Hyperpolarized Carbon-13 MRI in Breast Cancer

Abstract

One of the hallmarks of cancer is metabolic reprogramming, including high levels of aerobic glycolysis (the Warburg effect). Pyruvate is a product of glucose metabolism, and ¹³C-MR imaging of the metabolism of hyperpolarized (HP) [1-¹³C]pyruvate (HP ¹³C-MRI) has been shown to be a potentially versatile tool for the clinical evaluation of tumor metabolism. Hyperpolarization of the ¹³C nuclear spin can increase the sensitivity of detection by 4-5 orders of magnitude. Therefore, following intravenous injection, the location of hyperpolarized ¹³C-labeled pyruvate in the body and its subsequent metabolism can be tracked using ¹³C-MRI. Hyperpolarized [¹³C]urea and [1,4-¹³C₂]fumarate are also likely to translate to the clinic in the near future as tools for imaging tissue perfusion and post-treatment tumor cell death, respectively. For clinical breast imaging, HP ¹³C-MRI can be combined with ¹H-MRI to address the need for detailed anatomical imaging combined with improved functional tumor phenotyping and very early identification of patients not responding to standard and novel neoadjuvant treatments. If the technical complexity of the hyperpolarization process and the relatively high associated costs can be reduced, then hyperpolarized ¹³C-MRI has the potential to become more widely available for large-scale clinical trials.

Keywords: breast cancer; carbon-13; hyperpolarization; magnetic resonance imaging; metabolism.

Figure 1

Dynamic Nuclear Polarization (DNP), and intracellular uptake and enzymatic conversion of [1-¹³C]pyruvate to lactate. (a) During DNP, the solution containing [1-¹³C]pyruvate, is doped with a stable radical containing unpaired electrons (electron paramagnetic agent, EPA), which become polarized when frozen at ~1 K (−272 °C) and exposed to a strong magnetic field (3.35–5 T). Microwave irradiation transfers spin polarization from the electron spins to the ¹³C nuclei. The hyperpolarized frozen sample is then rapidly dissolved using superheated water. Before intravenous injection into a cubital vein, the EPA is removed, and its concentration, as well as the pyruvate concentration, pH, and temperature, are assessed. (b) After intravenous injection and transit to the tissue of interest, the hyperpolarized [1-¹³C]pyruvate is taken up intracellularly by the monocarboxylate transporters (MCTs). Lactate dehydrogenase (LDH) in the cytosol catalyzes the exchange of the hyperpolarized ¹³C-label between pyruvate and the endogenous lactate pool.

Figure 2

Imaging with hyperpolarized [1-¹³C]pyruvate detects response and resistance to treatment with a PI3K inhibitor in ER+ breast cancer patient-derived xenografts. In drug-sensitive xenografts, the PI3K inhibitor inhibited phosphorylation of FOXO3A, which migrates into the nucleus inhibiting the expression of the transcription factor FOXM1, which in this tumor type drives the expression of LDHA. The resulting decrease in LDHA expression results in decreased hyperpolarized ¹³C label exchange between injected [1-¹³C]pyruvate and the endogenous tumor lactate pool. There was no change in the expression of the c-Myc and HIF-1α transcription factors, which can explain why there was no change in [¹⁸F]FDG uptake. In drug-resistant tumors, phosphorylation of FOXO3A led to its retention in the cytosol and sustained expression of FOXM1 and LDHA, and sustained labeling of lactate from hyperpolarized [1-¹³C]pyruvate. (a,c) fluorescence microscopy images showing migration of FOXO3A into the nucleus in a drug-sensitive tumor. (b) scheme showing control of LDHA expression by FOXO3A and localized ¹³C spectra showing a decrease in lactate labeling in the drug-sensitive tumor posttreatment. The lactate and pyruvate peaks are labeled. Spectra were acquired at 7.0-T using a ¹³C/¹H volume transmit coil with a 20 mm diameter ¹³C receiver coil. (d,g) [¹⁸F]FDG-PET images overlaid on CT images in drug-sensitive and drug-resistant tumors before and after treatment. The tumors are in the bottom left quadrant, and the images show no change in [¹⁸F]FDG uptake post-treatment. (e,f) Hyperpolarized ¹³C-labeled metabolite false-color images overlaid on T₂-weighted ¹H images. The tumor is outlined. ¹³C images were acquired using a 3D single-shot sequence that used spectral-spatial pulses for selective excitation of the pyruvate and lactate resonances. Figure adapted from figures in [36].

Figure 3

Triple-negative breast cancer on HP ¹³C-MRI. (a) Coronal T₁ 3D spoiled gradient echo (SPGR) MR image. (b) Coronally reformatted DCE image at peak enhancement after intravenous injection of a gadolinium-based contrast agent. (c) Summed hyperpolarized ¹³C-pyruvate and (d) summed hyperpolarized ¹³C-lactate images. The area of low ¹³C-pyruvate and ¹³C-lactate signals in the center of the tumor is likely to correspond to an area with low enhancement on DCE MRI. (e) LAC/PYR map showing intratumoral heterogeneity. The dominant intratumoral heterogeneity was concordant between the DCE-MRI and hyperpolarized ¹³C-MR images and represents decreased delivery of both the gadolinium-based contrast agent and ¹³C-pyruvate to the center of the tumor. (f,g) Dynamic hyperpolarized ¹³C-pyruvate and ¹³C-lactate images acquired over 15 time points after intravenous injection of hyperpolarized [1-¹³C]pyruvate (delay = 12 s; temporal resolution = 4 s). (h) Top: ¹³C metabolite spectrum from a coronal dynamic IDEAL spiral CSI slice covering the tumor summed over 15 time points; Bottom: The axial image from the equivalent DCE-MRI data was taken at the time point of maximum tumor enhancement. Abbreviations: ppm parts per million; IC NST invasive cancer of no specific type. Figure reproduced with permission from [23,33].

Figure 4

Changes in the lactate-to-pyruvate ratio between baseline and following treatment in responding and nonresponding breast cancer patients. (a,c,f,h) Coronal T₁-weighted 3D spoiled gradient echo (SPGR) images with the lactate-to-pyruvate ratio map overlaid on the breast tumor. (b,d,g,i) Coronal reformatted DCE images were obtained 150 s after intravenous injection of a gadolinium-based contrast agent. A patient with HER2⁺ breast cancer was imaged at baseline (a,b) and for ultra-early response assessment (c,d) following standard-of-care treatment and showed a decrease in the lactate-to-pyruvate ratio of 41% (e). At surgery, non-pCR with residual invasive cancer was identified. Another patient with TNBC was imaged at baseline (f,g) and for ultra-early response assessment (h,i) following treatment with chemotherapy and a PARP inhibitor and showed an increase in the lactate-to-pyruvate ratio of 157% (j). At surgery, pCR without residual invasive breast cancer was found. Reprinted after modifications with permission from [54].