| HOME | HELP | CONTACT US | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
aInstitut für Physiologie, Universität Duisburg-Essen, Essen, Germany; bHematology-Oncology Service, Luxembourg Medical Centre, Luxembourg
Key Words. Erythropoietin receptor • Erythropoiesis-stimulating agents • Epoetin • Angiogenesis • Hypoxia • Thromboembolism • Venous thromboembolic events
Correspondence: Joachim Fandrey, M.D., Institut für Physiologie, Universität Duisburg-Essen, Hufelandstr 55, D-45147 Essen, Germany. Telephone: 49-201-723-4600; Fax: 49-201-723-4648; e-mail: joachim.fandrey{at}uni-due.de
Received February 25, 2009; accepted for publication June 30, 2009.
Disclosures: Joachim Fandrey: Honoraria: Janssen-Cilag, Ortho Biotech; Mario Dicato: Honoraria: Amgen, Janssen-Cilag.
The content of this article has been reviewed by independent peer reviewers to ensure that it is balanced, objective, and free from commercial bias. No financial relationships relevant to the content of this article have been disclosed by the independent peer reviewers.
 |
ABSTRACT |
|---|
 Top Abstract IntroductionDo ESAs Promote Tumor...Do ESAs Increase the...Author ContributionsReferences |
|---|
|
INTRODUCTION |
|---|
|
TopAbstractIntroductionDo ESAs Promote Tumor...Do ESAs Increase the...Author ContributionsReferences |
|---|
Safety concerns also arose over the potential of ESAs to promote tumor growth through stimulation of EPORs on tumor vessels and/or enhanced tumor oxygenation. Open questions relating to the meaning of EPORs in tumors remain, and these questions include:
Another principal safety concern with ESAs is the associated greater risk for venous thromboembolic events (VTEs). VTEs are frequent in cancer patients as a result of the effects of malignant disease, cancer treatments, and comorbidities. This manuscript examines the involvement of ESAs relating to tumor growth and VTEs.
|
DO ESAS PROMOTE TUMOR GROWTH? |
|---|
|
TopAbstractIntroductionDo ESAs Promote Tumor...Do ESAs Increase the...Author ContributionsReferences |
|---|
Indirect evidence with immunohistochemistry (IHC) staining studies, with an antibody reputed to recognize EPOR, showed that prior absorption of EPOR proteins by antiserum eliminates tumor EPOR immunoreactivity [3, 4]. However, it was subsequently discovered that the antibodies were nonspecific and thus not suitable. Indirect clinical trial evidence demonstrating a shorter locoregional progression-free survival or overall survival time among EPOR+ patients treated with ESAs was also not convincing, because the respective antibodies were used to detect EPOR protein [1, 2, 4].
Direct evidence of EPORs on human tumors includes the demonstration of EPORs on tumor cell lines and tumor samples with polyclonal rabbit antisera [5]. However, using available antibodies, immunoreactivity was not found on the surface of the cells but was localized to the cytoplasm [3]. If those proteins detected within the cells were indeed EPORs, it is unlikely that they would confer signaling upon EPO treatment. In addition, for erythropoietic progenitors, which undoubtedly can express EPORs, it was shown that cytoplasmic EPOR immunoreactivity can be caused by immature forms of the receptor that do not contribute to EPOR signaling [6].
The question of whether or not tumor cells can bind EPO was addressed in two independent studies in which EPO-binding assays were performed using radiolabeled recombinant human EPO (rHuEPO) [7, 8]. A particular focus was laid on mammary tumor cells by including >68 primary breast tumor samples in one study [7] and the breast carcinoma cell lines MCF-7 and MDA-MB-231, which were reported to be EPO responsive in previous studies and to express EPOR mRNA [9–11]. Both studies revealed no significant binding of radiolabeled EPO to tumor cells, which is consistent with a lack of EPOR immunoreactivity at the cell surface [7, 8]. These data indicate that the detection of EPOR mRNA in solid tumor cells does not predict surface expression of the protein nor the potential to bind EPO.
