Testosterone and docetaxel treatment effect on mortality risk in nonmetastatic high‐risk prostate cancer: A predictive biomarker analysis

BACKGROUND: Prostate cancer (PC) arising in patients with low testosterone is often more aggressive and less responsive to androgen deprivation therapy (ADT). This study evaluated whether baseline testosterone group (low vs. normal) modifies the mortality benefit of adding docetaxel to radiation therapy (RT) and ADT in nonmetastatic high-risk PC. METHODS: From two randomized trials of docetaxel plus RT and ADT versus RT and ADT for T1c-4N0M0 PC, 255 patients with minimal or no comorbidity comprised the discovery cohort (median age, 65 years; follow-up, 10.4 years) and 563 patients with Eastern Cooperative Oncology Group performance status 0/1 formed the validation cohort (median age, 66 years; follow-up, 10.6 years). Multivariable Cox regression with an interaction between treatment arm and testosterone group (normal or low level) was used to evaluate all-cause mortality (ACM) risk by adjusting for age, T category, Gleason score, prostate-specific antigen (PSA) level, percent positive biopsy cores, and performance status. RESULTS: Randomization to RT, ADT, and docetaxel was associated with a significant reduction in ACM risk among patients with normal but not low testosterone, with a significant treatment-by-testosterone group interaction in both the discovery (p = .048) and validation cohorts (p = .042). Among patients with normal testosterone, those most likely to benefit from docetaxel had a PSA level of >20 ng/mL, T3/4 disease, or a Gleason score of 9/10 in both cohorts. CONCLUSIONS: Low testosterone at diagnosis does not appear to predict a reduced mortality risk when adding docetaxel to RT and ADT in otherwise healthy patients with nonmetastatic high-risk PC, whereas normal testosterone does, which provides evidence to support testosterone group as a predictive biomarker.