Acute myeloid leukaemia or myelodysplastic syndrome following use of granulocyte colony-stimulating factors during breast cancer adjuvant chemotherapy

Abstract

It is still controversial whether granulocyte colony-stimulating factors (G-CSF) per se increases the risk of secondary acute myeloid leukaemia (AML) or myelodysplastic syndrome (MDS) or whether the requirements for its use cause an increased risk of AML/MDS. In fact, the apparent effects of G-CSF may be caused by the high doses of epirubicin and cyclophosphamide received by the patients included in the treatment arms to which G-CSF was assigned [1]. In the manuscript by Hershman et al. [2], of the 906 women with stages I–III breast cancer, who were treated with G-CSF concurrently with adjuvant chemotherapy, 16 (1.77%) developed AML/MDS. The hazard rate ratio for AML or MDS among those treated with G-CSF or granulocyte–macrophage colony-stimulating factor compared with those who were not was 2.14. One of the issues concerning this observational study is that no information was available about the cumulative dose and duration of G-CSF treatment [3]. Smith et al. [4] found a positive association between the use and the dose of G-CSF and the risk of secondary acute leukaemia (sAL) in patients receiving standard anthracycline and dose-intensified cyclophosphamide. Eighteen of the 22 patients enrolled in B-25 trial and diagnosed with AML/MDS received G-CSF doses in excess of the median dose administered to all patients (242 lg/kg). Controlling for treatment arm, patient age, and operation, the estimated risk of AML/MDS for patients receiving more than the median dose of G-CSF, relative to those receiving the median dose or less, was 3.58 (P = 0.02). On the other hand, in the study by Praga et al. [5], the patients allocated to receive G-CSF according to protocol received median cumulative doses of epirubicin and cyclophosphamide significantly (P = 0.0001) higher than the corresponding median cumulative doses in patients not mandated to receive G-CSF. According to Wilking et al. [6], the major pitfall with the tailored and dose-escalated FEC arm (5-fluorouracil, epirubicin, cyclophosphamide) was the increased risk for secondary AML or MDS. All patients received prophylactic filgrastim (5 lg/kg daily subcutaneously) on days 2–14. The dose of epirubicin in the tailored FEC was much higher than in the standard FEC followed by marrow-supported high-dose therapy (780 vs. 181 mg/m, respectively). By contrast, Crump et al. [7] found no case of sAL among patients given epirubicin-based adjuvant chemotherapy plus G-CSF and Citron et al. [8] reported no correlation between the use of G-CSF and the incidence of sAL among patients randomized to standard or dose-dense chemotherapy. In our experience [9], the crude incidence of sAL after adjuvant high dose epirubicin plus cyclophosphamide with G-CSF support was 0.41%. Future analyses should investigate the possible role of G-CSF cumulative dose and duration of exposure in causing leukaemia in patients treated with adjuvant epirubicin and cyclophosphamide.