For many studies, commercially available polyclonal anti-EPOR antibodies were used that Elliott et al. [12] showed were not specific and detect multiple proteins, and are therefore not suitable for use in the detection of EPORs on tumor cells. All Western blot analyses in that study included a positive control, the UT-7/Epo dependent cell line. UT-7/Epo cells are hemopoietic in origin and originally derived from a megakaryoblastic leukemia. Their growth depends on EPO and EPOR signaling, which makes them an appropriate control. The anti-EPOR antibodies tested by Elliott et al. [12] included C20 (sc-695; anti-human EPOR; Santa Cruz Biotechnology, Santa Cruz, CA), which has been extensively used to detect EPORs in tumor cells using the IHC methods [12]. In that study, the investigators found that, when using C20, similar intensities of staining were observed in UT-7/Epo, MCF-7, and 769-P cells, which, respectively, contained high, moderate, and undetectable levels of EPOR [12]. Further examinations revealed that the antibody did not detect the EPOR protein but an isoform of heat-shock protein (HSP)70 [12]. C20 possibly recognized HSP70–2 because of regions of similarity between the HSP70 protein and the immunizing peptide that was used to generate the antibodies. Thus, what was reputed to have been EPORs was a crossreaction with the HSP70–2 and HSP70–5 isoforms.
Henke et al. [1] conducted an additional analysis of their patients from the study published in 2003 using this just described C20 antiserum for EPOR staining in head and neck cancer patients receiving radiotherapy with or without EPO [2]. The two EPO-treated groups were classified as EPOR+ or EPOR–. In the EPOR+ group, the locoregional progression-free survival rate was substantially lower in patients receiving epoetin beta than in those receiving placebo [2]. In the EPOR– group, epoetin beta did not impair outcome [2]. However, these results might not be valid because the reagent is nonspecific, and what was probably detected was HSP70. HSP70 is known to be overexpressed in highly malignant tumors, and expression correlates with shorter survival and resistance to apoptosis [13, 14]. HSP70 expression is increased in hypoxic conditions and in fast-growing tumors. Thus, their IHC may have simply detected tumors that were more aggressive because of faster growth and more hypoxic regions.
Can ESAs Stimulate EPORs on Tumor Cells In Vivo?
The majority of in vivo and in vitro studies on the effect of ESAs on human cancers have shown a neutral effect, and none of them have shown a negative effect [15].
A few in vitro studies, using different tumor cell lines, have shown some effect on proliferation and/or apoptosis resistance. However, pharmacologic concentrations (e.g., 200,000 mU/ml) of ESAs that exceed the concentrations measured in untreated healthy humans (
20 mU/ml plasma) or after s.c. administration of ESAs in clinical practice (increased by 150 mU/ml plasma on administration of 150 U rHuEPO per kg body weight) by several orders of magnitude were used in some studies [3]. In some studies, cells were kept under artificial conditions with the use of synchronized cells that were deprived of serum. As described above, well-validated anti-EPOR antibodies were not available and could not identify the presence of EPORs at the protein level. A detailed list of these studies is found in the article by Osterborg et al. [3].
Two additional studies deserve particular mention. Jeong et al. [16] treated human ovarian cancer cell lines (A2780, CaOV, SKOV, and OVCAR-3) with EPO (50,000 mU/ml) and found increased phosphorylation of a key signaling molecule, extracellular signal–related kinase (ERK)-1/2, of the EPOR. However, in none of the four cell lines did EPO directly stimulate the growth or survival of the cells in culture over a concentration range of 0–100,000 mU/ml [16]. In the other study, by Dunlop et al. [17], EPO, at pharmacological concentrations, activated signaling kinases downstream of the EPOR in non-small cell lung carcinoma (NSCLC) cell lines. Increased phosphorylation of the kinases signal transducer and activator of transcription (STAT)5, Akt, and ERK, however, was not associated with a growth advantage for the NSCLC cells over a range of EPO concentrations of 0–1,000 mU/ml [17].
