To determine prognostic factors to select high-risk men receiving dose-escalated radiation therapy (RT) who will have favorable outcomes with short-term (ST) or no androgen deprivation therapy (ADT).

Medical records of 458 men treated with definitive RT for high-risk, nonmetastatic prostate cancer at 3 academic referral centers from 1988 to 2009 were examined. Median dose was 76.4 Gy. Men received no ADT (n=105), STADT (<12 mo, n=194), or long-term ADT (LTADT: ≥12 mo, n=160). Univariate and multivariable analysis for freedom from distant metastases (FFDM) and cause-specific survival (CSS) were performed. Median follow-up was 71 months.

Seven-year FFDM was 83% and CSS was 91%. Multivariable analysis demonstrated that prostate-specific antigen (PSA) nadir ≤0.2 (HR=0.36; 95% CI, 0.20-0.64) and Gleason score (GS) were associated with FFDM and CSS (all P<0.05). ADT duration was not associated (P>0.05). Those with PSA nadir ≤0.2 ng/mL had improved outcomes. Men with GS 9 disease did poorly despite a PSA nadir ≤0.2 ng/mL and had improved CSS with LTADT (95% vs. 71%, P<0.05).

Select men with high-risk disease treated with dose-escalated RT may not require LTADT. In men treated with ADT, PSA nadir ≤0.2 is an independent prognostic factor associated with FFDM and CSS. Men without GS 9 may have acceptable outcomes with STADT if PSA nadir is ≤0.2 ng/mL. Further investigation is necessary to elucidate the role of PSA nadir in determining the optimal length of adjuvant ADT.