placed into the MRI scanner. Slice- selective free induction decays were acquired with a 10◦ flip angle RF pulse, TR= 3 s with 60 repetitions. Spectra were analyzed in the time domain using a Java-based MR user interface (jMRUI, http://www.jmrui.eu62,63), and sig- nal amplitudes for [1,4-13C2]fumarate and the combined signal from [1-13C]malate and [4-13C]malate were fitted. Enzymatic activity of FH in tumors was estimated from the cumulative ratio of 13C-malate/13C-fumarate over time in both the FH knockdown (N = 4) and empty vector (N = 3) animals. After baseline imaging, animals received an intraperitoneal injection of bevacizumab (Avastin; Roche, Basel, Switzerland), an anti-VEGF agent to induce tumor necrosis, at 5 mg/kg.24,39 Animals were subse- quently imaged at 48 and 72 h after treatment. 2.2 Toxicological studies in animals Fumarate is an endogenous molecule and was there- fore not expected to exhibit biological toxicity. However, because doses required for clinical imaging are supraphys- iological, an assessment of toxicity was performed. All investigations were performed using double 13C-labeled [1,4-13C2,2,3-d2]fumarate, with deuteration lengthening the spin lattice relaxation time and therefore enhancing the in vivo signal.40,41 Ames and mouse lymphoma assay testing42 were undertaken (Covance Laboratories, Huntingdon, UK) to http://www.jmrui.eu 4 BIRCHALL et al. assess mutagenic potential (see section S1), along with analysis of in vitro hemocompatibility of 13C-fumarate with rat blood in a 1.8% DMSO/0.08% AH111501/16.1% Trometamol buffer solution in water for injections to match the human formulation (see section S2). A repeated dose study with 13C-fumarate in buffer solution was per- formed to assess systemic toxicity by monitoring changes in peripheral blood measurements and urinalysis follow- ing two doses (1 h apart), administered on days 1 and 8 to Sprague Dawley rats (N = 18). Potential medium-term effects were assessed after a 2-week recovery period. Ani- mals were then culled to measure organ weights and note differences in macroscopic organ appearance and any histopathological changes. Platelet clumping and blood film assessments were undertaken to assess the effect of the 13C-fumarate formulation on coagulation. The highest dose formulation assessed was 12 mg/mL, and the maxi- mum dose volume for administration was 5 mL/kg twice daily. Therefore, the maximum dose assessed in this work was 60 mg/kg, representing a single exposure 15.6-fold above the maximum human doses described below. Alto- gether, three male and three female rats were injected at each individual dose level of 5, 30, and 60 mg/kg. The 13C-fumarate concentration in rat blood plasma was mea- sured 5 min after each injection. Additional information on the toxicology study design, as well as additional details of blood and urinary parameters assessed and organs and tissues inspected at necropsy, are available in sections S3 and S4, respectively. 2.3 First-in-human safety and tolerability study Nine healthy volunteers (6 male/3 female) received injec- tions at the National Institute for Health Research (NIHR) Cambridge Clinical Research Facility. The first three vol- unteers received a low concentration (20 mM, 0.96 mg/kg) dose of 13C-fumarate solution prepared under supervision of a pharmacist. Three volunteers then received a medium concentration (40 mM, 1.92 mg/kg) dose, with the final three receiving the maximum (80 mM, 3.84 mg/kg) dose. All volumes for injection were prepared as 0.4 mL/kg, and the three fumarate concentrations were injected at rates of 0.04, 0.4, and 5 mL/s. Participants were recruited and screened using blood tests; physical examination; medical history review; and monitoring of resting heart rate, tem- perature, and blood pressure. Favorable ethical approval for this study was awarded by East of England–Essex Research Ethics Committee and registered on a public website (REC: 20/EE/0090, IRAS number 266343). At the start of each day of injection, a sample dose of [1,4-13C2,2,3-d2]fumarate (Sigma-Aldrich, St. Louis, MO) was analyzed in a Good Clinical Laboratory Practice facil- ity (Cambridge University Hospitals Clinical Investigation Ward, Cambridge, UK) to ensure the concentration was within specified limits (±10% tolerance of the target dose; see section S5). Filter integrity and sample pH tests were performed to ensure contrast agent sterility and safety for injection (target pH between 6.0 and 9.0). 13C-fumarate was injected into the antecubital vein at the appropriate volume, flow rate, and dose concentration using a MedRad Spectris Solaris EP injection pump (Bayer AG, Leverkusen, Germany). After injection, participants were observed for a minimum of 1 hour while vital signs—temperature, blood pressure, heart rate, and oxygen saturation—were recorded. Blood samples were acquired before and after injection to assess safety parameters. Follow up by a clin- ician was performed within 72 h of injection, although a provision was in place for participants to contact a medical professional or study team member if required. 2.4 Hyperpolarization of [13C]fumarate at clinical doses Dissolution DNP requires a glassing matrix and free radical3: DMSO (Wak-Chemie Medical GmbH, Stein- bach, Germany) and AH111501 trityl radical (Syncom B.V., Groningen, Netherlands), respectively, were used in this study. A dose of 0.38 g of clinical-grade [1,4-13C2,2,3-d2]fumarate was first mixed with 0.66 g DMSO for 2 h at ∼30◦C. Afterward, 19.1 mg of AH111501 was added, and the mixture was stirred for another 2 h at ∼30◦C. The radical concentration was optimized (see below) for maximum polarization. The prepared sample was subsequently transferred into a cryovial and loaded into a 5 T SPINlab hyperpolarizer (GE HealthCare). 2.4.1 Optimization of polarization A pair of experiments were conducted in optimally hyper- polarized samples of 13C-fumarate formulation identical to those mentioned above to optimize spin-transfer effi- ciency. Microwave frequency was first incremented in steps of 1 MHz from 140.040 to 140.077 GHz (7.5 min acquisition time between frequencies). Then, microwave attenuation was varied over a range of 5 to 12 dB (step size 0.5 dB). Under each set of conditions, the steady-state HP 13C-fumarate signal was recorded. The maximum achiev- able HP 13C-fumarate signal was assessed at steady state following microwave irradiation at 140.055 GHz (chosen based on the optimal frequency for pyruvate on our sys- tem; see section S6) and with 8 dB attenuation in the SPIN- lab hyperpolarizer at 0.8 K using different concentrations BIRCHALL et al. 5 of AH111501 (17.5, 20, 25, and 30 mM). An otherwise fixed formulation of 0.38 g fumarate and 0.66 g DMSO with vary- ing radical content (16.7, 19.1, 23.9, and 28.7 mg) was used throughout. 