Are EPORs on Tumor Cells Functional?
Preclinical data do not provide compelling support for a role for the EPOR in tumor progression. The EPOR gene itself is not an oncogene, it is not amplified in tumor cells, and there is no selective advantage for tumors to overexpress it [3]. In most studies, only EPOR mRNA levels are detectable, but not elevated, in tumor cell lines, compared with nontumor tissues [3]. There is no proof that the detected EPOR mRNA is translated into the protein. Tumor cell lines show weak or no detectable EPO binding using radioactive-labeled EPO [3]. Surface expression of EPORs on tumors has not been unambiguously demonstrated, and many studies were unable to detect physiologically relevant surface EPORs on tumor cells [3]. Twenty-three independent nonclinical studies showed superior outcomes, with no tumor progression occurring [3].
Treatment with recombinant EPO is not associated with lower efficacy of antitumor therapy. In two preclinical studies, both in vitro and in vivo in mammary carcinomas, EPOR expression and the effect of treatment with recombinant EPO on tumor growth were examined [8, 18]. In these breast cancer cell lines, EPOR expression was detected. No specific binding of EPO to the receptor was observed, suggesting that the EPOR was not functional. The treatment of these cells with EPO, at concentrations that can reasonably be achieved in vivo, did not result in tumor growth. EPO did not elicit a physiological tumor response, such as proliferation, migration, or signal transduction. Moreover, when the estrogen-positive breast cancer cells lines MCF-7 and MDA-MB231 were treated with cytostatics (i.e., paclitaxel and tamoxifen), EPO treatment did not interfere with the treatment response or antiproliferative effects of the cytotoxics. In mice, the addition of EPO to paclitaxel did not affect treatment outcome [8, 18].
Can ESAs Affect Cancer Biology Through Stimulation of Tumor Angiogenesis?
The role of EPO in cancer biology is not straightforward, and multiple factors may potentially be involved in tumor response to exogenous EPO [19].
There might be direct effects and/or indirect systemic effects in tumors (Fig. 1) that could lead to the modification of tumor cell growth and survival and chemoradiation response [19]. Through indirect effects on the tumor microenvironment, elevated Hb levels and changes in perfusion may have an effect on tumor outcome, although this was not recognized in many of the preclinical studies [19].
|
Moreover, in the same study, Hardee et al. [20] achieved high EPO concentrations within the tumor by intratumoral injection of 280,000 mU EPO in a volume of 7 µl. Again, no systemic effect on hematocrit levels was observed. Because erythroid progenitor cells express EPORs on their surface and are highly sensitive, a hematopoietic response should have been initiated if only moderate amounts of the 280,000 mU of EPO had reached the systemic circulation. Vice versa, it is questionable how a much lower systemic concentration of EPO during anemia therapy could ever reach a concentration in the tumor tissue high enough to affect angiogenesis or the tumor itself.
Finally, EPO binding to the EPOR surface homodimer causes a conformational change and results in the activation of the phosphorylation of its intracellular domain (Fig. 1) [19, 21]. This activation triggers a cascade of intracellular downstream events, including the Janus kinase 2–STAT5 pathway, which triggers proliferation, apoptosis resistance, and differentiation in erythroid progenitor cells [19, 21]. There are two important considerations to bear in mind regarding the effect of EPO on the signaling pathway in tumor cells. Although increased phosphorylation and thus activation of signaling kinases has been found in tumor cells, treatment with EPO did not cause an increased growth rate in the tumor cells [16, 17]. Moreover, the EPO concentration that will be used in the clinical setting for the treatment of anemia will be much lower than the concentrations that have been used in preclinical trials.
With regard to the three open questions mentioned earlier on the effects of ESAs on tumor proliferation, the following answers to these are based on our current understanding of the EPOR.
