Abstract
Tumor hypoxia has been shown to promote tumor progression, cause resistance to both radiotherapy and chemotherapy and is a marker for poor prognosis. Due to its ability to modulate the clinical outcome, hypoxia has been recognized as a tumor-specific target for development of novel hypoxia-activated/targeted bioreductive pro-drugs. In the present study, a (1)H MRI based pO(2) measuring technique, PISTOL (Proton Imaging of Siloxanes to map Tissue Oxygenation Levels), was used to determine the baseline oxygenation of pre-clinical models of non-small cell lung cancer (NCI-H1975) and epidermoid carcinoma (A431) and their response to hypoxia-activated pro-drug (HAP) Tirapazamine (TPZ). A431 tumors had lower baseline oxygenation than the NCI-H1975 tumors (p < 0.05). TPZ was effective in slowing the tumor growth only in NCI-H1975 tumors, even though A431 tumors were significantly less oxygenated (50 ± 14 torr for NCI-H1975 vs. 30 ± 11 torr for A431, p < 0.0001) and had similar hypoxic fractions (0.08 ± 0.09 for NCI-H1975 vs. 0.18 ± 0.19 for A431, p > 0.05) pre-treatment.TPZ resulted in a significant decrease (p < 0.05) in the oxygenation of central regions of NCI-H1975 tumor, and an increase in the edema fraction which can be attributed to its reported anti-vascular effect. In conclusion, PISTOL was able to report the changes in tumor oxygenation in both the tumor xenografts over time and in response to therapy. The baseline mean pO(2) and hypoxic fraction alone were not a reliable indicator of therapeutic efficacy and the results indicate that other factors (e.g. tumor perfusion and relative expression of nitroreductases) may play an important role in determining efficacy of hypoxia-targeted treatments in conjunction with tumor oxygenation status. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1038/s41598-025-32618-2.