Objective: To determine optimal predictors with which to select the crucial patients enrolled in delayed-combined androgen blockade (CAB) trials, based on risk factors.
Methods: From January 2001 to December 2004, 92 prostate cancer patients with T1c, T2 and T3aN0M0 were enrolled in a clinical trial. Medical castration and anti-androgen treatment were used sequentially as delayed-CAB. The prostate specific antigen (PSA) nadir was determined following medical castration only. Anti-androgen treatment was administered if a PSA progression was observed and the subsequent PSA response was evaluated. Time to PSA biochemical failure, induced by medical castration or with anti-androgen treatment, was estimated. Risk factors of PSA failure were evaluated by multivariate analysis.
Results: During luteinizing hormone-releasing hormone (LH-RH) monotherapy, a Kaplan-Meier analysis estimated that the proportion of patients without PSA progression was 64.8% at 5 years. In the multivariate analysis of the prediction of PSA progression with LH-RH monotherapy, a Gleason score over 8, initial PSA >20 ng/ml and PSA nadir >0.2 ng/ml were significant independent risk factors that affected PSA biochemical failure. The PSA progression-free rate in the lower PSA nadir group was significantly lower than that in the other. The 25 patients in the higher PSA nadir group were treated with anti-androgen therapy. Under anti-androgen therapy, the PSA progression-free rate was 62.6% at 5 years. Only PSA nadir >0.2 ng/ml was a significant independent risk factor. The PSA progression-free rate in the lower PSA nadir group was significantly lower than the other.
Conclusions: PSA nadir was the optimal predictive for low stage, non-metastatic population during delayed-CAB.