Dynamics of conditional survival and risk factors in androgen deprivation therapy for prostate cancer using a multi-institutional Japan-wide database.

OBJECTIVES: The objectives of this study were to analyze the conditional survival and prognostic factors in androgen deprivation therapy for prostate cancer using the Japan Study Group of Prostate Cancer database. METHODS: Data on patients treated with primary androgen deprivation therapy between 2001 and 2003 from a nationwide database of the Japan Study Group of Prostate Cancer were used. The conditional 5-year progression-free rate, cancer-specific survival and overall survival, as well as the conditional mortality owing to prostate cancer and other causes were calculated as per subgroups. Prognostic factors for progression-free rate, cancer-specific survival and overall survival at each time after androgen deprivation therapy initiation were calculated using the Cox proportional hazards model. RESULTS: The conditional 5-year progression-free rate and cancer-specific survival, but not overall survival, gradually increased with time. The prognostic impact of stage IV characteristics (T4, N1 and M1) changed over time; however, the prognostic impact of the Gleason score remained unchanged. In the subgroup analysis, prostate-specific mortality risk reduced over time in patients with stage IV prostate cancer, whereas non-prostate cancer mortality increased over time in elderly patients. CONCLUSIONS: Information regarding conditional survival and mortality obtained in this study would provide a benchmark for physicians and cancer survivors.