Abstract
Background
Prostate cancer is the most common cancer diagnosis in men, and one of the leading indications for radiation therapy. The risk of resultant secondary leukemias has not been consistently established. We investigated the risk of all leukemias in a population-based cohort of patients (pts) with locoregional prostate cancer definitively treated with radiotherapy.
Methods
We queried the Surveillance, Epidemiology, and End Results (SEER) 17 registries to identify a cohort of men >20 years old (n = 183,268) with locoregional prostate adenocarcinoma newly diagnosed between January 1973 and December 2008. Pts who underwent initial treatment with radical prostatectomy (RP) were compared to pts receiving RP with external beam radiotherapy (RP w/EBRT) to investigate the risk of radiation-induced leukemias. These cohorts tend to be well matched regarding age, medical comorbidities and disease characteristics. All new leukemias occurring as a second primary cancer at least one year after the first diagnosis of prostate cancer were identified in SEER using the International Classification of Diseases for Oncology, Third Edition (ICD-O-3) morphology codes. Secondary leukemias included acute myeloid leukemia (AML); chronic myelogenous leukemia (CML); acute and chronic lymphocytic leukemia (ALL & CLL) and other categories as reported in SEER. Pts were observed from date of prostate cancer diagnosis until leukemia occurrence, death, or last date of follow-up. Univariate and multivariate analyses were performed using the Fine and Gray competing risk regression analysis with leukemia as a time-dependent endpoint and death from any cause or the diagnosis of any other second cancer as competing events. RP w/ EBRT group was compared with the RP cohort as the reference group, controlling for age. Hazard ratios (HR) with 95% confidence intervals (CIs) are reported.
Results
Median age was 67 years (yrs, range 22 – 105) at prostate cancer diagnosis: 67 yrs in RP and 68 yrs in RP w/ EBRT pts (p<0.0001); 158,913 (86.7%) were treated with RP and 24,355 (13.3%) with RP w/EBRT. Median follow-up was 7.6 yrs [(range, 1 – 35.5): 7.5 yrs in the RP group and 8.3 yrs in the RP w/ EBRT group, (p<0.0001)].
In total, 949 (0.5%) leukemia cases were identified: 336 (0.2%) acute leukemias [266 (0.2%) in the RP group and 70 (0.3%) in the RP w/ EBRT]; 538 (0.3%) chronic leukemias [462 (0.3%) in the RP group and 76 (0.3%) in the RP w/ EBRT] and 75 (0.04%) of unspecified histology. Histologic subtypes (per ICD-O-3 codes) were: AML (n=249), acute monocytic leukemia (n=18), ALL (n=24), other acute leukemias (n=45), other myeloid/monocytic/lymphocytic leukemias (n=48), aleukemic/subleukemic/NOS (n=27), CML (n=131) and CLL (n=407).
Median age at acute leukemia diagnosis was 77 yrs [(range, 50 – 101): 78 yrs in the RP group and 76 yrs in RP w/EBRT pts, (p=0.0271)] and for chronic leukemias was 76 yrs [(range, 47 – 101): 76 yrs in the RP group and 77 yrs in the RP w/EBRT pts, (p=0.50)].The median time to develop acute leukemias was 6.0 yrs [(range, 1 – 28.2): 6.1 yrs in the RP group and 5.7 yrs in the RP w/EBRT pts, (p=0.20)] and chronic leukemias was 6.9 yrs [(range, 1 – 29.8): 6.7 yrs in the RP group and 8.6 yrs in the RP w/EBRT pts, (p=0.0020)]. The cumulative incidence rate (CIR) at 20 years for acute leukemias was 0.24% for the RP pts vs. 0.32% for the RP w/EBRT pts (p=0.0196). The CIR at 20 years for chronic leukemias was 0.47% for the RP pts vs. 0.36% for the RP w/EBRT pts (p=0.10).
In univariate analyses, age >70 yrs (HR=1.40; CI, 1.13 – 1.74; p=0.0023), and those who received RP w/ EBRT (HR=1.49; CI, 1.14 – 1.94; p=0.0033) were significantly more likely to develop acute leukemias. In multivariate analysis, both advanced age (HR=1.40; CI, 1.13 – 1.74; p = 0.0023) and RP w/ EBRT (HR=1.49; CI, 1.14 – 1.94; p=0.0032), remained significantly associated with increased risk of acute leukemias. Radiation treatment was not significantly associated with the risk of developing chronic leukemias among pts treated with RP w/EBRT vs. RP [HR=0.91; CI, 0.72 – 1.16; p=0.45).
Conclusions
Among the best matched prostate cancer treatment cohorts, those who underwent EBRT following RP had a 49% increased risk of subsequent acute leukemias, although the absolute number of cases was low. Risk assessment in this cohort spans a time frame where radiation technologies have rapidly advanced and hence treatment period effects need to be considered in interpretation of results
Disclosures:
No relevant conflicts of interest to declare.
Is PSA density of the peripheral zone as a useful predictor for prostate cancer in patients with gray zone PSA levels?