2.4.2 Dissolution Hyperpolarized samples were dissolved in 51 mL of super- heated (>100◦C) and pressurized sterile water. In the acidic fluid, AH111501 precipitated and was filtered to reduce the final injected radical content to <10 μM. Sub- sequently, the sample was neutralized by a sterile buffer solution (Royal Free London NHS Foundation Trust, Pharmaceutical Quality Control Laboratory, London, UK; manufactured under Specials Manufacturing Licence MS11149) containing sterile water (25.5 mL), sodium hydroxide (11.35 g), and tromethamine (TRIS, 7.25 mL); pH= 13.4. The sample was passed through a sterile filter with a pore size of 0.22 μm and was drawn into a MedRad syringe (Bayer AG) while an aliquot (∼5 mL) was used for quality control prior to injection. The time for dissolution and sample ejection was ∼30 s. 2.4.3 Quality control (QC) A multi-probe QC unit was used to measure solution tem- perature as well as fumarate and AH111501 concentration prior to injection. Sample pH was measured spectropho- tometrically using a calibration file tailored for fumarate. QC was performed on a similar timescale as dissolution and ejection from the hyperpolarizer and was conducted simultaneously with sample transfer to either a 50-mL Fal- con tube (for phantom experiments) or a MedRad syringe pump (Bayer AG) for injection in vivo. Therefore, the sample was ready for injection after approximately 60 s. 2.5 Phantom studies HP 13C-fumarate imaging in phantoms was performed using a 26 cm inner diameter dual-tuned 1H/13C trans- mit/receive head coil (Rapid Biomedical GmbH) at ∼32.1 MHz and a field strength of 3 T (MR750, GE Health- Care). Unlocalized MRS applying 18 sequential scans of 32 μs pulse length was used to observe longitudinal relaxation of 13C nuclear spins at a TR of 10 s and over a 5 kHz bandwidth. T1 estimates were not corrected for RF losses but were assumed to be negligible due to the low nominal flip angle used (3◦). This was followed by a hard pulse 2D MRSI acquisition: FOV= 20 cm, TR= 116 ms, TE= 0.237 ms, pulse width= 100 μs, flip angle= 10◦, 10× 10 weighted circular k-space coverage with 61 tran- sients, 5 kHz bandwidth, 7 s total duration. Dotarem (Guerbet, Villepinte, France) was then added to the 13C-fumarate solution as a source of gadolinium contrast agent in a 1:200 ratio, and the Falcon tube was inverted several times to mix. This served to increase the spin relax- ation rate and enable calculation of thermal 13C polariza- tion using 12 transients at a 90◦ flip angle and 500 μs pulse length each, with a TR of 1 s as described in section S8. Lastly, a 3D fast spoiled gradient echo 1H pulse sequence (FOV= 20 cm, TR= 6.2 ms, TE= 1.88 ms, 64 locations, flip angle= 5◦, 2 averages, 42 s total duration) was acquired to facilitate coregistration of HP 13C images with structural 1H MRI. Spectroscopic imaging of FH activity at differing enzyme concentrations was also performed. Upon disso- lution of the HP 13C-fumarate solution, 15 mL was rapidly transferred into three separate 50-mL Falcon tubes, each containing either 0, 50, or 100 unit quantities of FH (Sigma-Aldrich), chosen to cover a wide range of physi- ological FH concentrations from normal tissue to tumor cells as reported previously.17,23,43 Tubes were inverted three times to ensure proper mixing prior to transfer into the MRI scanner, where an unlocalized MRS sequence (described above) was run every 10 s after insertion. Fumarate-to-malate conversion was assessed with the 2D dynamic MRSI sequence described previously. Peak iden- tification and integration over each voxel were performed using MatLab (version 2023b, MathWorks, Natick, MA), and intensity ratios of HP fumarate (single peak) and malate (two peaks, arising from the inequivalent 13C1 and 13C4 resonances) were compared to produce spatial maps of the malate/fumarate ratio in each phantom after∼6 min of metabolism. Spectrophotometric analysis at 290 nm was used to confirm these findings (see section S9). 2.6 Large animal studies For the porcine IRI experiments, hyperpolarization was performed using a SpinAligner DNP hyperpolarizer (Polarize, Frederiksberg, Denmark)44 operating at 6.7 T and 1.4 K. The vial content for hyperpolarization was 0.38 g [1,4-13C2,2,3-d2]fumarate/0.66 g DMSO/34.3 mg AH111501 (36 mM radical concentration after disso- lution). Injection was performed following dilution in 40 mL water for injection at a rate of 5 mL/s, followed by a 20 mL saline flush. Fumarate concentration was measured using a SpinSolve benchtop NMR spectrometer (Magritek GmbH, Aachen, Germany), comparing signal intensity against a linear fit of [1,4-13C2,2,3-d2]fumarate titrations at five different concentrations. 6 BIRCHALL et al. Imaging was performed 20 s after 13C-fumarate injection on a 3 T MRI scanner (Signa, GE Healthcare) using a single 20 mm slice CSI sequence (FOV= 28 cm, TR= 115 ms, TE= 2.71 ms, flip angle= 10◦, 20× 20 mm voxel size, 196 transients, 5 kHz bandwidth, 23 s total duration). A combination of clamshell transmit (Rapid Biomedical GmbH) and eight-channel flexible receiver array coil (JD Coils, Hamburg, Germany) were employed. Zero-filling was performed during reconstruction to dou- ble the 2D MRSI image resolution from 20× 20× 20 mm3 (14-by-14 voxels) to 10× 10× 20 mm3 (28-by-28 vox- els). Phase correction and peak fitting of fumarate and malate spectra were performed using the Oxford Spec- troscopy Analysis toolbox for MatLab (version 2023b, MathWorks).45 This study was undertaken in six pigs (all female, weight= 40± 2 kg) as approved by the Danish Animal Inspectorate (Ref. 2019-15-0201-00367). In each animal, the right kidney underwent 90 min of ischemia and sub- sequent warm IRI46 (see Ref. 47 for detailed experimental methods), and then was imaged after 3–4 h by intravenous injection of HP 13C-fumarate. No IRI was induced in the contralateral (left) kidney of each animal serving as a noninjured control. 3 RESULTS 3.1 Measurements of metabolism in a FH knockdown tumor model Figure 2 shows results from the in vivo tumor exper- iments. In control tumors (empty vector) where FH enzyme was present, a higher 13C-malate/13C-fumarate ratio was observed across all timepoints compared to knockdown tumors, indicating physiological conver- sion of fumarate to malate within the tumor. In the FH knockdown tumors, there was between 42% and 71% reduction in the 13C-malate/13C-fumarate ratio across all timepoints. A progressive increase in this ratio after 48 and 72 h of anti-VEGF therapy with bevacizumab indicated increasing necrosis over this period. The dif- ferences in 13C-malate/13C-fumarate at 48 and 72 h post-therapy were statistically significant as determined via a two-sample t-test with equal variance (p= 0.037 and 0.003, respectively). 3.2 Fumarate safety and toxicology 3.2.1 Preclinical assessment of toxicology Ames and mouse lymphoma assay testing showed no evidence of mutagenic activity (see Table S1). Platelet 0 48 72 Time post-treatment (hrs) 0 0.05 0.10 0.15 0.20 0.25 0.30 C um . 13 C -m al at e/ fu m ar at e ra tio Control empty vector FH shRNA knockdown 0.35 ns * ** F I G U R E 2 Comparison of the cumulative 13C-malate/13C-fumarate ratio over time following treatment with bevacizumab in control tumors (red, empty vector) and FH knockdown tumors (blue, shRNA for FH). 