The use of hyperpolarized carbon-13 magnetic resonance for molecular imaging

Abstract

Until recently, molecular imaging using magnetic resonance (MR) has been limited by the modality's low sensitivity, especially with non-proton nuclei. The advent of hyperpolarized (HP) MR overcomes this limitation by substantially enhancing the signal of certain biologically important probes through a process known as external nuclear polarization, enabling real-time assessment of tissue function and metabolism. The metabolic information obtained by HP MR imaging holds significant promise in the clinic, where it could play a critical role in disease diagnosis and therapeutic monitoring. This review will provide a comprehensive overview of the developments made in the field of hyperpolarized MR, including advancements in polarization techniques and delivery, probe development, pulse sequence optimization, characterization of healthy and diseased tissues, and the steps made towards clinical translation.

Keywords: Cellular imaging; DNP; Hyperpolarization; MRSI; Metabolic imaging; Molecular imaging; PHIP; Pyruvate.

Figure A

Signal-time courses of multiple metabolites derived from dynamic HP [1-¹³C]pyruvate spectroscopy in a rat. The boxed regions specify the time window for a 17-s CSI acquisition at the following time delays: 15 s, 25 s, and 35 s. The choice of delay determines the best SNR for the metabolites of interest. The insert shows a waterfall plot of the different metabolite peaks: bicarbonate, pyruvate, alanine, pyruvate-hydrate, and lactate (from left to right). Each spectrum was acquired every 3 s. Reproduced from [113].

Figure B

Metabolite maps generated from a single-point 2D CSI pulse sequence, overlaid over a ¹H T₂-weighted FSE anatomical image of the kidneys. The spectrum (B) is from the 1.25×1.25×1cm³ voxel in (A). Reproduced from [104].

Figure C

A (A) single shot three-dimensional double spin echo pulse sequence and its (B) 3D k-space trajectory. The k-space was acquired as a stack of interleaved spirals. Specific metabolites were excited via a spectral-spatial pulse. Reproduced from [152].

Figure D

(A) The fate of the ¹³C labeled carbon-1 (red). In most organs, the label will be detected on the lactate, alanine, bicarbonate (via PDH). In the liver, the label can also appear on other TCA cycle intermediates via the enzyme pyruvate carboxylase (PC). Spectra from (B) fasted and (C) fed states of a perfused, isolated rat liver after injection of HP [1-¹³C]pyruvate. Figures adapted from [178].

Figure E

The three images are HP [1-¹³C]lactate/[1-¹³C]pyruvate maps overlaid on axial T₂-weighted ¹H images of the prostate from three patients in the first ever hyperpolarized MR clinical study. All three patients had a biopsy-proven Gleason grade 3 + 3 prostate cancer. Adapted from [103].

Figure F

(A) The signal-time course of HP lactate after the injection of HP [1-¹³C]pyruvate in isolated, perfused rat lungs of a bleomycin-induced lung injury model. The maximum summed lactate signal is seen on day 7, during the inflammatory phase of this model. The HP lactate signal correlates highly with the semi-quantitative neutrophil score (B), and moderately with the macrophage score (C), based on H&E histology. Adopted from [124].

Figure G

A multi-bolus functional HP MRS experiment based on an electrical stimulation of hind-limb skeletal muscles in mice. (A) Four boluses of HP [1-¹³C]pyruvate were delivered (blue bands), each followed by electrical stimulation (red bands). Each bolus was delivered about 30 s apart; each stimulation was 10 s long and involved 10 ms trains of 200 μs/10 V pulses repeated at 10 Hz. (B) The metabolite-to-total-carbon-ratio (MtoC) of all four boluses with best-fit quantification of each bolus’s kinetic data (solid lines). (C) The metabolic rate constants, as determined for each bolus (p <0.05). Reproduced from [270].