The results of the present meta-analysis show no evidence that treatment with epoetin beta impairs survival or promotes tumour progression in patients with cancer, at least during the period of observation.
These data are concordant with evidence from some preclinical studies that suggest that epoetin may improve cyto- and radio-sensitivity and impair progression of various tumours. Correction of anaemia by epoetin has been reported to improve cyclophosphamide cytotoxicity in a rat model (Thews et al, 2001
) and to restore radiosensitivity of experimental human tumours in nude mice (Stuben et al, 2003
). Also, one study using a murine myeloma model reported that epoetin induced tumour regression and antitumour immune responses (Mittelman et al, 2001
). However, other preclinical data, primarily obtained from cell lines, have suggested that epoetin may diminish the effects of cytostatic agents or promote tumour cell growth in vitro
(Acs et al, 2003
; Farrell and Lee, 2004
Some clinical studies have suggested reduced tumour progression and increased survival in anaemic patients with cancer treated with epoetin. In a non-randomised study of 191 patients undergoing neoadjuvant chemoradiotherapy and resection for squamous cell carcinoma of the oral cavity or oropharynx, treatment with epoetin was associated with significantly better local control and survival compared with an untreated historical control group (Glaser et al, 2001
). Similarly, in a preliminary report, treatment with epoetin improved tumour control and survival in a randomised controlled trial of 385 patients with various pelvic malignancies receiving radiotherapy (Antonadou et al, 2001
). A nonsignificant trend towards a survival benefit with epoetin has also been suggested by the results of a randomised, double-blind, placebo-controlled trial of 375 patients with solid or non-myeloid haematological malignancies receiving non-platinum-based chemotherapy (Littlewood et al, 2001
). In addition, a recent meta-analysis by Bohlius et al (2005)
of randomised controlled trials in patients with cancer also reported a trend towards improved survival with epoetin (HR 0.84, 95% CI 0.69–1.02, n
=2805). Although subsequent expansion of this analysis showed a shift towards increased mortality risk and increased risk from thromboembolic events, this was suggested by the authors as being possibly due to methodological limitations such as baseline imbalances (Bohlius et al, 2006
). One prospective study investigating the role of dose-dense chemotherapy in patients with early breast cancer has demonstrated that epoetin alfa had no adverse influence on survival (Möbus et al, 2004
). A similarly neutral effect on survival was also reported in a meta-analysis of four randomised, double-blind, placebo-controlled studies of darbepoetin alfa in patients with a variety of tumour types (Hedenus et al, 2005
In contrast, two prospective, randomised studies in which survival was a primary end point have been less positive, and have even suggested that therapy with epoetin could have a detrimental effect on survival (Henke et al, 2003
; Leyland-Jones et al, 2005
). The study by Henke et al (2003)
was a double-blind study in 351 patients with carcinoma of the oral cavity, oropharynx, hypopharynx, or larynx treated with curative radiotherapy. In this study, epoetin beta was reported to increase Hb levels relative to placebo. However, there was also a reduction in both locoregional progression-free survival (adjusted RR 1.62; 95% CI 1.22, 2.14; P
<0.001) and overall survival (adjusted RR of death 1.39; 95% CI 1.05, 1.84; P
=0.02) relative to placebo. The study by Leyland-Jones et al (2005)
, in which patients receiving first-line chemotherapy for metastatic breast cancer were treated with epoetin alfa for the prevention of anaemia, was terminated early because of a significant (P
=0.01) difference in 12-month survival between patients in the epoetin alfa (70%) and placebo (76%) groups (Leyland-Jones et al, 2005
However, these two studies need to be interpreted with caution, as baseline imbalances in prognostic factors favoured placebo in both. Other limitations of these two studies have also been highlighted (Dunst, 2004
; Leyland-Jones and Mahmud, 2004
; Vaupel and Mayer, 2004
). Moreover, it should be noted that both trials were investigational in nature and both used epoetin outside of its currently approved indications (in predominantly mild or non-anaemic patients, many of whom attained higher than recommended Hb levels with epoetin therapy). These observations, along with reports that the apparent negative effect of epoetin beta observed in the Henke et al (2003)
study simply reflects over-treatment (Vaupel and Mayer 2004
), led to exclusion of the Henke et al study (2003)
from this meta-analysis.
This meta-analysis takes into account epoetin beta studies not included in the previous meta-analysis of epoetin randomised controlled trials (Bohlius et al, 2005
). The findings, together with evidence from other studies (Food and Drug Administration, 2004
) suggest there is no indication of an increase in early disease progression and that epoetin does not impair survival in patients with anaemia when used as currently approved. Furthermore, the results of the present study indicate a trend towards a reduced rate of tumour progression with epoetin beta treatment.
Long-term follow-up data from one of the studies included in this analysis provide further evidence that epoetin beta has a neutral effect on survival. In this randomised, double-blind trial of severely anaemic patients with lymphoproliferative malignancy, median survival was 17 months with epoetin beta and 18 months with placebo and Kaplan–Meier curves for survival were similar for both treatment groups (Österborg et al, 2005
Cancer and its treatment are known predictors of risk for thromboembolism, with absolute risk depending on tumour type, stage, and extent of cancer and treatment, and on other factors such as age, surgery, immobilisation, and comorbid features (Lee and Levine, 2003
). Treatment with epoetin has been associated with occasional reports of thromboembolic events and this is reflected in the product label. The present meta-analysis, as well as that by Bohlius et al (2005)
, showed a marginal increase in incidence of thromboembolism in patients receiving epoetin. Importantly, in our analysis, epoetin was not associated with any increase in the proportion of patients with thromboembolic events leading to death.
This meta-analysis of nine controlled trials, which represents all randomised, controlled trials of epoetin beta within its current indication, shows no evidence of a negative effect on survival or thromboembolic-related mortality of epoetin beta in patients with cancer. Moreover, these data show a small but statistically significant benefit in slowing tumour progression compared with placebo or standard transfusion therapy. Within its licensed indication, our results indicate that epoetin beta is a safe treatment of anaemia for patients with cancer.