13C, carbon-13; FH, shRNA, short hairpin RNA. clumping assessments and blood film reviews showed no evidence of coagulation. Statistical analysis of in vitro hemocompatibility with rat blood suggested that injec- tions at a fumarate formulation/blood ratio of 1.35:1 may affect some coagulation parameters, including activated partial thromboplastin clotting time, prothrombin time, and fibrinogen concentrates (see Table S2). However, there was no evidence of hemolysis, suggesting that although the buffer solution may affect prothrombin time, the fumaric acid formulations had little effect on hemocompatibility. Elongated clotting times observed at a formulation/blood ratio of 2.70:1 were likely due to the large resultant dilu- tion factor. Platelet clumping assessments, blood film reviews, and hemolysis assessments showed no evidence of hemotoxicity. whilst [1,4-13C2,2,3-d2]fumarate was well tolerated, slight physiological changes were observed, including ele- vated white blood cell count, increased group mean alka- line phosphatase activity, elevated plasma phosphorus concentrations, and higher organ weights at doses of 60 or 120 mg/kg/day. Additionally, minimal diffuse cortical vac- uolation in the zona reticularis was seen in female rats only at very high doses of 120 mg/kg/day. This dose was 15.6-fold higher than the highest injected human dose and was therefore not deemed to be clinically relevant. Full recovery of these findings was observed after the 2-week recovery period. The no-observed-adverse-effect level, the experimental dose level immediately below that which produced a statistically significant increase in the rate of adverse effects observed relative to healthy control animals,48 was 120 mg/kg/day for both sexes in this study. BIRCHALL et al. 7 F I G U R E 3 Mean rat blood plasma concentrations of 13C-fumarate following administration of different doses of 13C-fumarate in solution after the: (A) first dose on days 1 and 8, (B) second dose on days 1 and 8, (C) first and second doses on the first day, (D) first and second doses on day 8. Error bars are included as the SD of the population average. The t-test p-values are quoted separately for male rats (pm) and female rats (pf ). 10 60 120 0 0.2 0.4 0.6 0.8 1.0 M ea n pl as m a co nc . ( m M ) 10 60 120 0 0.2 0.4 0.6 0.8 1.0 M ea n pl as m a co nc . ( m M ) Males; Day 1 Females; Day 1 Males; Day 8 Females; Day 8 10 60 120 13C-fumarate dose (mgkg-1 / day) 0 0.2 0.4 0.6 0.8 1.0 M ea n pl as m a co nc . ( m M ) DOSE 1 DOSE 2 DAY 1 DAY 8 Males; Dose 1 Females; Dose 1 Males; Dose 2 Females; Dose 2 Males; Day 1 Females; Dose 1 Males; Dose 2 Females; Dose 2 Males; Day 1 Females; Day 1 Males; Day 8 Females; Day 8 10 60 120 0 0.2 0.4 0.6 0.8 1.0 M ea n pl as m a co nc . ( m M ) pm = 0.4353 pf = 0.6293 pm = 0.2106 pf = 0.1788 pm = 0.1609 pf = 0.2778 pm = 0.3700 pf = 0.8937 13C-fumarate dose (mgkg-1 / day) 13C-fumarate dose (mgkg-1 / day) 13C-fumarate dose (mgkg-1 / day) (A) (B) (C) (D) More information on significant toxicity and pathological observations can be found in Tables S5–S11. Results from the repeated dose toxicity study are illus- trated in Figure 3. A series of paired, two-tailed t-tests determined that no statistical significance (0.16≤ p≤ 0.89) could be confirmed between 13C-fumarate concentration in the blood plasma and the size of the 13C-fumarate solu- tion dose over the 7 day period (Figure 3A,B), or over the 1 h period between doses (Figure 3C,D). This indicated rapid metabolism and/or excretion of 13C-fumarate with no significant accumulation in blood over time. 3.2.2 First-in-human healthy volunteer injections Figure 4 displays the recorded participant temperature, heart rate, blood pressure, and oxygenation, as well as blood biochemistry including sodium, bicarbonate, and glucose levels before and after injection as a function of 13C-fumarate dose and injection rate. All nine injections were within the acceptable pH range (min 7.23, max 8.29). Additional parameters assessed are included in Table S12 and Figure S1. No substantial changes in participant temperature, heart rate, blood pressure, or oxygenation were recorded over the 2 h post-injection. Comparison of blood sam- ples acquired from each participant before and after 13C-fumarate injection also demonstrated no statistically significant changes. However, a >10% increase in the quantity of unlabeled bicarbonate was observed in two of the nine participants (14% and 16%; see Figure 4T). Par- ticipants were otherwise clinically asymptomatic with no adverse events reported. 3.3 Optimization of [13C]fumarate polarization Peak signal intensity was recorded at 140.055 GHz (Figure 5A). Optimal signal amplitude and SNR were achieved at a microwave attenuation level of 8 dB (Figure 5B). Results from varying the radical concentra- tion are shown in Figure 5C. An example timecourse of 13C polarization buildup using optimal conditions is shown in Figure 5D: Hyperpolarization buildup curves at each radical concentration are shown in Figure S2. Comparable sweeps of signal intensity as a function of microwave frequency and attenuation for [1-13C]pyruvate are presented in Figure S3. At these optimized settings, a polarization of ∼19% was recorded from the sample containing 20 mM AH111501 radical, with an associated polarization buildup rate of 𝛾 = 0.026± 0.003 min−1. Significantly elevated buildup rates were observed at higher radical concen- trations, and these were accompanied by only marginal 8 BIRCHALL et al. F I G U R E 4 Participant vital measurements before and after injection with 13C-fumarate: (A–D) 0.96 mg/kg, (E–H) 1.92 mg/kg, (I–L) 3.84 mg/kg. Blood biochemistry measurements before and after injection of 13C-fumarate, (M–O) 0.96 mg/kg, (P–R) 1.92 mg/kg, (S–U) 3.84 mg/kg. Injection rates can be identified by line color throughout (see injection scheme in the top-left of the image). losses in the maximum polarization achieved: ∼14% at 𝛾 = 0.089± 0.023 min−1 when utilizing the 30 mM radical concentration. 3.4 Phantom imaging Dynamic MR spectroscopic measurements of HP fumarate and malate signals observed in the 50-mL Falcon tube phantom containing the highest FH concentration are shown in Figure 6A. The T1 of HP 13C-fumarate was calculated via exponential fitting to be ∼76 s, with mag- netization decaying to around 45% of its initial solid-state polarization during the ∼60 s dissolution and sample transfer time. The insert shows a roughly linear increase in malate/fumarate ratio over time, as determined by area-under-curve integration at each timepoint. Figure 6B demonstrates quantitative MRSI measurement of the malate/fumarate ratio in three otherwise identical phan- toms containing HP 13C-fumarate solution and different quantities of FH. Spectrophotometric characteriza- tion measuring the fumarate-to-malate conversion in a BIRCHALL et al. 9 F I G U R E 5 Optimization of microwave (A) frequency and (B) attenuation for hyperpolarization of 13C-fumarate on the SPINlab hyperpolarizer (GE HealthCare, Waukesha, WI). Vertical dashed red lines indicate chosen optimal values utilized in later experiments. (C) Steady-state 13C polarization and 13C polarization buildup rate of HP 13C-fumarate in the presence of various AH111501 radical concentrations. (D) Example 13C polarization buildup over time (20 mM radical variant pictured; other concentrations shown in Figure S3). HP, hyperpolarized. (A) (B) 140.04 140.05 140.06 140.07 140.08 Microwave Frequency (GHz) A m pl itu de (a .u .) 