A Comparative Investigation of Therapeutic Approaches for Primary Prostate Cancer and Risk of Acute Myeloid Leukemia.

Radiation is a common therapeutic approach for prostate cancer. However, previous studies have reported that exposure to ionizing radiation may initiate leukemogenesis. We utilized a case-cohort design to investigate whether prostate cancer patients treated with external beam radiation therapy (EBRT) demonstrated greater risk of developing acute myeloid leukemia (AML) than prostate cancer patients that underwent surgery. Cases (n=224) were identified using Surveillance, Epidemiology, and End Results (SEER), a population-based cancer registry representative of the United States. Cases were restricted to men with primary prostate cancer diagnosed from January 1986-December 1994 and subsequent diagnosis of AML within 15 years of follow-up. Controls (n=256) were randomly sampled from all available primary prostate cancer diagnoses in SEER that corresponded to the study period selected for cases. We excluded patients that developed secondary cancers other than AML when selecting cases and controls because of competing risks. The outcome (AML) was dichotomized and mutually exclusive exposure categories (no treatment [reference], EBRT, surgery, or combination EBRT and surgery) were included as dummy variables in the model. Logistic regression was used to directly estimate 5-, 10-, and 15-year risk ratios (RRs) for AML. Prostate cancer patients treated with EBRT, surgery, or a combination of EBRT and surgery demonstrated a greater risk of developing AML than untreated prostate cancer patients after controlling for age, race, grade, and stage of prostate cancer. However, patients who received EBRT had consistently greater risk of developing AML across the time intervals than patients that underwent surgery (EBRT: 5-yr RR=1.77, 95% CI=0.75, 4.20; 10-yr RR=2.70, 95% CI=1.29, 5.65; 15-yr RR=2.76, 95% CI=1.34, 5.69. Surgery: 5-yr RR=1.42, 95% CI=0.63, 3.20; 10-yr RR=1.51, 95% CI=0.76, 3.00; 15-yr RR=1.56, 95% CI=0.80, 3.04). Patients who received a combination of EBRT and surgery exhibited risk estimates similar to EBRT alone (5-yr RR=1.79, 95% CI=0.65, 4.94; 10-yr RR=2.34, 95% CI=0.97, 5.66; 15-yr RR=2.61, 95% CI=1.09, 6.23). Our data indicate that prostate cancer patients who received EBRT had almost a 2-fold greater risk of developing AML 10-years after prostate cancer diagnosis than patients who underwent surgery. The finding may be of interest when considering treatment options, but requires further elucidation of specific predictors that may contribute to increased AML risk among prostate cancer patients. Furthermore, our data indicate that AML risk begins to plateau 10-years after prostate cancer diagnosis for patients that received EBRT and may assist with defining induction and latency. Data limitations prevented addressing potential confounding by independent or concomitant chemotherapy and socioeconomic status. The magnitude of bias introduced by the inability to incorporate such variables is difficult to speculate without further data. Future investigations with access to robust data may be able to address the limitations herein. Our investigation thus far provides the largest comparison of therapeutic approaches and AML risk among prostate cancer patients.