Synopsis

The purpose of this study was to investigate the feasibility of in vivo ¹³C->¹H hyperpolarization transfer, which has significant potential advantages for detecting the distribution and metabolism of hyperpolarized ¹³C probes, in a clinical MRI scanner. A standalone pulsed ¹³C RF transmit channel was developed for operation in conjunction with the standard ¹H channel of a clinical 3T MRI scanner. Operation of the custom pulsed ¹³C RF channel resulted in effective ¹³C->¹H hyperpolarization transfer, as confirmed by the characteristic anti-phase appearance of ¹H-detected, ¹J_CH-coupled doublets. ¹H detection of HP [2-¹³C]lactate generated in vivo was achieved in a rat liver slice.

Introduction

While the vast majority of prior work on hyperpolarized (HP) ¹³C MRI has relied on direct ¹³C detection, shifting the hyperpolarization to nearby ¹H spins offers significant potential advantages for detecting HP ¹³C magnetization. These include increased sensitivity and reduced imaging gradient requirements, due to the 4x higher gyromagnetic ratio of ¹H. Moreover, specialized ¹³C receiver coils and electronics can be supplanted by standard ¹H equipment, which more readily attains body noise dominance¹ and is already tightly integrated with existing MRI methods.

Initial studies in preclinical systems have reported the feasibility of ¹³C->¹H hyperpolarization transfers in vivo^2,3. However, clinical scanners are not generally configured for heteronuclear transfers. In this work, we have developed an external pulsed ¹³C RF transmit channel for operation in synchronization with water-suppressed ¹H MR spectroscopic acquisition on a clinical 3T scanner. The newly developed framework was applied for ¹³C->¹H hyperpolarization transfer experiments in ¹J_CH-coupled systems: 1) phantom ¹H MR experiments with HP [2-¹³C]glycerol, and 2) in vivo ¹H detection of HP [2-¹³C]lactate generated from injected HP [2-¹³C]pyruvate via LDH in rat liver.

Methods

Development of an external pulsed ¹³C RF transmit channel- The external ¹³C RF architecture is illustrated in Figure 1, based on extensive modification of a commercial standalone ¹H decoupler⁴ (GE). Commands were issued over a PC USB/GPIO connection from a custom software interface (Matlab), enabling execution of arbitrary complex RF waveforms with arbitrary timing delays with respect to a scanner TTL signal. The amplifier was also configured for RF blanking during readout intervals. ¹³C power was delivered into one of the two linear ¹³C modes of a dual-tuned ¹³C/¹H quadrature coil, whose ¹H channel was connected as usual to the 3T clinical MRI scanner. For power calibration, ¹³C signal was detected via the other ¹³C mode.

Pulse sequences- Two ¹³C pulse sequences were programmed-- one served for power calibration, and the other for polarization transfer. The power calibration sequence consisted of a single 500μs hard pulse. The polarization transfer sequence (Figure 2) consisted of two consecutive 500μs hard pulses with 90° RF phase shift. A custom water suppression module was programmed for the ¹H MRS sequence.

MR experiments- We first directly hyperpolarized 30μL [2-¹³C]glycerol in a commercial Hypersense polarizer. Following dissolution in superheated D₂O, ¹³C hyperpolarization was transferred to the proton attached to the labeled carbon. Next, the potential for ¹H detection of [2-¹³C]lactate generated from HP [2-¹³C]pyruvate in vivo was tested in a normal rat. In this case, an 8mm axial ¹H slice through rat liver was excited, 25s after injection of 2.5mL 80mM HP [2-¹³C]pyruvate. An equimolar quantity of sodium lactate (natural abundance) was co-injected with the pyruvate to stimulate HP exchange into the lactate pool.

Results

Effective hyperpolarization transfer in the [2-¹³C]glycerol phantom was evidenced by the characteristic anti-phase appearance of the ¹H doublet on the second readout (Figure 3), following the ¹³C sequence. Initial ¹H hyperpolarization of this doublet (in-phase) was also observed on the first readout, probably due to cross relaxation (23). About 23s was required to transport the HP sample to the MRI scanner. Since the ¹³C T₁ relaxation time of [2-¹³C]glycerol was separately measured to be ~7s in D₂O at 3T, we estimate that ~96% of the initial polarization was lost prior to MR data acquisition in this case. Nevertheless, the result shows clear feasibility for polarization transfer.

Effective transfer of hyperpolarization was also evident in the in vivo rat experiment with HP [2-¹³C]pyruvate. Water suppression was critical in this case, resulting in 190-fold suppression in the 8mm rat liver slice. Similar to the phantom results, a resonance corresponding to transferred HP lactate clearly appeared on the second ¹H MRS readout (Figure 4), following execution of the ¹³C sequence.

Discussion

Our results extend into a clinical MRI scanner environment prior initial reports demonstrating the feasibility of in vivo ¹³C->¹H hyperpolarization transfer^2,3. In contrast to a recent study showing transfer from HP [1-¹³C]lactate to distant methyl protons³ (4.1Hz), much larger couplings (~150Hz) characterize the systems investigated here, which experience less polarization decay associated with required evolution times. Furthermore, refocusing pulses are not required since (1/J_CH)<<T₂*. However, the ¹³C T₁ relaxation times of these species are also much shorter than [1-¹³C]lactate as a result of dipolar coupling, and additional sequence modifications are required to obtain optimized in-phase results.

Future work will focus on these optimizations, in addition to improving the effectiveness of water suppression, likely utilizing a combination of RF and gradient-enhanced approaches. We will also aim to investigate the potential for dynamic ¹H imaging of hyperpolarization transferred from ¹³C.

Acknowledgements

References

1. Edelstein WA, Glover GH, Hardy CJ, Redington RW. The intrinsic signal-to-noise ratio in NMR imaging. Magn Reson Med 1986;3:604–618.

2. Mishkovsky M, Cheng T, Comment A, Gruetter R. Localized in vivo hyperpolarization transfer sequences. Magn Reson Med 2012;68:349–352. doi: 10.1002/mrm.23231.

3. Wang J, Kreis F, Wright AJ, Hesketh RL, Levitt MH, Brindle KM. Dynamic (1) H imaging of hyperpolarized [1-(13) C]lactate in vivo using a reverse INEPT experiment. Magn Reson Med 2017;137:6418. doi: 10.1002/mrm.26725.

4. von Morze C, Tropp J, Chen AP, Marco-Rius I, Van Criekinge M, Skloss TW, Mammoli D, Kurhanewicz J, Vigneron DB, Ohliger MA, Merritt ME. Sensitivity enhancement for detection of hyperpolarized ¹³C MRI probes with ¹H spin coupling introduced by enzymatic transformation in vivo. Magn Reson Med. In press.