56789101112 Microwave Attenuation (dB) A m pl itu de (a .u .) 200 300 400 700 800 900 (C) (D) 16 18 20 22 24 26 28 30 32 Radical Concentration (mM) A m pl itu de (a .u .) 0 0.05 0.10 0.15 B ui ld -u p ra te (m in -1 ) Time (mins) A m pl itu de (a .u .) 0 400 800 1200 13Cmax = 1176 +/- 43 a.u. Tb = 38.9 +/- 4.3 mins = 0.026 +/- 0.003 min-1 0 50 100 150 0 500 1000 1500 DNP F I G U R E 6 (A) 13C MR spectra of HP fumarate and malate over time in the ∼7.7 U/mL FH-containing phantom (corresponding malate/fumarate ratio plot inset). The truncated signal from the HP [1,4-13C2,2,3-d2]fumarate is at 176.0 ppm, the signal from [1-13C]malate at 182.5 ppm, and the signal from [4-13C]malate at 181.4 ppm. (B) 1H structural image overlaid with a 2D colormap of HP malate/fumarate ratio per voxel in each of the three phantoms acquired after ∼6 min of enzyme activity, assessed by peak-area integration of the respective MRSI spectra. ∼5 U/mL enzyme-containing phantom over time can be found in Figure S4. 3.5 In vivo imaging of renal ischemia in a porcine model HP fumarate and malate peak integrals were measured across all voxels during the 2D MRSI acquisition to produce spatial colormaps of the individual fumarate and malate signal intensities per voxel (Figure 7A,B). Example HP 13C-fumarate and 13C-malate spectra obtained from single voxels are shown in Figure 7C. HP 13C-malate peaks were visible only in the ischemic kidney, at approximately +6 ppm relative to the corresponding HP 13C-fumarate peak. The malate/fumarate ratios per voxel across both the ischemic and contralateral normal kidney are shown in Figure 7D, overlaid on an anatomical 1H image. Polariza- tion at the point of dissolution was estimated to be ∼27%. A histogram plot of this data comparing the ischemic and 10 BIRCHALL et al. F I G U R E 7 Spatial colormaps showing the peak area integral by voxel for the: (A) 13C-fumarate peak, and (B) two 13C-malate peaks from 2D MRSI acquisition. (C) Example MR spectra of HP 13C-fumarate and 13C-malate in the ischemic right kidney (blue line) and contralateral healthy kidney (red line), 20 s after injection. The integration ranges for the respective spectra are indicated by dashed vertical lines. (D) Spatial colormap of the ratio of malate-to-fumarate peak area integrals outlined in displays (A, B), coregistered with the 1H MRI localizer. healthy kidneys, along with corresponding data from other animals imaged in the study, are shown in Figure S5. In the example presented, an elevated 13C-malate/13C-fumarate ratio was observed in the ischemic porcine kidney relative to the healthy control: 0.08 (𝜎 = 0.05) compared to 0.03 (𝜎 = 0.02) for all voxels considered above the noise threshold (i.e., those visible in Figure 7D). 4 DISCUSSION The results demonstrate the feasibility of using HP 13C-fumarate as a clinical probe for imaging cellular necrosis using MRI. We have optimized sample prepa- ration for clinical use and demonstrated the safety of the probe preclinically and in healthy human volun- teers. Furthermore, we have shown that knocking down FH in a murine tumor model results in a reduction in the 13C-malate/13C-fumarate ratio, confirming that the method measures enzyme activity within tissue. Finally, when utilizing the clinical dose preparation of HP 13C-fumarate in a large animal model, 13C-malate sig- nal was detected following ischemia–reperfusion injury, demonstrating the potential of this approach for detecting necrosis in patients in the future. The in vivo tumor experiments in mice showed a two- to threefold reduction in the 13C-malate/13C-fumarate ratio following knockdown of FH expression using shRNA. The smaller reduction in enzyme activity in vivo compared to in vitro following shRNA knockdown likely reflects the contributions from cells within the tumor microenvironment that have normal FH expression, such as the stroma, immune cells, or vascular tissue. Alterna- tively, there may be some restoration of FH activity in tumor cells due to clonal expansion of populations with higher FH expression. The malate signal increased follow- ing the administration of an antivascular agent in keep- ing with previous work,24 but this remained significantly lower in the knockdown tumors compared to the con- trol tumors over the three timepoints. These results show that HP 13C-fumarate can be used to measure tumor FH activity and changes with the onset of cell death—and also potential complexity of the biological origin of these measurements in that multiple tissue compartments may contribute to the detected signal. Injections into healthy volunteers showed no sig- nificant physiological changes, confirming the safety of fumarate as an endogenous molecule. The >10% increase in bicarbonate concentration in two of the participants may have arisen from the pH of the injected probe being at the upper end of the acceptable range in these cases (pH: BIRCHALL et al. 11 8.29, 8.20), resulting in a transient change in the physio- logical bicarbonate/carbon dioxide ratio to compensate by buffering these changes. The maximum achievable HP 13C signal was obtained using a radical concentration of 20 mM. However, the more rapid polarization buildup rate recorded using a rad- ical concentration of 30 mM (∼3.4-fold higher) may be more desirable for scaling up contrast agent production in a future clinical setting given the significant reduc- tion in sample polarization time. Previous characteriza- tion of [1-13C]pyruvate hyperpolarization indicated an optimal concentration of ∼25 mM AH111501,49 which is in reasonable agreement with our results. The observa- tion that the polarization buildup was efficient for both 13C-fumarate and 13C-pyruvate at identical microwave fre- quency and power implies the potential for co-polarization and sequential injection of these contrast agents to tar- get different metabolic processes.50 Simultaneous injec- tion has been demonstrated for HP 13C-labeled pyruvate and urea,20,25,36,51 although spatially resolving malate and pyruvate–hydrate peaks may be challenging at clinical field strengths in the case of 13C-labeled pyruvate and fumarate co-injection.52 The presence of increased HP 13C-malate signal in the ischemic pig kidney (Figure 7B,C) is likely to be as a result of either increased FH activity in the extracellular space or increased permeability of the membrane to fumarate following acute cellular necrosis. The dose used in these experiments approximates to those that will be used in future clinical studies and further supports the use of HP 13C-fumarate as a probe for measuring tissue necrosis in a wide range of large animal models and in patients. How- ever, we acknowledge the small sample size and variability of results between animals (Figure S5A–C) as a limitation of this work. HP 13C-fumarate has the potential to characterize cell death in many conditions such as cancer, ischemia, inflammation, transplant rejection, and infection, as well as to detect the necrosis