Epor Is Not Expressed or Functional in Primary Tumor Cells Isolated Directly From Human Primary and Metastatic Epithelial Tumors Including Breast, Non-Small Cell Lung, Colorectal, and Ovarian Tissues

EpoR expression and function was analysed in disaggregated primary tumor cells obtained from surgically resected human tumor tissues from 186 patients: colorectal (n=46), breast (n=34, including 8 Her2+ tumors defined by HercepTest™), non small cell lung (n=41), ovarian (n=35), Head and Neck (n=5) and other tumor types (n=8). Metastatic tissues were also analysed: colon (n=2), ovary (n=3), lung (n=2), pancreatic (n=1) as well as tumors from pre-treated patients: colorectal (oxaliplatin, bevacizumab; n=1), breast (cyclophosphamide, methotrexate, fluorouracil; n=2), breast (radiation, docetaxel cyclophosphamide, Paclitaxel; n=1), non small cell lung (radiation; n=2), ovarian (carboplatin and Paclitaxel; n=2).

To examine if cell surface EpoR is expressed in the tumor cell compartment, flow cytometry analysis was carried out using a validated, EpoR-specific antibody (mAb 307). To evaluate if tumor cells respond functionally to recombinant human Epo (rHuEpo), levels of phosphorylated downstream signaling proteins (Akt, Erk1/2 and STAT5) were analysed by intracellular flow cytometry using phospho-specific antibodies following ex vivo stimulation with a titration of rHuEpo (0U [vehicle] to 300U/mL for 5 and 30 min). Expression of known tumor growth factor receptors (IGF-1R, c-Met and EGFR) was analyzed as positive controls. Additionally, the response of Akt, Erk and STAT pathways to stimulation with a tumor growth factor cocktail (“GF”: EGF, HGF and IGF-1), was used as a control for pathway integrity. The sensitivity of all flow based assays was validated through the analysis of EpoR expression and function in erythroid precursor cells (EPCs) across a time course of differentiation from primary human bone marrow samples.

Though tumor samples exhibited cell-surface expression of EGFR, c-Met, and IGF-1R, no cell-surface EpoR was detected in tumor cells from any of the 186 tumors examined by flow cytometry. Mean expression relative to negative control: EGFR = 2.80 (95% CI, 1.63–3.97), c-Met = 5.28 (95% CI, 4.39–6.17), IGF-1R = 8.03 (95% CI, 6.81–9.24) and EpoR (0.69 (95% CI, 0.64–0.73). No EpoR was detected by Western blot analysis of tumor lysates. As shown in Table 1, treatment with the GF cocktail stimulated phosphorylation of Akt and Erk in all cohorts. However, no activation of these pathways (or STAT5) by rHuEpo was observed in any of tumors even at high supra-pharmacological concentrations of rHuEpo (up to 300U/mL) and extended stimulation times (up to 30 mins).

Analysis of EpoR expression and function during differentiation in-vitro of EPCs revealed detectable EpoR cell surface expression (2.13-fold above baseline) and function (e.g. pSTAT5: 1.39-fold relative to vehicle at 300 U/mL rHuEpo) as early as day 1 when EPCs (CD36+/CD34-) comprised only 6.9% of the population. By day 8, (>90% EPCs) EpoR expression was detectable at 6.36-fold relative to baseline and pSTAT5 activation was observed 5.6-fold relative to vehicle at 300U/mL rHuEpo.

Using sensitive methods no evidence of EpoR expression or function in primary human tumor cells isolated directly from primary and metastatic tumors was observed in multiple epithelial tumor types. Expression and functional activation of EGFR, c-Met and IGF1R was readily detectable. Therefore, this study suggests that it is unlikely that rHuEpo acts as a growth factor for primary human tumor cells.

McCaffery:Amgen: Employment, Equity Ownership. Rossi:Amgen: Employment, Equity Ownership. Paweletz:Amgen: Employment, Equity Ownership. Tudor:Amgen: Employment, Equity Ownership. Elliott:Amgen Inc: Employment. Fitzpatrick:Amgen Inc.: Employment, Equity Ownership. Patterson:Amgen: Employment, Equity Ownership.