Abstract
Risk of myelodysplastic syndrome (MDS) and acute myeloid leukemia (AML) post-breast cancer treatment with adjuvant chemotherapy and granulocyte colony-stimulating factors (G-CSF) is not fully characterized. Our objective was to estimate MDS/AML risk associated with specific breast cancer treatments. We conducted a retrospective cohort study of women aged ≥66 years with stage I-III breast cancer between 2001 and 2009 using the Surveillance, Epidemiology, and End Results-Medicare database. Women were classified as receiving treatment with radiation, chemotherapy, and/or G-CSF. We used multivariable Cox proportional hazards models to estimate adjusted hazard ratios (HR) and 95 % confidence intervals (CI) for MDS/AML risk. Among 56,251 breast cancer cases, 1.2 % developed MDS/AML during median follow-up of 3.2 years. 47.1 % of women received radiation and 14.3 % received chemotherapy. Compared to breast cancer cases treated with surgery alone, those treated with chemotherapy (HR = 1.38, 95 %-CI 0.98-1.93) and chemotherapy/radiation (HR = 1.77, 95 %-CI 1.25-2.51) had increased risk of MDS/AML, but not radiation alone (HR = 1.08, 95 % CI 0.86-1.36). Among chemotherapy regimens and G-CSF, MDS/AML risk was differentially associated with anthracycline/cyclophosphamide-containing regimens (HR = 1.86, 95 %-CI 1.33-2.61) and filgrastim (HR = 1.47, 95 %-CI 1.05-2.06), but not pegfilgrastim (HR = 1.10, 95 %-CI 0.73-1.66). We observed increased MDS/AML risk among older breast cancer survivors treated with anthracycline/cyclophosphamide chemotherapy that was enhanced by G-CSF. Although small, this risk warrants consideration when determining adjuvant chemotherapy and neutropenia prophylaxis for breast cancer patients.