The expression of erythropoietin (EPO) and the EPO receptor (EPOR) have been recognized in a variety of human cancers, including breast, prostate, colon, ovary, uterine, cervical, glioblastoma, and head and neck squamous cell carcinoma (HNSCC). A growing number of studies demonstrate functional EPO/EPOR signaling events in cancer cells that contribute to disease progression. These findings have been difficult to reconcile with the current clinical practice of treating or preventing anemia in cancer patients with recombinant EPO (rEPO). Anemia is a wellrecognized negative prognostic factor for cancer patients treated with radiation therapy and chemotherapy. EPO has been an appealing alternative to blood transfusion for these patients to increase hemoglobin levels and to improve quality of life by decreasing fatigue. In addition to the identification of EPO and EPOR on cancer cells, a number of recent clinical trials have heightened concerns regarding the potential effects of EPO on cancer progression. The Breast Cancer Erythropoietin Trial (BEST) examined rEPO treatment in nonanemic, metastatic breast cancer patients. This study was terminated early when a higher mortality rate was observed among the rEPO treatment group. In a previous report, Henke and colleagues performed a randomized, double-blind, multi-institutional trial comparing anemic head and neck cancer patients receiving radiation therapy with or without rEPO treatment. These authors reported that rEPO treatment corrected hemoglobin levels, but was associated with decreased locoregional cancer control and patient survival. Although there was some concern regarding study design and methodology, that study and the BEST trial raised legitimate concerns regarding current recommendations for the routine use of rEPO in cancer patients. In the study reported in this issue of the Journal of Clinical Oncology, Henke and colleagues examined the expression of EPOR in a subset of 154 patients from a single clinical center derived from their original multicenter population (351 patients). Expression of EPOR was assessed by immunohistochemistry and reviewed by two independent, blinded pathologists. Within their patient population, 104 patients (67.5%) demonstrated EPOR expression. Their findings demonstrated that decreased locoregional progression-free survival due to treatment with rEPO was significant only in patients with HNSCC that expressed EPOR. rEPO treatment did not impact outcome in patients who had tumors that did not express EPOR. Although originally identified as a cytokine that promoted RBC progenitor survival and differentiation, the role of EPO in normal cells has been dramatically redefined within the past decade. EPO/EPOR signaling regulates physiologic cyclic uterine angiogenesis. In the CNS, EPO is produced by astrocytes and EPOR is expressed by neurons. EPO appears to have a neuroprotective effect against ischemic injury. Similarly, rEPO protects cardiac myocytes against ischemic injury in preclinical studies. The ability of cancer cells to subvert the EPO/EPOR system should not be surprising. A significant number of oncogenes represent signaling molecules that have been appropriated by cancer cells to facilitate proliferation and survival. The overexpression of the epidermal growth factor receptor in HNSCC is crucial for growth, invasion, metastasis, and angiogenesis. A number of preclinical studies have demonstrated EPO-mediated activation of the mitogenactivated protein kinase (MAPK), phosphatidylinositol 3-kinase (PI3K)–Akt, JAK-STAT (Janus kinase-Signal Transducer and Activator of Transcription), and nuclear factor-kappa B (NF B) signaling pathways in a variety of human cancers. Activation of these signaling cascades has been associated with cellular functions that promote tumor progression. rEPO appears to induce cancer cell proliferation and inhibit apoptosis. In addition, EPO stimulates not only chemotaxis of endothelial cells, but also migration and invasion of breast cancer and HNSCC cells. Pretreatment with rEPO protects some cancer cell lines from the cytotoxic effects of the chemotherapeutic agent, cisplatin. Thus, EPO/EPOR signaling appears to contribute to a wide variety of tumor-promoting functions in different cancer types. The contribution of the EPO/EPOR signaling axis to cancer progression is not completely straightforward. The influence of EPO/EPOR on different cancer types appears to be quite variable and remains incompletely understood. A number of different tumor cell lines do not demonstrate any proliferation response to rEPO. Similarly, certain cancer types undergo increased apoptosis or are more sensitive to fluorouracil treatment. Clearly, the complex biology of EPO/EPOR signaling in normal and cancer cells requires continued research. JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L VOLUME 24 NUMBER 29 OCTOBER 1