Obviously, the inability to detect EPOR protein on the cell surface caused by unreliable reagents is a lack of evidence, but not evidence of absence. The difficulty in generating appropriate antibodies may, in fact, be a problem of the antigen, because the EPOR as a preformed dimer in the cell membrane could be a difficult epitope to generate specific antibodies [19]. The absence of specific binding of radiolabeled EPO mentioned above [7, 8] excludes the possibility that EPOR mRNA-expressing tumor cells, in general, carry EPOR protein on their surfaces, but leaves open the possibility that single tumor cells may have EPOR proteins that bind EPO. Still, as discussed above, the concentrations needed to elicit any effect on a tumor cell in vitro were much higher than anyone would expect to achieve in a malperfused tumor. If biologically active EPO is produced by tumor cells themselves, it might, like many other growth factors produced by tumors, affect other cells in a paracrine way. But this setting is not the concern of physicians treating anemia associated with cancer or using cancer treatment with systemic EPO within the approved indication.
|
DO ESAS INCREASE THE INCIDENCE OF VTES? |
|---|
|
TopAbstractIntroductionDo ESAs Promote Tumor...Do ESAs Increase the...Author ContributionsReferences |
|---|
Cancer patients have a significantly higher risk for VTEs, with an incidence of thrombosis in the range of 4%–20% [23, 24]. VTEs are a leading cause of death in cancer patients. The risk for a VTE is 4.1 times higher in cancer patients than in patients without cancer and 6.5 times higher in patients receiving chemotherapy [25]. Whereas the clinical rates of VTEs are reported to be in the range of 4%–20% in cancer patients, the incidence of VTEs was found to be 50% in cancer patients at autopsy [23, 24, 26, 27]. VTEs are thus often not diagnosed while the patient is alive, and constitute a serious complication in cancer patients. A fatal pulmonary embolus and VTE recurrence are about three times more likely to occur in a cancer patient than in a patient without cancer [27]. A patient with a VTE has up to a 25% chance of having a pulmonary embolism, and a patient with a pulmonary embolism has a death rate up to 20% [23–27].
Risk Factors for VTEs
The risk for a cancer patient developing a VTE is complicated and can include multiple factors [28]. Risk factors for VTEs in cancer patients include patient-, cancer-, and treatment-related factors (Table 1). The age of the patient, the existence and nature of comorbidities, the type of primary cancer, the stage of disease, the type of cancer treatment, etc., can all affect the incidence of VTEs. Most VTEs occur within 6 months of diagnosis. Also, patients with progressive disease or current metastatic disease have a higher risk for a VTE, and surgery and chemotherapy or hormonal therapy are the major risk factors in these patients [28].
|
VTE Pathophysiology
Advanced cancer is associated with a hypercoagulable state and tissue factor (TF) expression [30]. TF is the physiologic initiator of coagulation and a regulator of tumor angiogenesis [31]. Pathways of activation of coagulation in cancer are summarized in Figure 2. Tumor cells produce TF and cancer procoagulant (CP), which start the extrinsic pathway by activating factors VIIa and Xa. Tumor necrosis factor and interleukin-1 induce TF expression on monocytes and on endothelial cells, activating factor VIIa [30]. Thrombin induces platelet aggregation. Thus, these factors increase the thrombophilic state. TF can initiate a hypercoagulable state and thrombosis. TF can also be an angiogenic-promoting factor in primary tumor growth and it also interferes with the spread of metastatic cells [30].
|
The thrombophilic mutations factor V G1691A (Leiden), prothrombin C677T (PT), and methylene tetrahydrofolate reductase (MTHFR) are the three most frequent mutations found in the white population, and these patients have a greater risk for developing VTEs [34]. However, in cancer patients with VTEs, testing for these thrombophilic mutations (Factor V Leiden, PT, MTHFR) is useful only if there is a previous personal or family history of VTE [34].
