To the Editor:

Henke and colleagues hypothesize, in the October 10, 2006, issue of the Journal of Clinical Oncology, that epoetin might adversely affect the prognosis of patients with head and neck cancer if malignant cells express erythropoietin receptors (EpoR).¹ However, the data supporting this hypothesis are not convincing.

While there are other publications suggesting that EpoR are expressed in tumors and cancer cell lines,^2,3 there are indisputable problems with the techniques used for identifying EpoR location (either on the cell surface or intracellular) and its functionality. Henke et al¹ attempted to demonstrate retrospectively the presence of EpoR in tumor biopsies using the C20 polyclonal rabbit antihuman EpoR antibody. However, studies using Western blot analysis show convincingly that the C20 antibody, and many other commercially available anti-EpoR antibodies, are nonspecific and bind to multiple targets.⁴

We have recently investigated whether EpoR are expressed and transmit a proliferative signal in a panel of human cancer cell lines by using reverse-transcriptase polymerase chain reaction, immunofluorescence, and Western blot analysis.⁵ We detected EpoR mRNA in all cell lines. The available commercial antibodies tested (including the C20 antibody) detected multiple proteins. However, experiments with small interfering RNA showed that this staining was not specific and that no EpoR protein was expressed by the cancer cells. This brings into question the significance of immunohistochemical analysis using these antibodies. For example, the putative EpoR identified by Henke et al¹ could include other proteins, such as heat shock protein 70, which is known to be elevated in human cancer and to cross react with the C20 antibody.⁴ Henke et al¹ did not attempt to colocalize receptor mRNA within the biopsies. Indeed, such an approach would have had its own limitation in only detecting EpoR transcripts and not functional receptor protein.

Moreover, studies suggest that EpoR identified in tumor cells are often extensively located in the cell cytoplasm, making the receptors inaccessible to epoetin. Sinclair et al⁶ demonstrated that, despite high levels of EpoR protein identified by Western blot analysis in several cancer cell lines, little or none of this receptor was located on the cell surface. Similarly, Abdalla et al⁷ could not demonstrate EpoR expression on the cell surface of tumor cells from patients with lymphoproliferative malignancies, although EpoR mRNA was identified by reverse-transcriptase polymerase chain reaction in a significant number of tumor samples.

In support of these findings, various in vitro studies have failed to show any growth promoting effect of epoetin over a wide dose range in many tumor cell lines, including those with evidence suggestive of EpoR expression.^3,7,8 Other studies suggesting that epoetin increases the proliferation of EpoR expressing cell lines generally suffer from methodologic issues.^2,9 For example, epoetin levels 10- to 1000-fold greater than the maximum plasma level observed in patients receiving approved doses of epoetin¹⁰ were required to demonstrate EpoR stimulation in vitro. Indeed, Acs et al² had to culture cells in a medium containing epoetin 250,000 U/L to observe EpoR-mediated tyrosine phosphorylation.

While Henke et al¹ suggest that administration of epoetin may stimulate cell growth in patients with EpoR-expressing tumors, no attempt is made to explain how stimulation of EpoR might reduce the efficacy of radiotherapy in their patients. It would be expected that cells proliferating in response to epoetin would remain susceptible to the cytotoxic actions of radiotherapy.

It is also possible that methodologic issues in the original study by Henke et al,¹¹ from which the biopsy samples were obtained, may partially explain the reduced survival observed in patients treated with epoetin; these have not been adequately addressed by Henke et al in their publications.^1,11 Importantly, baseline imbalances in key prognostic factors in the original study favored the placebo group.¹¹ These imbalances were evident mainly in patients with tumors of the hypopharynx and, notably, epoetin was associated with a significantly poorer outcome only in these patients and not in patients with other tumor types.^11,12 The significance of these imbalances is potentially compounded in the secondary analysis¹ because it is based on a subset of only 44% of patients from the original study. Importantly, within the subset of patients claimed to be EpoR positive (n = 104), substantially more patients treated with epoetin (27%) than patients treated with placebo (19%) had tumors of the hypopharynx, suggesting a potential for bias towards poorer survival in the epoetin group. Conversely, in patients classified as EpoR negative (n = 50), there were more placebo patients (31%) with hypopharyngeal cancer than epoetin patients (21%). These differences may at least partially explain any survival differences observed between patients classified as EpoR positive compared with those classified as EpoR negative. Moreover, the percentage of patients on epoetin therapy who also received radiation therapy according to protocol was higher in EpoR negatively classified patients (96%) than in EpoR positively classified patients (87%).

The question of whether epoetin promotes tumor progression via stimulation of EpoR is important but the data presented by Henke et al¹ in patients with head and neck cancer are not convincing. Patients with cancer and anemia derive great benefits from epoetin therapy, including reduced transfusion need and improved quality of life. Recently updated evidence-based guidelines on the use of epoetin in patients with cancer suggest initiating epoetin at hemoglobin levels of 9 to 11 g/dL and maintaining hemoglobin levels of 12 to 13 g/dL during treatment.¹³ The available data suggest that, when used according to this schedule, epoetins are a safe and well-tolerated treatment of anemia.

AUTHORS’ DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Although all authors completed the disclosure declaration, the following authors or their immediate family members indicated a financial interest. No conflict exists for drugs or devices used in a study if they are not being evaluated as part of the investigation. For a detailed description of the disclosure categories, or for more information about ASCO's conflict of interest policy, please refer to the Author Disclosure Declaration and the Disclosures of Potential Conflicts of Interest section in Information for Contributors.

Employment: N/A Leadership: N/A Consultant: Wolfgang Jelkmann, Roche, Amgen, Ortho Biotech Stock: Wolfgang Jelkmann, Roche, Amgen Honoraria: Wolfgang Jelkmann, Roche, Amgen, Ortho Research Funds: N/A Testimony: N/A Other: N/A