associated with early response to a wide range of therapies in many diseases. Com- pared to existing preclinical HP 13C probes such as [1-13C]pyruvate, the lack of a requirement for additional cofactors such as nicotinamide adenine dinucleotide may be advantageous.22,27 However, it is important to also consider that, despite the signal boost from hav- ing two labeled carbon atoms, the inequivalent C1 and C4 nuclei within [1,4-13C2,2,3-d2]malate have different chemical shifts, which restrict the maximum achievable SNR because the signals are not directly additive. The lower solubility of fumarate relative to pyruvate presents an additional limitation on the maximum achievable probe concentration at the point of injection, although the concentrations utilized here are higher than those employed for the recent successful clinical translation of HP 13C-urea.36,51 Recent preclinical studies have shown that higher concentrations of HP 13C-fumarate may be achievable using parahydrogen-induced polarization53 or using d-DNP with a meglumine glassing matrix to improve solubility,54 although these approaches are yet to be scaled up to clinically relevant doses. An important question for future research is the tim- ing of the onset of necrosis because it is likely that HP 13C-fumarate will be most sensitive shortly after the onset of membrane permeability but before estab- lished necrosis has set in, when FH may either be excreted from the extracellular space or undergo proteol- ysis. An analogy of this temporal change in metabolism can be found when HP 13C-pyruvate is injected into patients undergoing neoadjuvant treatment for breast can- cer patients: The subsequent 13C-lactate signal produced may initially increase55 but then decreases after one cycle of treatment.35 Further information could be acquired by comparing the metabolism of dead cells using HP 13C-fumarate with live cell imaging using HP 13C-pyruvate or [18F]fluorodeoxyglucose uptake on positron emission tomography. Improvements in early detection of cel- lular necrosis in cancer and other diseases may have important implications for assessing experimental sys- temic treatments and monitoring existing therapies in the clinic. [2,3-d2]fumarate is an alternative probe that has been investigated with deuterium (2H) metabolic imaging (DMI) without the requirement for hyperpolarization.43,56 DMI with oral 2H-glucose has recently been translated to clinical field strength.57 Unlike HP 13C-MRI, the 2H tis- sue signal does not decrease rapidly, facilitating its use for detecting slower metabolism,56 although the longer metabolite wash-in time results in an increased time period for breakdown into further downstream products. In vivo linewidths are typically broader with DMI; there- fore, SNR may be lower compared to 13C-MRI. Preclinical studies using [2,3-d2]fumarate have shown a decreased SNR but an enhanced malate/fumarate ratio when com- pared to HP [1,4-13C2]fumarate.43,58,59 Future comparative studies of DMI and HP 13C DNP are required in a clinical setting to evaluate this in humans. 5 CONCLUSIONS The work presented here demonstrates the evaluation and development of HP [1,4-13C2,2,3-d2]fumarate as a new clinical hyperpolarized probe. HP 13C-fumarate has shown potential as an injectable MR contrast agent. We have undertaken the first-in-human injection of non-HP 13C-fumarate in healthy volunteers, optimized 12 BIRCHALL et al. the hyperpolarization of the molecule and the imag- ing protocol used, and illustrated its applicability on clinical imaging systems in an in vivo large animal model of ischemia–reperfusion. This workflow mir- rors the previous translational pathways for 13C-labeled pyruvate,30,60 urea,36,51 and α-ketoglutarate.37,61 As with these emerging probes, the next stage in the translation of HP 13C-fumarate is the imaging of HP 13C-fumarate metabolism in healthy human volunteers and subsequently patients. AFFILIATIONS 1Department of Radiology, University of Cambridge, Cambridge, UK 2Cancer Research UK Cambridge Centre, Cambridge, UK 3Department of Clinical Medicine, Aarhus University MR Research Centre, Aarhus, Denmark 4Institute of Clinical Medicine, University of Eastern Finland, Kuopio, Finland 5Radiopharmacy Department, Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK 6Division of Experimental Medicine and Immunotherapeutics, Department of Medicine, University of Cambridge, Cambridge, UK 7Cambridge Clinical Trials Unit, Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK 8The Discovery Centre, AstraZeneca, Cambridge Biomedical Campus, Cambridge, UK 9A.I. Virtanen Institute for Molecular Sciences, University of Eastern Finland, Kuopio, Finland 10Cancer Research UK Cambridge Institute, Cambridge, UK ACKNOWLEDGMENTS This research was supported by the National Institute for Health and Care Research (NIHR) Cambridge Biomedi- cal Research Centre (NIHR203312). The views expressed are those of the authors and not necessarily those of the NIHR or the Department of Health and Social Care. This research was additionally supported by Cancer Research UK, Cancer Research UK Cambridge Centre and Cam- bridge Institute, and The Wellcome Trust. The authors also acknowledge support from the National Cancer Imaging Translational Accelerator (NCITA) and The Mark Foun- dation Institute for Integrative Cancer Medicine (MFICM) at the University of Cambridge. p.w. acknowledges sup- port from the Gates Cambridge Trust (OPP1144). m.a.m. acknowledges support from the Cambridge Experimental Cancer Medicine Centre. The authors acknowledge the contributions of the Cancer Research UK Cambridge Insti- tute and Cambridge University Hospitals National Health Service (NHS) Foundation Trust Radiopharmacy Unit for their assistance with healthy human volunteer injections. j.r.b. and p.w. contributed equally as principal authors. m.a.m. and f.a.g. contributed equally as senior authors. CONFLICT OF INTEREST STATEMENT The authors acknowledge research support from GE HealthCare. FAG has grants from AstraZeneca and NVi- sion Imaging. ORCID Jonathan R. Birchall https://orcid.org/0000-0003-3920 -4038 Pascal Wodtke https://orcid.org/0000-0002-6109-4261 Ashley Grimmer https://orcid.org/0000-0001-6013-5271 Esben S. S. Hansen https://orcid.org/0000-0001-5512 -9870 Lotte B. Bertelsen https://orcid.org/0000-0002-2491 -3184 Nikolaj Bøgh https://orcid.org/0000-0002-0321-3269 Ines Horvat-Menih https://orcid.org/0000-0002-1373 -6944 Mikko I. Kettunen https://orcid.org/0000-0002-2004 -660X Christoffer Laustsen https://orcid.org/0000-0002-0317 -2911 Kevin Brindle https://orcid.org/0000-0003-3883-6287 Mary A. McLean https://orcid.org/0000-0002-3752 -0179 Ferdia A. Gallagher https://orcid.org/0000-0003-4784 -5230 REFERENCES 1. Therasse P, Arbuck SG, Eisenhauer EA, et al. New guidelines to evaluate the response to treatment in solid tumors. J Natl Cancer Inst. 2000;92:205-216. 2. Eisenhauer EA, Therasse P, Bogaerts J, et al. New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1). Eur J Cancer. 2009;45:228-247. 3. Ardenkjær-Larsen JH, Fridlund B, Gram A, et al. Increase in signal-to-noise ratio of >10,000 times in liquid-state NMR. Proc Natl Acad Sci U S A. 2003;100:10158-10163. 