Recombinant epoetins (epoetin alfa, darbepoetin alfa, epoetin beta) increase hemoglobin (Hb), reduce transfusions, and improve quality of life in patients with chemotherapy-related anemia. However, results from 2 phase III trials reporting lower survival rates relative to placebo in cancer patients treated with recombinant epoetins beyond anemia correction (ie, to Hb >12 g/dL) (Leyland-Jones. Lancet Oncol 2003; Henke et al. Lancet 2003) have prompted further investigation into the potential proliferative action of these agents on non-erythroid cells, including tumor cells, which express EPOR. The current analyses evaluated the clinical significance of EPOR expression after in vitro administration of recombinant human erythropoietin (rHuEPO, epoetin alfa), as well as the in vivo effect of recombinant epoetins on tumor growth, in 2 well-established preclinical models of breast carcinoma (MDA-MB-231 and MCF-7 cell lines). The in vitro analysis evaluated EPOR expression under hypoxic and normoxic conditions by immunoblotting, flow cytometry, and immunohistochemistry. EPOR binding was assessed with radioactive iodine-labeled rHuEPO (125I-rHuEPO). The in vivo analysis evaluated the effect of recombinant epoetin therapy on tumor growth in orthotopically implanted MDA-MB-231 and MCF-7 breast carcinoma xenograft models in athymic mice. Mice received 1 of 4 treatments:

  1. saline (control) every other day (QOD),

  2. 0.0025 mg/kg epoetin alfa subcutaneously (SC) QOD,

  3. 0.0075 mg/kg darbepoetin alfa SC once weekly, and

  4. 0.0025 mg/kg epoetin beta SC QOD.

Effect on tumor growth was measured by calculating the difference in the final (Day 23) mean tumor volume between the treated group and the control group. Both cell lines demonstrated EPOR staining almost exclusively in the cytosol, with minimal cell surface expression. Intracellular EPOR was comparable under normoxic and hypoxic conditions, and hypoxia did not affect the expression or localization of EPOR. Epoetin alfa did not stimulate the migration, proliferation, or activation of signal transduction cascades in the 2 breast cancer models, although these pathways are normally activated in hematopoietic cells. There was no significant effect on tumor volume after 23 days of recombinant epoetin therapy compared with control. Mean tumor volumes ± standard error (SE) in the MDA-MB-231 cells on Day 23 were as follows: control, 847.6 ± 91.9 mm3; epoetin alfa, 560.6 ± 57.4 mm3; darbepoetin alfa, 809.8 ± 129.9 mm3; epoetin beta, 730.7 ± 66.4 mm3. In the MCF-7 cells, mean tumor volumes ± SE, respectively, were 1004.3 ± 72.9 mm3, 914.5 ± 245 mm3, 884.5 ± 97.1 mm3, and 809.4 ± 103.3 mm3. Recombinant epoetin therapy did not affect tumor inhibition or survival when coadministered with paclitaxel. In both cell lines, recombinant epoetin therapy resulted in mean final Hb values that were significantly (P<.01) higher than those observed with control (all agents produced Hb increases >1.0 g/dL/week), validating that pharmacologic doses were administered. Our findings suggest that although EPOR was expressed, it was nonfunctional and not involved in tumor growth promotion in these 2 models of breast carcinoma.

Author notes

Corresponding author