The impact of androgen deprivation and pelvic radiation on the development of bladder cancer.

Abstract

439 Background: The male predilection of urothelial bladder cancer (UBC) as well as the expression of the androgen receptor in bladder tissue point to the role for androgens in UBC tumorigenesis. Animal studies demonstrate a potential role for androgen deprivation in diminishing UBC. More recently, two separate groups demonstrated decreased rates of both primary and recurrent UBC in prostate cancer patients previously receiving androgen deprivation therapy (ADT). Given the common use of radiation therapy (RT) in the treatment of localized prostate cancer, and previous data supporting the increased frequency of UBC in prostate cancer patients treated with RT, the interaction between ADT and RT in UBC remains an important consideration. Methods: Using the linked SEER-Medicare database, we investigated the interactions among ADT, RT and UBC by performing a retrospective cohort study of elderly (age 66-99) prostate cancer patients diagnosed between 1999-2011. Kaplan-Meier analysis and Cox proportional modeling were used to determine the risk of developing secondary bladder cancer after prostate cancer treatment (based on exposure to ADT, RT, both, or neither). All analyses were two-sided. Results: Of 121,927 patients with primary prostate cancer, 1,466 (1.20%) developed subsequent UBC with a median follow up of 5.08 years (range 0.003-12.00). Compared with patients never receiving ADT or RT (n = 43,809), the hazard ratios for the development of secondary bladder cancer in patients ever receiving ADT but no RT (n = 14,009), RT but no ADT (n = 16,672), or both ADT and RT (n = 17,465) were 0.76 (95% confidence interval [CI]: 0.63-0.91 ), 0.73 (95% CI: 0.64-0.83), and 0.69 (95% CI: 0.61-0.79), respectively. Conclusions: Both ADT and RT are independently associated with a reduced risk of secondary bladder cancers in prostate cancer patients. The finding of decreased UBC incidence in patients receiving RT was surprising, and in direct contradiction to previous studies of similar patient populations. Possible explanations include differences in cohort selection, changes in RT delivery, and differences in control groups.