The incidence of renal tumors has risen dramatically in the last 20 years, largely due to the increased utilization of imaging with incidental discovery of many localized tumors. Management of renal tumors relies heavily on imaging. For these localized tumors, current imaging techniques cannot distinguish approximately 20% of the tumors that are benign from the malignant renal cell carcinomas (RCCs). As such, localized tumors are most commonly treated by surgery, leading to more than 10,000 unnecessary operations for benign tumors every year in the U.S. alone. RCCs exhibit increased glycolysis (aka Warburg effect), resulting in elevated lactate production. This key feature may be used to differentiate RCCs from benign renal tumors using the hyperpolarized carbon-13 (HP ¹³C) magnetic resonance (MR), a powerful molecular imaging technique that can measure real-time dynamic pathway-specific metabolic fluxes. The purpose of this study was to investigate the pyruvate-to-lactate flux in living patient-derived renal tumor tissues using HP [1-¹³C]pyruvate.Renal tissues were obtained from patients undergoing nephrectomy for renal tumors (4 benign tumors, 10 clear cell RCC, ccRCC, and 12 normal renal tissues from the nephrectomy specimen that were not involved by tumors. Cores of 8mm diameter were obtained and further sliced into 300 -350 micron thick slices (optimal thickness to allow maximal oxygen and nutrient diffusion) and cultured as previously described¹. 4-5 tissue slices (~60-80 mg of primary tissue) were perfused in a 5mm MR-compatible bioreactor with reduced sample mass requirements². The bioreactor is a custom-designed 3D perfusion system which provides a physiological setting for the tissues, and which has been shown to facilitate the metabolic characterization of hyperpolarized substrates metabolism³. NMR data were acquired on a narrow-bore 11.7T Varian INOVA (125MHz ¹³C) equipped with a 5mm broadband probe. 31P spectra were obtained to monitor tissue viability (via quantification of βNTP) during the bioreactor experiment. HP ¹³C MR was acquired dynamically (10° pulses, 3 s interval for 300 s) following injection of 0.75 mL of 4mM of [1-13C] pyruvate in the tissue slices. Additionally, some slices were incubated with [3-¹³C] pyruvate on a rotating plate⁴, and [3-¹³C]lactate in the media were quantified using high resolution NMR (Bruker 800MHz) over a period of 6-8 hours. Tissue staining, including H&E, and monocarboxylate transporters (MCT), was performed on all the slices and a clinical pathologist classified and graded the tissue slices. mRNA expression levels of MCT4 (MCT isoform responsible for lactate transport) and LDH-A (lactate dehydrogenase , isoform A) and the enzyme activity of LDH (lactate dehydrogenase, enzyme that interconverts pyruvate and lactate) were quantified. All data are expressed as mean ± std.error. Two tailed student’s t-test was used with a significance level of p<0.05. Fig.1 shows a representative ³¹P spectra of a ccRCC with elevated PMEs (phosphomonoesters, which comprises of phophosholine,PC, and phosphoethanolamine, PE) similar to prior work⁵. The inset shows stable bNTP concentrations of the tissue slices maintained in the bioreactor for over 24 hours, which provides evidence of metabolic viability during the course of HP MRS bioreactor study. A representative HP ¹³C MR spectra acquired at 50 sec after infusion of HP pyruvate in a grade 2 ccRCC, and the kinetics of lactate production are shown in fig.2A. Benign tumor tissue produced 2.5 fold higher HP lactate (p<0.005) than normal renal tissue (fig.2B). The observed HP lactate of ccRCC is, however, lower than that of benign renal tumors despite significantly higher LDH expression and activity in ccRCCs (fig.3). This likely reflects rapid lactate efflux in ccRCC. In the bioreactor with a continuous flow of medium, rapidly exported lactate is expected to flow away from the active region of the MR coil, and would not contribute to the observed hyperpolarized lactate signal. The rapid lactate efflux in ccRCC is supported by significantly higher mRNA expression and immunohistochemical stain of MCT4, which exports lactate out of the cell (fig.4). We further confirmed the increased lactate efflux in ccRCC via thermal labeling of renal tissues with [3-¹³C] pyruvate and measuring [3-¹³C]lactate in the media. We observed three fold higher rate of lactate efflux in ccRCC compared to the normal kidney and benign tumor tissue (fig.4C). We have shown that rapid lactate efflux is a characteristic feature of ccRCCs. This feature can be used to differentiate ccRCCs (which comprise the majority of all RCCs) from benign tumors using HP ¹³C pyruvate MR. Because of the implication of lactate efflux in tumor invasiveness and the prognostic value of MCT4 expression in ccRCC⁶, HP pyruvate MR has the potential to provide noninvasive metabolic biomarkers of renal tumor aggressiveness.