Early secondary acute myelogenous leukemia in breast cancer patients aftertreatment with mitoxantrone, cyclophosphamide, fluorouracil and radiation therapy

Background: The topoisomerase II-targeted drugs, epipodophyllotoxins and an thracyclines, have been shown to induce therapy-related AML (t-AML) charact erized by a short latency period after chemotherapy, the absence of prior m yelodysplastic syndrome and stereotyped chromosome aberrations. Few reports have been published on patients treated with the anthracenedione mitoxantr one which also targets topoisomerase II. We observed 10 cases of such t-AML over a 7-year-period in breast cancer patients treated with mitoxantrone c ombined with fluorouracil, cyclophosphamide and regional radiotherapy, and in three cases with vindesine. Patients and methods: We retrospectively analyzed patients referred to our hospital for AML with a past history of polychemotherapy for breast cancer, including mitoxantrone, either as adjuvant (8 patients)/neoadjuvant (1 pat ient) therapy or for metastatic disease (1 patient). We studied the probabi lity of developing t-AML in a prospective series of 350 patients treated wi th an adjuvant FNC regimen (mitoxantrone, fluorouracil, cyclophosphamide) a nd radiation therapy. Results: The median age was 45 years (range 35-67). t-AML developed 13-36 m onths (median 16) after beginning chemotherapy for breast cancer, and 4-28 months (median 10.5) after ending treatment. As described in t-AML followin g treatment with epipodophyllotoxins or anthracyclines, we found a majority of FAB M4, M5 and M3 phenotypes (7 of 10), and characteristic karyotype ab normalities that also can be found in de novo AML: breakpoint on chromosome 11q23 (3 patients), inv(16)(p13q22) (2 patients), t(15;17)(q22;q11) (1 pat ient), t(8;21)(q22;q22) (1 patient) and del(20q)(q11) (1 patient). The prog nosis was poor. All patients died of AML shortly after diagnosis. Since two patients had been enrolled in a prospective trial for the treatment of bre ast cancer which included 350 patients, the probability of developing t-AML was calculated to be 0.7% from 25-40 months, using the Kaplan-Meier method (95% confidence interval (95% CI): 0.1-4.5). Conclusions: The combination of mitoxantrone with cyclophosphamide, fluorou racil, and radiation therapy can induce t-AML, as with other topoisomerase II-targeted drugs. Despite a low incidence, the prognosis appears to be poo r.