4. O’Connor JP, Aboagye EO, Adams JE, et al. Imaging biomarker roadmap for cancer studies. Nat Rev Clin Oncol. 2017;14:169-186. 5. Brindle KM, Bohndiek SE, Gallagher FA, Kettunen MI. Tumor imaging using hyperpolarized 13C magnetic resonance spec- troscopy. Magn Reson Med. 2011;66:505-519. 6. Gallagher FA, Kettunen MI, Day SE, et al. Magnetic resonance imaging of pH in vivo using hyperpolarized 13C-labelled bicar- bonate. Nature. 2008;453:940-943. 7. Wodtke P, Grashei M, Schilling F. Quo Vadis hyperpolarized 13C MRI? Z Med Phys. 2025;35:8-32. 8. Grashei M, Wodtke P, Skinner JG, et al. Simultaneous mag- netic resonance imaging of pH, perfusion and renal filtra- tion using hyperpolarized 13C-labelled Z-OMPD. Nat Commun. 2023;14:5060-5076. 9. Düwel S, Hundshammer C, Gersch M, et al. Imaging of pH in vivo using hyperpolarized 13C-labelled zymonic acid. Nat Com- mun. 2017;8:15126. https://orcid.org/0000-0003-3920-4038 https://orcid.org/0000-0003-3920-4038 https://orcid.org/0000-0003-3920-4038 https://orcid.org/0000-0002-6109-4261 https://orcid.org/0000-0002-6109-4261 https://orcid.org/0000-0001-6013-5271 https://orcid.org/0000-0001-6013-5271 https://orcid.org/0000-0001-5512-9870 https://orcid.org/0000-0001-5512-9870 https://orcid.org/0000-0001-5512-9870 https://orcid.org/0000-0002-2491-3184 https://orcid.org/0000-0002-2491-3184 https://orcid.org/0000-0002-2491-3184 https://orcid.org/0000-0002-0321-3269 https://orcid.org/0000-0002-0321-3269 https://orcid.org/0000-0002-1373-6944 https://orcid.org/0000-0002-1373-6944 https://orcid.org/0000-0002-1373-6944 https://orcid.org/0000-0002-2004-660X https://orcid.org/0000-0002-2004-660X https://orcid.org/0000-0002-2004-660X https://orcid.org/0000-0002-0317-2911 https://orcid.org/0000-0002-0317-2911 https://orcid.org/0000-0002-0317-2911 https://orcid.org/0000-0003-3883-6287 https://orcid.org/0000-0003-3883-6287 https://orcid.org/0000-0002-3752-0179 https://orcid.org/0000-0002-3752-0179 https://orcid.org/0000-0002-3752-0179 https://orcid.org/0000-0003-4784-5230 https://orcid.org/0000-0003-4784-5230 https://orcid.org/0000-0003-4784-5230 BIRCHALL et al. 13 10. Chen J, Hackett EP, Singh J, Kovács Z, Park JM. Simul- taneous assessment of intracellular and extracellular pH using hyperpolarized [1-13C]alanine ethyl ester. Anal Chem. 2020;92:11681-11686. 11. Batsios G, Taglang C, Cao P, et al. Imaging 6-phosphogluconolactonase activity in brain tumors in vivo using hyperpolarized δ-[1-13C]gluconolactone. Front Oncol. 2021;11:589570. 12. Bohndiek SE, Kettunen MI, Hu D, et al. Hyperpolarized [1-13C]-ascorbic and dehydroascorbic acid: vitamin C as a probe for imaging redox status in vivo. J Am Chem Soc. 2011;133:11795-11801. 13. Keshari KR, Kurhanewicz J, Bok R, Larson PEZ, Vigneron DB, Wilson DM. Hyperpolarized 13C dehydroascorbate as an endogenous redox sensor for in vivo metabolic imaging. Proc Natl Acad Sci U S A. 2011;108:18606-18611. 14. Chaumeil MM, Larson PEZ, Yoshihara HAI, et al. Non-invasive in vivo assessment of IDH1 mutational status in glioma. Nat Commun. 2013;4:2429-2440. 15. Keshari KR, Wilson DM, Chen AP, et al. Hyperpolarized [2-13C]-fructose: a hemiketal DNP substrate for in vivo metabolic imaging. J Am Chem Soc. 2009;131:17591-17596. 16. Moreno KX, Satapati S, DeBerardinis RJ, Burgess SC, Mal- loy CR, Merritt ME. Real-time detection of hepatic glu- coneogenic and glycogenolytic states using hyperpolarized [2-13C]dihydroxyacetone. J Biol Chem. 2014;289:35859-35867. 17. Gallagher FA, Kettunen MI, Hu D, et al. Production of hyperpo- larized [1,4-13C2]malate from [1,4-13C2]fumarate is a marker of cell necrosis and treatment response in tumors. Proc Natl Acad Sci U S A. 2009;106:19801-19806. 18. Yogev O, Yogev O, Singer E, et al. Fumarase: a mitochondrial metabolic enzyme and a cytosolic/nuclear component of the DNA damage response. PLoS Biol. 2010;8:e1000328. 19. Mescam M, Vinnakota KC, Beard DA. Identification of the catalytic mechanism and estimation of kinetic parameters for fumarase. J Biol Chem. 2011;286:21100-21109. 20. Düwel S, Durst M, Gringeri CV, et al. Multiparametric human hepatocellular carcinoma characterization and therapy response evaluation by hyperpolarized 13C MRSI. NMR Biomed. 2016;29:952-960. 21. Witney TH, Brindle KM. Imaging tumour cell metabolism using hyperpolarized 13C magnetic resonance spectroscopy. Biochem Soc Trans. 2010;38:1220-1224. 22. Mignion L, Dutta P, Martinez GV, Foroutan P, Gillies RJ, Jordan BF. Monitoring chemotherapeutic response by hyper- polarized 13C-fumarate MRS and diffusion MRI. Cancer Res. 2014;74:686-694. 23. Bohndiek SE, Kettunen MI, Hu D, et al. Detection of tumor response to a vascular disrupting agent by hyperpolar- ized 13C magnetic resonance spectroscopy. Mol Cancer Ther. 2010;9:3278-3288. 24. Bohndiek SE, Kettunen MI, Hu D, Brindle KM. Hyperpolar- ized 13C spectroscopy detects early changes in tumor vascu- lature and metabolism after VEGF neutralization. Cancer Res. 2012;72:854-864. 25. Eldirdiri A, Clemmensen A, Bowen S, Kjær A, Ardenkjær-Larsen JH. Simultaneous imaging of hyperpolar- ized [1,4-13C2]fumarate, [1-13C]pyruvate and 18F–FDG in a rat model of necrosis in a clinical PET/MR scanner. NMR Biomed. 2017;30:e3803. doi:10.1002/nbm.3803 26. Miller JJ, Lau AZ, Nielsen PM, et al. Hyperpolarized [1,4-13C2]fumarate enables magnetic resonance-based imag- ing of myocardial necrosis. JACC Cardiovasc Imaging. 2018;11:1594-1606. 27. Clatworthy MR, Kettunen MI, Hu D, et al. Magnetic resonance imaging with hyperpolarized [1,4-13C2]fumarate allows detec- tion of early renal acute tubular necrosis. Proc Natl Acad Sci U S A. 2012;109:13374-13379. 28. Nielsen PM, Eldirdiri A, Bertelsen LB, et al. Fumarase activity: an in vivo and in vitro biomarker for acute kidney injury. Sci Rep. 2017;7:40812. 29. Laustsen C, Nielsen PM, Qi H, Løbner MH, Palmfeldt J, Ber- telsen LB. Hyperpolarized [1,4-13C]fumarate imaging detects microvascular complications and hypoxia mediated cell death in diabetic nephropathy. Sci Rep. 2020;10:9650. 30. Nelson SJ, Kurhanewicz J, Vigneron DB, et al. Metabolic imaging of patients with prostate cancer using hyperpolarized [1-13C]pyruvate. Sci Transl Med. 2013;5:198ra108. 31. Aggarwal R, Vigneron DB, Kurhanewicz J. Hyperpolarized [1-13C]pyruvate magnetic resonance imaging detects an early metabolic response to androgen ablation therapy in prostate cancer. Eur Urol. 2017;72:1028-1029. 32. Tran M, Latifoltojar A, Neves JB, et al. First-in-human in vivo non-invasive assessment of intra-tumoral metabolic heterogene- ity in renal cell carcinoma. BJR Case Rep. 2019;5:20190003. 33. Chung BT, Chen H-Y, Gordon J, et al. First hyperpolarized [2-13C]pyruvate MR studies of human brain metabolism. J Magn Reson. 2019;309:106617-106623. 34. Stødkilde-Jørgensen H, Laustsen C, Hansen ESS, et al. Pilot study experiences with hyperpolarized [1-13C]pyruvate MRI in pancreatic cancer patients. J Magn Reson Imaging. 2020;51:961-963. 35. Woitek R, McLean MA, Gill AB, et al. Hyperpolarized 13C MRI of tumor metabolism demonstrates early metabolic response to neoadjuvant chemotherapy in breast cancer. Radiol Imaging Cancer. 