Targeting hypoxic microenvironment of pancreatic xenografts with the hypoxia-activated prodrug TH-302

Abstract

Previous reports have suggested that the hypoxic microenvironment provides a niche that supports tumor stem cells, and that this might explain clinical observations linking hypoxia to metastasis. To test this, we examined the effects of a hypoxia-activated prodrug, TH-302, on the tumor-initiating cell (TIC) frequency of patient-derived pancreatic xenografts (PDX).The frequencies of TIC, measured by limiting dilution assay, varied widely in 11 PDX models, and were correlated with rapid growth but not with the levels of hypoxia. Treatment with either TH-302 or ionizing radiation (IR), to target hypoxic and well-oxygenated regions, respectively, reduced TIC frequency, and the combination of TH-302 and IR was much more effective in all models tested. The combination was also more effective than TH-302 or IR alone controlling tumor growth, particularly treating the more rapidly-growing/hypoxic models. These findings support the clinical utility of hypoxia targeting in combination with radiotherapy to treat pancreatic cancers, but do not provide strong evidence for a hypoxic stem cell niche.

Keywords: TH-302; hypoxia; pancreatic cancer; patient-derived xenograft; tumor-initiating cells.

Figure 1. Patient-derived pancreatic xenograft models display different growth rates and hypoxia level and closely resemble the patient tumor

A. Representative H&E staining of the patient surgical specimen and tumors derived from the orthotopic and subcutaneous site of OCIP51 and OCIP110 show stable tumor morphology after implantation at either site if compared to the patient tumor (scale bars = 100micron). B. Percentage of tumor hypoxia as indicated by EF5 staining in the patient-derived xenograft models. C. EF5 staining vs. tumor growth shows that hypoxia mostly correlates with rapid growth pattern. Tumor growth is defined by the time elapsed between two passages. D. Correlation between Ki67 expression and growth rate in xenograft tumors. E. Representative sections of double immunofluorescent staining for Ki67 (green) and EF5 (red) of OCIP19, OCIP23 and OCIP51 tumors (DAPI: blue).

Figure 2. TH-302 treatment reduces tumor weight and growth rate in fast-growing hypoxic xenografts

A. Treatment schedule for chronic TH-302 treatment of subcutaneous tumors. Animals were treated with 50mg/kg TH-302 on days 1, 5 and 9 and 2Gy of irradiation on days 2, 3 and 4. B. Tumor growth of OCIP19 (n = 6 per group), OCIP23 (n = 6 per group), OCIP51 (n = 6 per group), OCIP83 (n = 4 per group) and OCIP110 (n = 4 per group) in response to treatment with TH-302 or IR alone or with the combination of both according to the treatment schedule. Tumor volume was measured using calipers and referenced back to the starting volume to calculate the change in tumor volume over the course of the treatment. Error bars represent SEM. * p≤0.05; ** p≤ 0.01

Figure 3. TIC frequency strongly correlates with tumor growth rate but not hypoxia

A. Statistical analysis shows that TIC frequency strongly correlates with tumor growth (R = 0.88), B. but not with the magnitude of tumor hypoxia (R = 0.4). The apparent rapid growth seen in OCIP130, which had the lowest TIC frequency, is explained by the accumulation of mucin-filled ducts resulting in a tumor mass disproportionate to the cellular content (Supplementary Figure 2E). C. Treatment with 10Gy of IR reduced the TIC frequency in most of the tested models. TH-302 as a single agent reduced TIC in a dose dependent manner in most of the tested models with the exception of OCIP23. A similar dose-dependency is observed in the TH-302 + IR group. With the exception of OCIP19, the combination of TH-302 and IR was beneficial compared to IR or TH-302 alone in all tested models.