Chemotherapy treatment is associated with an increased risk for thrombosis [33]. Doxorubicin and epirubicin downregulate endothelial protein C receptor and impair the activated protein C (APC) pathway [35]. This creates a state similar to factor V Leiden. Therefore, the conversion of protein C to APC is hampered [35]. After treatment with these anthracyclines, 25% of patients had a low APC level [35]. This might be a contributing factor in chemotherapy-induced thrombophilia [35]. In 62 patients with multiple myeloma, 23% had APC resistance at baseline and 50% of these patients developed a VTE [36]. The incidence of VTEs was higher in the thalidomide group of patients [36]. In 1,178 patients with solid tumors, 109 had resistance to APC; of these, 36 patients had the factor V Leiden mutation and were excluded from the analysis [37]. On data available for 30 of 31 patients who were treated, the acquired APC resistance normalized after therapy [37].
VTEs with ESAs in Oncology Studies
Aapro and colleagues conducted the Breast Cancer–Anemia and the Value of Erythropoietin study [38]. In that randomized study, metastatic breast cancer patients were treated with anthracycline- and/or taxane-based chemotherapy and were randomly assigned to either epoetin beta or control. Although patients who received epoetin beta experienced more thromboembolic events (TEEs) than controls (13% versus 6%; p = .012), there was no significant difference in serious TEEs and no difference in TEE-related deaths (Table 2) [38]. There was also no significant difference in terms of overall survival (hazard ratio [HR], 1.07; 95% confidence interval [CI], 0.87–1.33; p = .522) or in progression-free survival (HR, 1.07; 95% CI, 0.89–1.30; p = .448) between the two groups.
|
|
In clinical trials, the Hb level at which treatment with ESAs is stopped may relate to the relative risk for a thrombotic event [24, 42]. When the target Hb is 13g/dl, the relative risk for a VTE is 0.70 [24, 42]. When the target Hb level is 13–14 g/dl, the relative risk rises steeply to 1.71 [24, 42]. When the Hb level is allowed to rise up to 15 g/dl, the relative risk for a thrombotic event is 1.92 [24, 42]. Thus, if the target Hb level is >13 g/dl, there is a higher risk for a thrombotic event occurring.
The 2006 Cochrane meta-analysis of clinical trial data involved 35 trials and 6,769 cancer patients receiving either epoetin or darbepoetin [43, 44]. Treatment with epoetin or darbepoetin was statistically significant in reducing the risk for RBC transfusions and improved the hematologic response [43, 44]. In this meta-analysis, the overall relative risk for TEEs in patients receiving treatment with epoetin or darbepoetin was 1.67 (95% CI, 1.35–2.06), representing a higher risk than with placebo [43, 44]. When the analysis is confined to studies involving epoetin alfa within and outside labeling, this increased risk is somewhat smaller (relative risk, 1.42; 95% CI, 1.13–1.93). The trials included patient groups that fell within the current labeling indications for epoetin or darbepoetin and also off-labeling (e.g., nonanemic patients and/or targeted Hb levels higher than product label recommendations) [43, 44]. It seems as though a higher incidence of TEEs occurs in off-label patients, particularly in trials that targeted Hb levels higher than those recommended by current ESA labeling. The authors of this meta-analysis mention two studies in particular that were off-label. In one phase III study, the majority of enrolled metastatic breast cancer patients undergoing chemotherapy were not anemic at baseline, and epoetin alfa was used with the aim of maintaining Hb levels of 12–14 g/dl [22]. In that study, the proportion of patients who experienced fatal TEEs was 1.3% in the epoetin alfa arm, versus 0.6% in the placebo arm [22]. In the second study, 300 U/kg epoetin beta was used in an attempt to sensitize head and neck tumors to radiation therapy, and the targeted Hb levels were 12–14 g/dl in women and 13–15 g/dl in men [1]. In the epoetin beta arm, the cardiac death rate was 5.5%, and this figure was 3% in the patients receiving placebo [1].
Bennett and colleagues conducted a meta-analysis of VTE and mortality rates in phase III trials of ESAs versus control (placebo or standard care) for the treatment of cancer-associated anemia [28]. Information on VTEs was extracted from 38 clinical trials involving 8,172 patients. Cancer patients treated with ESAs had a greater VTE risk than control patients (7.5% versus 4.9%; relative risk, 1.57; 95% CI, 1.31–1.87) and a higher mortality risk (HR, 1.10; 95% CI, 1.01–1.20) [28].
