Hyperpolarized dissolution dynamic nuclear polarization can monitor the kinetics of in vivo metabolism by magnetic resonance spectroscopy and imaging. Despite its favorable polarization properties the most common substrate used by this technique, ¹³C₁-pyruvate, is typically injected systemically at a supra-physiological concentration to achieve adequate signal to noise ratio. Hyperpolarizing drug candidates could allow insights into their mode of action and metabolic fate. However, administering drug molecules at high concentrations can lead to adverse effects in animals. The vascular targeting agent combretastatin-A1 (CA1) causes rapid and selective shutdown of blood flow in solid tumors, resulting in extensive oxygen and nutrient deprivation leading to tumour cell death¹. To investigate the in vivo metabolism of drug molecules we have developed a method for directly administering them to tumor tissue by infusion through a single supplying artery, thus maximizing tumor drug delivery and minimizing T1 relaxation and systemic toxicity.

Develop a method for directly injecting ¹³C-labeled compounds into a single tumor-supplying artery.

Obtain proof of principle for MR detection of signal following intra-arterial injection of hyperpolarized ¹³C₁-pyruvate, ¹³C_u-glucose and ¹³C₁-CA1.

BDIX rats (male mean weight 298g, range 233-344g.) were implanted with 10⁶ P22 fibrosarcoma cells into the right inguinal fat pad. Fat surrounding the tumor implantation site was semi-isolated from the main fat pad so that the tumor received its principle blood supply directly from the superior epigastric artery². Tumors were allowed to grow for 7-10 days prior to MR scanning, mean tumor weight after sacrifice = 1.6 ± 0.7 g (mean±S.D.). Rats were anaesthetized with isoflurane for surgery and MR spectroscopy/imaging. The saphenous artery up to the superior epigastric branch supplying the tumor or the contralateral femoral vein was cannulated for hyperpolarized compound delivery. Rats were positioned in a 7T MRI scanner and maintained at 37^oC using a homoeothermic blanket. Hyperpolarized ¹³C₁-pyruvate, ¹³C_u-glucose-d7 or ¹³C-CA1 was administered via the arterial cannulation for direct tumor delivery (0.2ml at 1ml/min) or via the venous cannulation for systemic delivery (5ml/kg over 13s, approximately 7ml/min). ¹³C MR spectra or images from tumor tissue were collected every 1s and processed using Matlab.The arterially injected ¹³C-pyruvate was first observed 7s after dissolution and its area under the time course curve (AUC) was 7 times larger than the systemically administered route; which was first observed at 11s, see figure 1. Accounting for the lower arterial dose compared to the venous injection, the net signal gain was x54. Rapid imaging (¹H FLASH, FOV ~40x40mm, 1x1 mm slice, 64x64, FA 30o, TR/TE 50.05/6 ms, image acquired every 3.2 s) whilst administering gadolinium contrast agent, using the same injection protocol as for hyperpolarized substrates, showed that the tumor tissue was significantly more perfused when compared to the surround normal tissue, which was not the case for venous administered gadolinium. A linear correlation was found between ¹³C-glucose AUC and concentration (10-156mM), see figure 2. A potential glucose metabolite, alanine, was tentatively assigned. Hyperpolarized custom ¹³C-labelled CA1 was arterially administered and its parent peak observed, , at its expected chemical shift (58ppm), see figure 3.We describe a new method for the efficient delivery of hyperpolarized substrates to a site of interest, i.e. tumor, which minimizes the systemic dose with increased signal to noise. The method permits in vivo ¹³C detection of molecules with shorter T1 and reduced levels of polarization, compared to ¹³C₁-pyruvate, potentially allowing drug candidate metabolism to be observed.