Background: Recently, increasing numbers of women receiving adjuvant chemotherapy for breast cancer have also received granulocyte colony-stimulating factors (G-CSFs) or granulocyte-macrophage colony-stimulating factors (GM-CSFs). Although these growth factors support chemotherapy, their long-term safety has not been evaluated. We studied the association between G-CSF use and incidence of leukemia in a population-based sample of breast cancer patients.
Methods: Among women aged 65 years or older in the Surveillance, Epidemiology, and End Results-Medicare database who were diagnosed with stages I-III breast cancer from January 1, 1991, to December 31, 1999, we identified those who received G-CSF or GM-CSF concurrently with chemotherapy. We used Cox proportional hazards models to estimate hazard ratios for the association of treatment with G-CSF or GM-CSF and subsequent (through December 31, 2003) diagnosis of acute myeloid leukemia (AML) or myelodysplastic syndrome (MDS). All statistical tests were two-sided.
Results: Of 5510 women treated with chemotherapy, 906 (16%) received G-CSF or GM-CSF therapy, and 64 (1.16%) were subsequently diagnosed with either MDS or AML before a cancer recurrence. Use of G-CSF and GM-CSF was associated with more recent diagnosis, younger age, urban residence, fewer comorbidities, receipt of radiation therapy, positive lymph nodes, and cyclophosphamide treatment. Of the 906 patients who were treated with G-CSF, 16 (1.77%) developed AML or MDS; of the 4604 patients not treated with G-CSF, 48 (1.04%) developed AML or MDS. The hazard rate ratio for AML or MDS among those treated with G-CSF or GM-CSF compared with those who were not was 2.14 (95% confidence interval [CI] = 1.12 to 4.08). AML or MDS developed within 48 months of breast cancer diagnosis in 1.8% of patients who received G-CSF or GM-CSF but only in 0.7% of patients who did not (hazard ratio = 2.59, 95% CI = 1.30 to 5.15).
Conclusions: The use of G-CSF was associated with a doubling in the risk of subsequent AML or MDS among the population that we studied, although the absolute risk remained low. Even if this association is confirmed, the benefits of G-CSF may still outweigh the risks. Meanwhile, however, G-CSF use should not be assumed to be risk free.