In summary, in the ESA treatment–related trials, the HR is around 1.09; however, this includes off-label use, which can result in a higher incidence of VTEs, particularly in trials that target Hb levels higher than those recommended by current ESA labeling and trials that enroll patients who are nonanemic at baseline. VTEs are common in cancer patients, and the relative risk for VTEs is in the range of 1.02–4.34 in cancer patients, depending on the type of cancer. In studies with a higher risk for VTEs in comparison with the control group, the cancer-inherent VTE risk is the same in both groups. From meta-analyses of ESA studies, the relative risk for a VTE falls within the ranges normally observed in the cancer patient population. The incidence of VTEs appears not to be influenced by the ESA dose. It appears that the elevated risk for a VTE associated with ESAs is a class effect because TEEs were seen in studies done with the three types of ESAs (epoetin alfa, epoetin beta, and darbepoetin). No head-on comparisons among the three ESAs have been done regarding VTEs. It also has to be stressed that the vast majority of the studies did not have an active objective control (d-dimer, ultrasound, or other) for the presence or absence of VTEs, but the latter was reported as an observational adverse event. The incidence of silent VTEs in these studies is not known. Therefore, in ongoing and further studies with ESAs, an active control for VTEs should be included. In order to limit VTEs, the aim of ESA treatment should be confined to targeted Hb levels and the patient population recommended by the current ESA labeling. A higher incidence of VTEs should rarely be a problem if ESAs are used within labeling indications and where necessary with the use of anticoagulants.
|
AUTHOR CONTRIBUTIONS |
|---|
|
TopAbstractIntroductionDo ESAs Promote Tumor...Do ESAs Increase the...Author ContributionsReferences |
|---|
Provision of study materials or patients: Joachim Fandrey
Data analysis and interpretation: Joachim Fandrey, Mario Dicato
Manuscript writing: Joachim Fandrey, Mario Dicato
Final approval of manuscript: Joachim Fandrey
Other: Literature review: Joachim Fandrey, Mario Dicato
The authors take full responsibility for the content of this article and thank Rob Stepney, medical writer, and Julia O'Regan, Bingham Mayne and Smith, Edinburgh, supported by an educational grant from Ortho Biotech, a division of Janssen-Cilag Europe, for their assistance in preparing a first draft of the manuscript based on an oral presentation at a meeting held on November 20, 2008 in Sitges, Spain, organized by a Scientific Committee of Matti Aapro, Mario Dicato, Pere Gascón, Francesco Locatelli, Jerry Spivak, and Jay Wish.
|
REFERENCES |
|---|
|
TopAbstractIntroductionDo ESAs Promote Tumor...Do ESAs Increase the...Author ContributionsReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. Aapro An Update on Twenty Years of Anemia Management with Erythropoiesis-Stimulating Agents in Nephrology and Oncology/Hematology Oncologist, September 1, 2009; 14(suppl_1): 1 - 5. [Full Text] [PDF]
|
|
|
|
|
|
F. Locatelli and H. Becker Update on Anemia Management in Nephrology, Including Current Guidelines on the Use of Erythropoiesis-Stimulating Agents and Implications of the Introduction of "Biosimilars" Oncologist, September 1, 2009; 14(suppl_1): 16 - 21. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. Spivak, P. Gascon, and H. Ludwig Anemia Management in Oncology and Hematology Oncologist, September 1, 2009; 14(suppl_1): 43 - 56. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Locatelli and P. Gascon Is Nephrology More at Ease Than Oncology with Erythropoiesis-Stimulating Agents? Treatment Guidelines and an Update on Benefits and Risks Oncologist, September 1, 2009; 14(suppl_1): 57 - 62. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | CONTACT US | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| THE ONCOLOGIST | STEM CELLS | CME | ALPHAMED PRESS JOURNALS |