2020;2:e200017. 36. Liu X, Tang S, Mu C, et al. Development of specialized mag- netic resonance acquisition techniques for human hyperpolar- ized [13C, 15N2]urea + [1-13C]pyruvate simultaneous perfusion and metabolic imaging. Magn Reson Med. 2022;88:1039-1054. 37. Kim Y, Dang D, Slater J, et al. Development for hyperpolar- ized [1-13C]alpha-ketoglutarate MRI for metabolic imaging of normal volunteers and mutant IDH glioma patients. In Proceed- ings of the Joint ISMRM & ISMRT Annual Meeting, Toronto, Ontario, Canada, 2023. p. 0843. 38. Gallagher FA, Sladen H, Kettunen MI, et al. Carbonic anhy- drase activity monitored in vivo by hyperpolarized 13C-magnetic resonance spectroscopy demonstrates its importance for pH reg- ulation in tumors. Cancer Res. 2015;75:4109-4118. 39. Gerber H-P, Ferrara N. Pharmacology and pharmacodynamics of bevacizumab as monotherapy or in combination with cyto- toxic therapy in preclinical studies. Cancer Res. 2005;65:671-680. 40. Chen AP, Keshari KR, Cunningham CH. Slowing T1 relaxation of hyperpolarized [2-13C]pyruvate with deuterium enrichment. In Proceedings of the 29th Annual Meeting and Exhibition of ISMRM & SMRT, An Online Experience, May 15-20, 2021. p. 3803. 41. Taglang C, Korenchan DE, von Morze C, et al. Late-stage deuter- ation of 13C-enriched substrates for T1 prolongation in hyperpo- larized 13C MRI. Chem Commun. 2018;54:5233-5236. http://dx.doi.org/0 14 BIRCHALL et al. 42. Ames BN, McCann J, Yamasaki E. Methods for detecting carcinogens and mutagens with the Salmonella/mammalian- microsome mutagenicity test. Mutat Res. 1975;31:347-364. 43. Hesse F, Somai V, Kreis F, Bulat F, Wright AJ, Brindle KM. Monitoring tumor cell death in murine tumor model using deuterium magnetic resonance spectroscopy and spectroscopic imaging. Appl Biol Sci. 2021;118:e2014631118. doi:10.1073/pnas .2014631118 44. Ardenkjær-Larsen JH, Bowen S, Petersen JR, et al. Cryogen-free dissolution dynamic nuclear polarization polarizer operating at 3.35 T, 6.70 T, and 10.1 T. Magn Reson Med. 2018;81:2184-2194. 45. Purvis LAB, Clarke WT, Biasiolli L, Valkovič L, Robson MD, Rodgers CT. OXSA: an open-source magnetic reso- nance spectroscopy analysis toolbox in MATLAB. PLoS One. 2017;12:e0185356. 46. Orvieto MA, Tolhurst SR, Chuang MS, et al. Defining maximal renal tolerance to warm ischemia in porcine laparoscopic and open surgery model. Urology. 2005;66:1111-1115. 47. Rasmussen CW, Bøgh N, Bech SK, et al. Fibrosis imaging with multiparametric proton and sodium MRI in pig injury models. NMR Biomed. 2023;36:e4838. 48. Leisenring W, Ryan L. Statistical properties of the NOAEL. Regul Toxicol Pharmacol. 1992;15:161-171. 49. Jähnig F, Kwiatowski G, Däpp A, et al. Dissolution DNP using trityl radicals at 7 T field. Phys Chem Chem Phys. 2017;19:19196-19204. 50. Wilson DM, Keshari KR, Larson PE, et al. Multi-compound polarization by DNP allows simultaneous assessment of multi- ple enzymatic activities in vivo. J Magn Reson. 2010;205:141-147. 51. Qin H, Tang S, Riselli AM, et al. Clinical translation of hyperpo- larized 13C pyruvate and urea MRI for simultaneous metabolic and perfusion imaging. Magn Reson Med. 2022;87:138-149. 52. Feuerecker B, Durst M, Michalik M, et al. Hyperpolarized 13C diffusion MRS of co-polarized pyruvate and fumarate to mea- sure lactate export and necrosis. J Cancer. 2017;8:3078-3085. 53. Gierse M, Nagel L, Keim M, et al. Parahydrogen-polarized fumarate for preclinical in vivo metabolic magnetic resonance imaging. J Am Chem Soc. 2023;145:5960-5969. 54. Huynh MT, Erfani Z, Al Nemri S, Chirayil S, Kovacs Z, Park JM. Enhanced solubility and polarization of 13C-fumarate with meglumine allows for in vivo detection of gluco- neogenic metabolism in kidneys. ACS Appl Mater Interfaces. 2024;16:37435-37444. 55. Woitek R, McLean MA, Ursprung S, et al. Hyperpolarized Carbon-13 MRI for early response assessment of neoadju- vant chemotherapy in breast cancer patients. Cancer Res. 2021;81:6004-6017. 56. De Feyter HM, Behar KL, Corbin ZA, et al. Deuterium metabolic imaging (DMI) for MRI-based 3D mapping of metabolism in vivo. Sci Adv. 2018;4:eaat7314. 57. Kaggie JD, Khan AS, Matys T, et al. Deuterium metabolic imag- ing and hyperpolarized (13)C-MRI of the normal human brain at clinical field strength reveals differential cerebral metabolism. Neuroimage. 2022;257:119284. 58. Hesse F, Wright A, Bulat F, Kreis F, Brindle K. Assessment of the sensitivity of 2H MR spectroscopy measurements of [2,3-2H2]fumarate metabolism for detecting tumor cell death. NMR Biomed. 2023;36:e4965. 59. Hesse F, Wright AJ, Bulat F, Somai V, Kreis F, Brindle KM. Deuterium MRSI of tumor cell death in vivo following oral delivery of 2H-labeled fumarate. Magn Reson Med. 2022; 88:2014-2020. 60. Chen AP, Albers MJ, Cunningham CH, et al. Hyperpolarized C-13 spectroscopic imaging of the TRAMP mouse at 3T—initial experience. Magn Reson Med. 2007;58:1099-1106. 61. Hong D, Batsios G, Viswanath P, et al. Acquisition and quantifi- cation pipeline for in vivo hyperpolarized 13C MR spectroscopy. Magn Reson Med. 2022;87:1673-1687. 62. Naressi, A., Couturier, C., Devos, J. M., Janssen, M., Mangeat, C., Beer, R. de, & Graveron-Demilly, D. (2001). Java-based graphi- cal user interface for the MRUI quantitation package. Magnetic Resonance Materials in Physics, Biology and Medicine, 12(2–3). doi:10.1007/bf02668096 63. Stefan, D., Cesare, F. D., Andrasescu, A., Popa, E., Lazariev, A., Vescovo, E., Strbak, O., Williams, S., Starcuk, Z., Cabanas, M., van Ormondt, D., & Graveron-Demilly, D. (2009). Quantitation of magnetic resonance spectroscopy signals: the jMRUI software package. Measurement Science and Technology, 20(10), 104035. doi:10.1088/0957-0233/20/10/104035 SUPPORTING INFORMATION Additional supporting information may be found in the online version of the article at the publisher’s website. FIGURE S1. Peripheral blood hematology and blood chemistry measurements obtained from the healthy human volunteer population before and after injection with 13C-fumarate at various dose levels and flow rates. FIGURE S2. Optimisation of microwave (A) frequency and (B) attenuation for hyperpolarization of 13C-pyruvrate on the SPINlab hyperpolarizer. Vertical dashed red lines correspond to the values utilized in 13C-fumarate experi- ments described in the main manuscript. FIGURE S3. Polarization build-up curves acquired from otherwise identical (0.38 g fumarate, 0.66 g DMSO formulation) samples of 13C-fumarate as a function of AH111501 radical concentration utilized: (A) 17.5 mM; (B) 20 mM; (C) 25 mM; (D) 30 mM. Microwave frequency= 140.055 GHz, 8 dB attenuation, temperature= 0.8 K. FIGURE S4. (A) Optical spectra acquired from a 1 mL sample of 13C-fumarate as a function of time following addition of 5 UmL−1 FH; (B) corresponding 13C-fumarate fraction of the total mixture at each time point as deter- mined by area under curve integration of the optical spectra. FIGURE S5. Spatial colormaps showing the measured malate-to-fumarate ratio in subjects #1 (A) and #5 (B); (C) Difference in mean malate-to-fumarate ratio between ischemic and healthy kidneys in all animals; (D) Histograms comparing malate-to-fumarate ratio in the ischemic and contralateral healthy kidneys for subject #6 (corresponding 2D MRSI colormap shown in Figure 7D of the main text). http://dx.doi.org/0 http://dx.doi.org/0 http://dx.doi.org/0 http://dx.doi.org/0 BIRCHALL et al. 15 TABLE S1. Cell cultures investigated in the 13C-fumarate mutagenic potential study. NT, not tested. TABLE S2. Statistical two-tailed t-test analysis of the effect of different 13C-fumarate formulations on blood coagulation parameters (partial thromboplastin time, PT; activated partial thromboplastin time, APTT; Clauss Fibrinogen, FIBC). Values denoted with an asterisk represent statistically significant comparisons between blood-to-formulation ration of 1:2.70, which were not to be considered relevant for in vivo human imaging studies where the maximum infusion rate was defined to be 1:1.35. NVR, no valid result for this measurement. TABLE S3. Sprague Dawley rat population groups inves- tigated in the 13C-fumarate toxicology study. TABLE S4. 