Emerging Risks of AML/MDS and Other Myeloid Neoplasms Following Chemotherapy for First Primary Malignancy, 2000-2012

Introduction. Treatment-related acute myeloid leukemia/myelodysplastic syndrome (tAML/MDS) is a rare but often fatal complication of systemic therapy for primary malignancy. Leukemogenicity of specific agents is variable, with particularly high risks associated with platinum-containing agents, certain alkylating agents, topoisomerase II inhibitors, and purine analogs. Current treatment practices increasingly include (neo)adjuvant and multiple courses of systemic therapy for a number of first primary malignancies. However, no large-scale study has quantified risks of tAML/MDS and other myeloid neoplasms after chemotherapy in the modern treatment era.

Methods. We identified a cohort of 746,007 adults who were initially treated with chemotherapy and survived ≥1 year following diagnosis with first primary malignancy during 2000-2012, as reported to 17 US population-based cancer registries from the Surveillance, Epidemiology, and End Results program. Risks for second primary AML/MDS, chronic myeloid leukemia (CML), and other myeloproliferative neoplasms (MPNs) and MDS/MPNs were compared to that expected in the general population (based on age-, race-, sex- and calendar period-specific incidence rates) using standardized incidence ratios (SIRs).

Results. tAML/MDS was identified in 2071 individuals following chemotherapy, four times more than expected based on general population rates (SIR=4.1, 95%CI=3.9-4.2). We identified novel elevations in tAML/MDS risk after chemotherapy for most gastrointestinal malignancies, including the oral cavity/pharynx (N=45, SIR=2.6, 95%CI=1.9-3.5), esophagus (N=28, SIR=4.3, 95%CI=2.9-6.2), liver (N=10, SIR=2.6, 95%CI=1.2-4.8), stomach (N=22, SIR=2.7, 95%CI=1.7-4.0), rectum (N=65, SIR=1.5, 95%CI=1.2-1.9), and anus (N=22, SIR=3.6, 95%CI=2.3-5.5), but not colon (N=67, SIR=1.1, 95%CI=0.8-1.3). Novel increased risks of tAML/MDS also were observed after chemotherapy for cancers of the pancreas (N=15, SIR=3.3, 95%CI=1.8-5.4), larynx (N=20, SIR=4.2, 95%CI=2.6-6.5), bladder (N=30, SIR=1.8, 95%CI=1.2-2.6), and melanoma (N=4, SIR=3.7, 95%CI=1.0-9.6)

Similar to previous studies, tAML/MDS occurred most commonly after female breast cancer (N=543, SIR=4.1, 95%CI=3.8-4.5), non-Hodgkin lymphoma (NHL; N=515, SIR=7.3, 95%CI=6.7-7.9), and lung cancer (N=185, SIR=4.1, 95%CI=3.5-4.7). We further confirmed previous observations of strikingly elevated risks of tAML/MDS after chemotherapy for cancers of the bone (N=10, SIR=35.1, 95%CI=16.9-64.6), testis (N=18, SIR=15.6, 95%CI=9.2-24.6), and soft-tissue (N=20, SIR=12.6, 95%CI=7.7-19.4), and more modestly elevated risks of tAML/MDS after chemotherapy for cancers of brain (N=18, SIR-7.8, 95%CI=4.6-12.4), ovary (N=84, SIR=5.5, 95%CI=4.3-6.7), endometrium (N=28, SIR=4.4, 95%CI=2.9-6.3), cervix (N=22, SIR=4.4, 95%CI=2.8-6.6), and prostate (N=15, SIR=2.7, 95%CI=1.5-4.4), as well as Hodgkin lymphoma (N=54, SIR=8.7, 95%CI=6.6-11.4), chronic lymphocytic leukemia (N=52, SIR=7.7, 95%CI=5.8-10.2), and myeloma (N=102, SIR=6.3, 95%CI=5.1-7.6). Risks were non-significantly heightened with radiotherapy plus chemotherapy for breast, lung, and stomach cancers compared with chemotherapy alone.

Elevated risks also were observed for CML after chemotherapy for lung cancer (N=12, SIR=2.5, 95%CI=1.3-4.4), breast cancer (N=35, SIR=1.8, 95%CI=1.3-2.5), and NHL (N=16, SIR=2.1, 95%CI=1.2-3.4), and for chronic myelomonocytic leukemia after chemotherapy for breast cancer (N=15, SIR=3.0, 95%CI=1.7-5.0) and NHL (N=16, SIR=4.2, 95%CI=2.4-6.9). In contrast, risks were not increased for other MPNs after chemotherapy for any first primary malignancy.

Conclusions. Despite the availability of modern cancer chemotherapy and targeted agents, risks of tAML/MDS are elevated across a broad spectrum of first primary cancers and extend to other myeloid neoplasms. Risks are consistent with those expected from expanded use of leukemogenic systemic therapy, particularly for specific cancers, such as cervical cancer, that are commonly treated with platinum-based regimens. More research is needed to quantify risks associated with specific agents and doses in the (neo)adjuvant and treatment setting. Risks for treatment-related myeloid neoplasms should be weighed against benefits of systemic therapy.