Distinguishing metabolic signals of liver tumors from surrounding liver cells using hyperpolarized ¹³C MRI and gadoxetate

Purpose: To use the hepatocyte-specific gadolinium-based contrast agent gadoxetate combined with hyperpolarized (HP) [1-C-13]pyruvate MRI to selectively suppress metabolic signals from normal hepatocytes while preserving the signals arising from tumors. Methods: Simulations were performed to determine the expected changes in HP C-13 MR signal in liver and tumor under the influence of gadoxetate. CC531 colon cancer cells were implanted into the livers of five Wag/Rij rats. Liver and tumor metabolism were imaged at 3 T using HP [1-C-13] pyruvate chemical shift imaging before and 15 min after injection of gadoxetate. Area under the curve for pyruvate and lactate were measured from voxels containing at least 75% of normal-appearing liver or tumor. Results: Numerical simulations predicted a 36% decrease in lactate-to-pyruvate (L/P) ratio in liver and 16% decrease in tumor. In vivo, baseline L/P ratio was 0.44 +/- 0.25 in tumors versus 0.21 +/- 0.08 in liver (p = 0.09). Following administration of gadoxetate, mean L/P ratio decreased by an average of 0.11 +/- 0.06 (p < 0.01) in normal-appearing liver. In tumors, mean L/P ratio post-gadoxetate did not show a statistically significant change from baseline. Compared to baseline levels, the relative decrease in L/P ratio was significantly greater in liver than in tumors (-0.52 +/- 0.16 vs. -0.19 +/- 0.25, p < 0.05). Conclusions: The intracellular hepatobiliary contrast agent showed a greater effect suppressing HP C-13 MRI metabolic signals (through T-1 shortening) in normal-appearing liver when compared to tumors. The combined use of HP MRI with selective gadolinium contrast agents may allow more selective imaging in HP C-13 MRI.