A Two-Component Assay for Hypoxia Incorporating Long-Term Nitroreduction and Short-Term DNA-Damage Allows Differentiation of the Three Hypoxia Sub-types.

Hypoxia in tumors has many well-characterized effects that are known to prevent optimal cancer treatment. Despite the existence of a large number of assays that have supported hypoxia as an important diagnostic, there is no routine clinical assay in use, and anti-hypoxia therapies have often not included parallel hypoxia measurements. Even with a functioning hypoxia assay, it is difficult to match the oxygen dependence of treatment resistance to that of the assay, and this mismatch can vary substantially from assay to assay and even from tumor to tumor [e.g., caused by endogenous variations in non-protein sulfhydryls (NPSH)]. An underlying concern is the current inability to measure the three types of hypoxia; in particular, cycling hypoxia can affect all aspects of detection and treatment strategy. Here we present data that help validate a new two-component hypoxia assay recently suggested by our laboratory. This assay incorporates the long-term bioreduction of the 2-nitroimidazole, EF5, and the short-term production of γ-H2AX (e.g., time of ionizing radiation exposure). The former can be calibrated to provide the average tissue pO2 over the EF5 exposure time while the latter provides the combined sum of microenvironmental radiation response modifiers (e.g., oxygen and NPSH) at the time of irradiation. Importantly, formation of γ-H2AX is not dependent on blood flow, while EF5 binding is only minimally so, due to the rapid and extensive diffusion characteristics of lipophilic compounds. While both individual assays have their limitations, which are addressed in this article, their combination can dissect the type of hypoxia present. In particular, a mismatch between the two assays can directly detect cycling hypoxia in a therapeutically relevant manner. Preliminary use of this two-component assay in small PC3 tumors showed essentially no binding of EF5. Similarly, there were no tumor regions (for uniform irradiation with 12 Gy) with the low levels of γ-H2AX expected for a condition of cycling hypoxia. Thus, both assays were consistent with an essentially aerobic, radiation-responsive tumor. In a larger PC3 tumor, all regions of high EF5 binding had low levels of γ-H2AX.