In Vivo pH and Metabolite MR Imaging Using Hyperpolarized 13C-Pyruvate
Nicholas Drachman1, Stephen J. Kadlecek1, Mehrdad Pourfathi1,2, Yi Xin1, Harrilla Profka1, and Rahim R. Rizi1

1Radiology, University of Pennsylvania, Philadelphia, PA, United States, 2Electrical and Systems Engineering, University of Pennsylvania, Philadelphia, PA, United States


In this study, we investigate the possibility of simultaneously imaging pH and lactate to pyruvate ratio in vivo in the lungs. We produce hyperpolarized 13C-bicarbonate by rapidly decarboxylating hyperpolarized [1-13C]pyruvate with hydrogen peroxide. By tuning the reaction rate by altering the pH, we produce roughly equal amounts of pyruvate and bicarbonate, which allows us to image both metabolic processes simultaneously.


Many pathological processes are associated with changes in endogenous pH, including cancer, inflammation, and ischemia [1]. In vivo pH imaging is possible using hyperpolarized 13C-bicarbonate, but achieving a high level of polarization in a short time remains a major challenge. This challenge can be overcome by polarizing [1-13C]-pyruvate by DNP and then rapidly converting it to [13C-]bicarbonate by decarboxylation via hydrogen peroxide while minimizing loss of polarization. The rate of decarboxylation can be tuned by altering the pH of the reactants, thereby allowing us to obtain the desired ratio of pyruvate to bicarbonate. By allowing the reaction to proceed to roughly 50% decarboxylation, both pyruvate and bicarbonate are present in high enough concentrations to measure local pH and lactate-to-pyruvate ratio. In this experiment, we demonstrate the effectiveness of this method by locally obtaining these measurements in inflamed rat lungs.


Two Sprague Dawley rats (400g) were ventilated with tidal volume of 12mL/kg and positive end-expiratory pressure (PEEP) of 3 cmH2O for t = 180 minutes. After 70 minutes the lungs were injured via tracheal HCl aspiration (pH 1.25, 2.5mL/kg) to simulate gastric acid aspiration. Rats were imaged in supine position using a 1H/13C Quadrature birdcage coil (m2m) in a 4.7T horizontal magnet (Varian Inc.). The hyperpolarized agent was produced by polarizing 100.1mg of [1-13C]pyruvate (Cambridge Isotopes Inc.) in a Hypersense DNP system. 5mL of 280mM hydrogen peroxide was prepared for the reaction by diluting 1.57mL of unstabilized 3% hydrogen peroxide solution with 3.43mL of DI water. After polarizing for one hour in the DNP system, the hyperpolarized pyruvate was diluted in 4mL of dissolution medium, which produces a neutral, isotonic, 280mM solution. 2.5mL of the 280mM peroxide solution and 2.5mL of the 280mM hyperpolarized pyruvate solution were mixed together with a small amount of NaOH to raise the pH near 12 where the reaction proceeds most rapidly. The solution was given roughly 8 seconds to react before being neutralized and injected into the rat via a tail vein catheter. The lungs were imaged using a slice selective 16x16 single-point CSI pulse sequence with an outward spiral k-space trajectory and an in-plane field of view 50x50 mm2 over a 30 mm axial slice to cover the entire lungs (TR/TE = 70/0.38 ms, SW = 6.0 kHz, α=12°). The data obtained was processed using custom routines in MATLAB 2014b (MathWorks Inc.). The pH in each voxel was calculated using the Henderson-Hasselbach equation, where the relative signal intensities are used in place of the relative concentrations (the validity of this substitution is based on the rapid interconversion of bicarbonate to CO2 due to the endogenous presence of carbonic anhydrase). The lactate-to-pyruvate ratio in each voxel was similarly calculated using the integrals under each metabolite’s respective peak. Both images were thresholded based on total carbon signal to remove noise seen on the perimeter of and outside of the lungs.


Carbonate polarizations of over 13% were achieved using this method. A reduction in pH was observed in nearly all regions of the inflamed lungs as compared to the value of 7.4 seen in healthy tissue. The pH in each voxel ranges from 7.0 to 7.4 as seen in figure 3, it is believed that these differences correspond to local variation in the level of inflammation. The mean pH measured over the entire volume of the lungs was 7.18±0.14, which is well within the expected range for inflamed tissue. At this pH, the CO2 signal is ~10 times smaller than the bicarbonate signal, however the polarization is high enough that the CO2 signal is quantifiable in all voxels in the lung. A sample voxel is shown in figure 1. In these experiments, the decarboxylation reaction was tuned to produce roughly equal amounts of hyperpolarized pyruvate and bicarbonate. With these parameters, there is a high enough concentration of pyruvate to quantify the amount of lactate produced metabolically; the lactate-to-pyruvate map obtained is shown in figure 4. The mean lactate-to-pyruvate ratio over the volume of the lungs was calculated to be ~0.07, which is consistent with other experiments done by our group.


In this study we demonstrated a method to locally measure the pH in inflamed lung tissue in vivo using 13C-bicarbonate produced via the decarboxylation of hyperpolarized [1-13C]-pyruvate. We have also shown the feasibility of simultaneously imaging both pH and lactate-pyruvate ratio in the lungs with a single injection.


No acknowledgement found.


[1] Grinstein, et al. (1991) Regulation of cyto- plasmic pH in phagocytic cell function and dysfunction. Clin. Biochem. 24, 241–247.


Figure 1. 13C spectrum from a sample voxel in the lungs. Pyruvate and bicarbonate are present in roughly equal amounts and both CO2 and lactate are plainly visible and quantifiable.

Figure 2. Spectra overlaid on a proton image of the injured lungs. The majority of the signal is present in the right lung, although the signal is still strong enough in the left lung to be quantifiable.

Figure 3. The calculated pH map of the injured lungs overlaid on the proton image. The calculated pH values range from 7.0 to 7.4.

Figure 4. The lactate-to-pyruvate map of the injured lungs overlaid on the proton image. The mean lactate-to-pyruvate ratio over the lungs is 0.07.

Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)