PurposeTo estimate the rate constant for pyruvate to lactate conversion in tumours in response to a hypoxic challenge, using hyperpolarised 13C1-pyruvate and magnetic resonance spectroscopy.Methods and materialsHypoxic inspired gas was used to manipulate rat P22 fibrosarcoma oxygen tension (pO2), confirmed by luminescence decay of oxygen-sensitive probes. Hyperpolarised 13C1-pyruvate was injected into the femoral vein of anaesthetised rats and slice-localised 13C magnetic resonance (MR) spectra acquired. Spectral integral versus time curves for pyruvate and lactate were fitted to a precursor-product model to estimate the rate constant for tumour conversion of pyruvate to lactate (kpl). Mean arterial blood pressure (MABP) and oxygen tension (ArtpO2) were monitored. Pyruvate and lactate concentrations were measured in freeze-clamped tumours.ResultsMABP, ArtpO2 and tumour pO2 decreased significantly during hypoxia. kpl increased significantly (p < 0.01) from 0.029 ± 0.002 s−1 to 0.049 ± 0.006 s−1 (mean ± SEM) when animals breathing air were switched to hypoxic conditions, whereas pyruvate and lactate concentrations were minimally affected by hypoxia. Both ArtpO2 and MABP influenced the estimate of kpl, with a strong negative correlation between kpl and the product of ArtpO2 and MABP under hypoxia.ConclusionThe rate constant for pyruvate to lactate conversion, kpl, responds significantly to a rapid reduction in tumour oxygenation.

Tumour vasculature is notoriously sinusoidal and leaky, and is hence susceptible to vascular disruption. Microtubule destabilising drugs such as the combretastatins form the largest group of tumour vascular disrupting agents and cause selective shutdown of tumour blood flow within minutes to hours, leading to secondary tumour cell death. Targeting the tumour vasculature is a proven anticancer strategy but early treatment response biomarkers are required for personalising treatment planning. Protein induction following treatment with combretastatin A4-phosphate was examined in a mouse fibrosarcoma model (fs188), where tumour cells express only the matrix-bound isoform of vascular endothelial growth factor A (VEGF188). These tumours are relatively resistant to vascular disruption by combretastatin A4-phosphate and hence a study of protein induction following treatment could yield insights into resistance mechanisms. The distribution of a number of proteins induced following treatment were visualised by MALDI-mass spectrometry imaging. Responses identified were validated by LC-ESI-MS/MS and immunohistochemical staining. Significant changes in proteins connected with necrosis, cell structure, cell survival and stress-induced molecular chaperones were identified. Protein-protein interactions were identified using STRING 9.0 proteomic network software. These relationship pathways provided an insight into the activity of the active tumour milieu and a means of linking the identified proteins to their functional partners.

Mathematical models are required to estimate kinetic parameters of [1-(13)C] pyruvate-lactate interconversion from magnetic resonance spectroscopy data. One- or two-way exchange models utilizing a hypothetical approximation to the true arterial input function (AIF), (e.g. an ideal 'box-car' function) have been used previously. We present a method for direct measurement of the AIF in the rat. The hyperpolarized [1-(13)C] pyruvate signal was measured in arterial blood as it was continuously withdrawn through a small chamber. The measured signal was corrected for T1 relaxation of pyruvate, RF pulses and dispersion of blood in the chamber to allow for the estimation of the direct AIF. Using direct AIF, rather than the commonly used box-car AIF, provided realistic estimates of the rate constant of conversion of pyruvate to lactate, kpl, the rate constant of conversion of lactate to pyruvate klp, the clearance rate constant of pyruvate from blood to tissue, Kip, and the relaxation rate of lactate T1la. Since no lactate signal was present in blood, it was possible to use a simple precursor-product relationship, with the tumor tissue pyruvate time-course as the input for the lactate time-course. This provided a robust estimate of kpl, similar to that obtained using a directly measured AIF.

A major application of dynamic nuclear polarization has been in the study of the conversion of hyperpolarized 13C1-pyruvate to lactate in various disease models. In a typical experimental protocol, hyperpolarized pyruvate is converted from solid to liquid state with superheated fluid and collected in a receiving vessel. The hyperpolarized solution is then rapidly transferred by hand from the polarizer to the imaging magnet, where it is manually injected through an intravenous cannula into the test subject by an experienced operator. Such a procedure leads to inconsistencies in timing, injection rate and volume, all of which can influence the time resolved signal as the pyruvate is metabolized. We have developed a fully magnetic-resonance-compatible withdraw/infuse syringe pump made entirely of plastic so that it can be operated within the bore of an unshielded 7 T (310 mm) magnet. The injector can be programed for variable injection volumes and rates to permit the rapid and reproducible injection of hyperpolarized material without human intervention. The injector was designed for use with 1 ml or 3 ml syringes with a maximum delivery volume of 2.4 ml. The standard deviation of delivered volume from the desired volume was found to be 0.7 % across a volume range of 0.6–2.4 ml.