Intermediate versus short-course hormone therapy and mortality in men with high-risk prostate cancer treated with radiation therapy.

Abstract

83 Background: Radiation therapy (RT) plus 28-36 months of hormonal therapy (HT) is standard-of-care for men with high-risk prostate cancer (HRPC) based on randomized trials comparing these HT durations to 4-6 months. However, it is unknown whether shorter durations of HT may also decrease mortality. We evaluate the impact of intermediate-course HT on the risk of all-cause mortality (ACM) in men with HRPC treated with RT. Methods: The study cohort comprised 554 men with HRPC (PSA > 20; Gleason score 8 or higher; or clinical stage T2c or higher) consecutively treated at the Chicago Prostate Cancer Center between 1997 and 2007. All men received brachytherapy with or without external beam RT and HT of intermediate (> 6 to 24; median 12 months) or short (up to 6; median 4 months) duration. A Cox regression multivariable analysis was performed assessing whether intermediate compared to short-course HT was associated with a decreased risk of ACM, adjusting for age, year and type of RT, treatment propensity score, and known PC prognostic factors. Results: After a median follow up of 4.3 years a total of 64 (11.6%) men died. Intermediate compared to short-course HT was associated with a significantly decreased risk of ACM (adjusted hazard ratio 0.44, 95% confidence interval 0.20 - 0.94, P = 0.03). Other significant covariates are shown in the table. The 5-year estimates of ACM for intermediate versus short-course HT were 7.0% and 15.7%, respectively. Conclusions: In men with HRPC treated with RT, a median HT duration of 12 months was associated with a significantly decreased risk of ACM when compared to a median HT duration of 4 months. This raises the hypothesis that HT durations shorter than 28-36 months may be sufficient to decrease mortality in men with HRPC. The ongoing RADAR trial by the Trans Tasman Radiation Oncology Group comparing 18 to 6 months of HT may provide level I evidence to validate this hypothesis. [Table: see text]