BACKGROUND: Epoetin beta (NeoRecormon®) 30 000 IU/week raises hemoglobin levels, reduces transfusion need and improves quality of life in patients with cancer. Recent studies have also suggested that epoetin therapy may impact upon outcomes in these patients. A meta-analysis was performed to investigate the effects of epoetin beta on survival, tumor progression and thromboembolic events in patients with hematological malignancies receiving chemotherapy.

METHODS: Data were pooled from all five randomized, controlled (placebo or standard care) clinical trials of epoetin beta that included anemic patients with hematological malignancies receiving chemotherapy. The study that showed epoetin beta once weekly to be as effective with the same safety profile as a three times weekly regimen was excluded from this analysis because of the lack of a non-epoetin-treated group. These studies were not designed to assess duration of survival and in all except one there was no follow-up beyond the duration of study treatment plus an additional 4-week period. Deaths reported during the study plus 4 weeks were therefore recorded in this analysis. All adverse event reports were assessed for evidence of disease progression or thromboembolism. Data were analyzed by standard Kaplan-Meier methods, Cox regression and log-rank tests.

RESULTS: A total of 791 patients with hematological malignancies were included (epoetin beta, n=461; control, n=330) (the difference in number was due to two of the five studies containing multiple epoetin beta treatment groups versus one control group). The majority was diagnosed with lymphoma (56%) or multiple myeloma (42%); the remaining patients (2%) were diagnosed with acute myeloid leukemia. Treatment groups were well balanced with regard to baseline demographic characteristics. There were no obvious differences in baseline tumor stage between treatment groups, although these data were not consistently collected across studies. The median weekly dose of epoetin beta was 30 000 IU. Death rates were similar in both the epoetin beta and control groups (0.39 vs 0.37 deaths/patient year). There was no indication of a difference in survival with epoetin beta compared with control (relative risk 1.04, 95% CIs 0.69, 1.55, p=0.86). The rate of disease progression was lower in the epoetin beta group compared with the control group (0.69 vs 0.81 events/patient year). The results from Kaplan-Meier estimates and Cox regression did not indicate an increased risk of disease progression with epoetin beta compared with control (relative risk 0.84, 95% CIs 0.62, 1.13, p=0.25). In fact, the hazard rate of 0.84 indicated a 16% reduction in the risk of an event when treated with epoetin beta. Thromboembolic event rates were also similar in the epoetin beta and control groups (0.17 vs 0.14 events/patient year), corresponding to crude rates of 5.4% and 4.8% (observation times: 147 and 112 patient years). These event rates are well within the range of those reported in the literature.

CONCLUSIONS: Epoetin beta has no effect on short-term survival, tumor progression or thromboembolic events when used to treat anemic patients with hematological malignancies, and the risk-to-benefit ratio of epoetin beta therapy remains favorable.

Author notes

Corresponding author

Sign in via your Institution