To evaluate whether the time to return to the baseline prostate-specific antigen (PSA) level (PSA response duration) is significantly associated with the time to death after randomization for men with hormone-refractory metastatic prostate cancer.

The study (n = 213) and validation (n = 281) cohorts constituted 494 men with hormone-refractory metastatic prostate cancer treated in four prospective randomized Phase II studies between February 1996 and October 2001. Cox regression analysis was used to evaluate whether a significant association existed between the PSA response duration and the time to death after randomization, controlling for treatment and known prognostic factors.

A decreasing PSA response duration was significantly associated with a shorter survival after randomization in the study (P = 0.001) and validation (P = 0.02) cohorts, controlling for treatment and known prognostic factors, which included serum PSA, lactate dehydrogenase, alkaline phosphatase, and hemoglobin levels and the Eastern Cooperative Oncology Group performance status. The adjusted hazard ratio for death was 1.9 (95% confidence interval 1.4 to 2.6; P = 0.0002) and 2.1 (95% CI 1.2 to 3.5; P = 0.01) for men in the study and validation cohorts, respectively, whose PSA response duration was shorter than the median value of 3 months. The PSA response duration dichotomized about the median remained significantly associated with death whether patients in the validation cohort experienced at least a 50% reduction (P = 0.05) in the PSA level or not (P = 0.03).

The duration of the PSA response to treatment is significantly associated with length of survival for men with hormone-refractory metastatic prostate cancer.