Outcomes in Patients with Acute Myeloid Leukemia Preceded by Breast Cancer

As breast cancer has become a more curable disease, the risk of second malignancy either related to therapy or genetic alterations has become apparent. Recent data from the National Cancer Institute's SEER database showed an increased risk of Acute Myeloid Leukemia (AML) in female patients with Stage III breast cancer. To investigate the outcomes of AML after breast cancer treatment, we reviewed all annotated cases with AML preceded by breast cancer (BrCa) at our institution.

Between 2001 and 2011, the Moffitt Cancer Center Total Cancer Care database was used to identify patients with a diagnosis of AML preceded by breast cancer. Individual charts were subsequently reviewed. Chi square test was used to compare categorical variables in univariate analysis. Kaplan Meier estimates were used to calculate OS. Log rank test was used for comparison between the 2 groups and Cox regression analysis was used for multivariable analysis of survival. All analyses were conducted using SPSS version 19.0 software.

We identified 60 female patients with AML preceded by a diagnosis of breast cancer. Median age at AML diagnosis was 65 years (range 43–81). Median time from BrCa diagnosis to AML diagnosis was 7 years. Twenty nine (48%) had stage ≥ 2 disease at time of BrCa diagnosis. Forty nine (82%) of patients had t-AML, 36 (60%) of whom had received chemotherapy, either alone or in combination with radiation, for breast cancer management. Nineteen (32%) had adverse karyotype and 22 (37%) had MDS that preceded the diagnosis of AML. Thirty-six (60%) underwent induction chemotherapy for AML with an cytarabine/anthracycline-based regimen (most commonly “7+3”), and CR/CR_iwas achieved in 21 (54%). Ten (17%) patients eventually proceded to allogeneic hematopoietic cell transplant, with a median survival of 20 months (95% C.I. 10 –29.9 mo) Median survival following AML diagnosis was 10.4 mo (95% C.I. 5.4 – 15.4 months).

Univariate analysis of prognostic variables associated with overall survival are shown in Table 1. Variables significantly associated with inferior survival included history of MDS preceding AML, adverse karyotype, and no induction therapy for AML. Interestingly, prior BrCa chemotherapy did not adversely impact survival. None of the individual variables retained prognostic significance in multivariate analysis.

AML arising in the setting of prior BrCa has a generally poor prognosis that mimics outcomes in older adults with AML. Many traditional adverse prognostic factors in AML are present in patients with AML from BrCa, likely accounting for the poor outcomes. However, many patients are able to receive and benefit from induction chemotherapy and ultimately proceed to allogeneic transplant, indicating that traditional therapy has an important role in the management of AML from BrCa. Larger numbers of patients will be necessary to confirm the importance of varied risk factors and treatment options in this population. A comparison with matched historic control patients will also be presented.

No relevant conflicts of interest to declare.