Over-expression of hypoxia-inducible proteins such as vascular endothelial growth factor (VEGF), metallothionein (MT), glucose transporter-1 (Glut-1), carbonic anhydrase IX (CA IX) and hypoxia inducible factor-1a (HIF-1a) is associated with poor prognosis in human cancers. The association is generally considered to be reflective of underlying tumor hypoxia but, clinically, little or no correlation is observed between the expression of VEGF, MT, Glut-1, CA IX and HIF-1a and the exogenous hypoxia marker, pimonidazole. VEGF, MT, Glut-1 and carbonic anhydrase are expressed in well-oxygenated basal cells of normal epithelia –not in hypoxic, suprabasal cells - and it seems that these factors might be more reflective of proliferation than hypoxia. In support of this hypothesis, the majority of hypoxic cells in human cancer express a marker for differentiation which might explain why many otherwise, hypoxia inducible factors are not expressed in the majority of hypoxic cells in human cancers. One endogenous hypoxia-inducible protein whose normal function is to sense hypoxia is the growth factor erythropoietin (EPO) and its receptor EPOR. We reported previously that EPO expression is correlated with pimonidazole binding in squamous cell carcinomas and now extend the study of EPO as a potentially useful endogenous hypoxia marker to breast cancer. Thirty-eight patients with confirmed malignancy of the breast who provided informed consent for a study approved by the UNC School of Medicine Institutional Review Board were infused with a dosage of 0.5 g/m2 of pimonidazole HCl dissolved in 0.9% saline. Twenty -four hours later multiple biopsies were taken, fixed in formalin, paraffin embedded, sectioned, immunostained and the area % labeled with pimonidazole adducts, EPO and EPOR measured on contiguous sections. 36 out of 38 tumors expressed EPO and 35 out of 38 expressed EPOR. Patterns of 1) overlap; 2) mixed overlap and no overlap; and, 3) no overlap between pimonidazole binding and EPO and EPOR expression were observed but this heterogeneity did not obviate overall, quantitative correlations between pimonidazole binding and EPO expression [p < 0.001 on a tumor-by- tumor basis (n = 38) and p < 0.0001 on a biopsy-by-biopsy basis (r = 0.60; n = 93)] or between pimonidazole binding and EPOR expression [p < 0.001 on a tumor-by tumor basis (n = 38) and p < 0.0001 on a biopsy-by biopsy basis (r = 0.63; n = 93)]. Supported by NIH CA 100844 (MOA), CA 85361(JAR & MAV), RR00046 (UNC Clinical Cancer Center), DOD BC 962506 (MAV & JAR) and DAMD 17-03-1-0968 (MOA).

[Proc Amer Assoc Cancer Res, Volume 47, 2006]