13C-fumarate formulations investigated dur- ing the toxicology study. TABLE S5. Statistically significant observations from peripheral blood hematology. Values listed as population means as recorded on Day 8 of the study (after receiv- ing 4 total injections of either saline control, vehicle, or 13C-fumarate). Asterisks denote p-values<0.05 (*) or<0.01 (**) for comparisons against the saline control (Group 1). Hashes (# and ##) denote corresponding p-values for com- parison against the vehicle (Group 2). No statistically sig- nificant observations were observed between any groups for the following parameters: Hct, RBC, Retic, MCHC, MCV, RDW, Plt, PT, APTT. Additionally, no statistically significant observations were noted for any parameter from female rats in Group 3 (5 mg/kg 13C-fumarate twice daily). TABLE S6. Statistically significant observations from peripheral blood hematology during the second week of the recovery period. Values listed as population means. Asterisks denote p-values <0.05 (*) or <0.01 (**) for com- parisons against the saline control (Group 1). A hash (#) denotes corresponding p-values for comparison against the vehicle (Group 2). No statistically significant observa- tions were observed between any groups for the following parameters: Hct, RBC, Retic, MCH, WBC, N, L, E, B, M, LUC, PT, APTT (t-test for Group 5 with Group 1 and Group 2, and for Group 2 with Group 1). TABLE S7. Statistically significant observations from blood chemistry. Values listed as population means as recorded on Day 8 of the study (after receiving four total injections of either saline control, vehicle, or 13C-fumarate). Asterisks denote p-values<0.05 (*) or<0.01 (**) for comparisons against the saline control (Group 1). Hashes (# and ##) denote corresponding p-values for com- parison against the vehicle (Group 2). No statistically sig- nificant observations were observed between any groups for the following parameters: ALT, AST, Urea, Chol, Trig, Cl, Total Prot, A/G (Williams’ test for Groups 3–5 with Group 1, t-test for Group 2 with Group 1). TABLE S8. Statistically significant observations from blood chemistry during the second week of the recov- ery period. Values listed as population means. Asterisks denote p-values <0.05 (*) or <0.01 (**) for comparisons against the saline control (Group 1). No statistically sig- nificant comparisons against the vehicle control (Group 2) were observed. Additionally, no statistically significant observations were observed between any groups for the following parameters: ALT, AST, Create, Gluc, Trig, Na, K, Cl, Ca, Phos, Total Prot, Alb, A/G (t-test for Group 5 with Group 1 and Group 2, and for Group 2 with Group 1). TABLE S9. Statistically significant observations from uri- nalysis during the second week of the recovery period. Val- ues listed as population means. Asterisks denote p-values <0.05 (*) or <0.01 (**) for comparisons against the saline control (Group 1). No statistically significant comparisons against the vehicle control (Group 2) were observed. Addi- tionally, no corresponding statistically significant observa- tions were observed between any groups for any parame- ters in the male populations. TABLE S10. Statistically significant observations from organ weighing at necropsy. Values listed as adjusted pop- ulation means as recorded on Day 8 of the study (after receiving four total injections of either saline control, vehi- cle, or 13C-fumarate). Asterisks denote p-values <0.05 (*) or <0.01 (**) for comparisons against the saline control (Group 1). A hash (#) denotes corresponding p-values for comparison against the vehicle (Group 2). No statisti- cally significant observations were observed between any groups for terminal body weight, nor for the following organs: adrenals, brain, heart, kidneys, ovaries, pituitary, seminal vesicles, spleen, testes, thymus, thyroid or uterus. Additionally, no statistically significant observations were noted for any organ within the male rat populations in Group 3 (5 mg/kg 13C-fumarate twice daily) and Group 4 (10 mg/kg 13C-fumarate twice daily). TABLE S11. Statistically significant observations from organ weighing at necropsy. Values listed as adjusted pop- ulation means as recorded during the second week of the recovery period. Asterisks denote p-values <0.05 (*) or <0.01 (**) for comparisons against the saline control (Group 1). A hash denotes corresponding p-values for com- parison against the vehicle (Group 2). No statistically sig- nificant observations were observed between any groups 16 BIRCHALL et al. for the following organs: adrenals, brain, epididymides, heart, kidneys, ovaries, pituitary, prostate, seminal vesi- cles, spleen or uterus. TABLE S12. Repeatability of 13C-fumarate formulation to various target dose concentrations in the healthy human volunteer injections. 0.96 mg/kg corresponds to a concen- tration of 20 mM, 1.92 mg/kg to 40 mM, and 3.84 mg/kg to 80 mM. How to cite this article: Birchall JR, Wodtke P, Grimmer A, et al. A pathway toward clinical translation of hyperpolarized [1,4-13C2,2,3-d2]fumarate as an imaging biomarker for early cellular necrosis in vivo. Magn Reson Med. 2025;1-16. doi: 10.1002/mrm.30519 A pathway toward clinical translation of hyperpolarized [1,4-13C2,2,3-d2]fumarate as an imaging biomarker for early cellular necrosis in vivo 1 INTRODUCTION 2 METHODS 2.1 Evaluating HP fumarate-to-malate conversion in vivo using colorectal tumors after FH knockdown 2.2 Toxicological studies in animals 2.3 First-in-human safety and tolerability study 2.4 Hyperpolarization of [13C]fumarate at clinical doses 2.4.1 Optimization of polarization 2.4.2 Dissolution 2.4.3 Quality control (QC) 2.5 Phantom studies 2.6 Large animal studies 3 RESULTS 3.1 Measurements of metabolism in a FH knockdown tumor model 3.2 Fumarate safety and toxicology 3.2.1 Preclinical assessment of toxicology 3.2.2 First-in-human healthy volunteer injections 3.3 Optimization of [13C]fumarate polarization 3.4 Phantom imaging 3.5 In vivo imaging of renal ischemia in a porcine model 4 DISCUSSION 5 CONCLUSIONS Affiliations ACKNOWLEDGMENTS CONFLICT OF INTEREST STATEMENT ORCID REFERENCES Supporting Information