Taxane chemotherapy versus anti-androgen agents as first-line therapy for metastatic castration-resistant prostate cancer.

Objectives To assess treatment patterns and outcomes of patients with metastatic castration-resistant prostate cancer (mCRPC) receiving first-line chemotherapy or antiandrogen therapy. Patients and Methods Patients initiating first-line antiandrogen therapy (abiraterone, enzalutamide) or chemotherapy (taxane) between October 2012 and September 2014 were retrospectively identified in the US Veterans Health Administration database. The impact of antiandrogen therapy vs chemotherapy on overall survival (OS) and time to discontinuation was assessed using Cox proportional hazard models, adjusting for prior androgen deprivation therapy (ADT) duration and available prognostic factors. Results Overall, 1445 patients were evaluable, of whom 1108 received antiandrogen therapy and 337 received chemotherapy (docetaxel). On multivariable analysis and propensity score analysis, the OS times for antiandrogen therapy vs chemotherapy were not significantly different (hazard ratio [HR] 1.041, 95% confidence interval (CI) 0.853-1.270, P = 0.694, and HR 1.047, 95% CI 0.861-1.273, P = 0.644, respectively). Time to discontinuation was shorter for chemotherapy vs antiandrogen therapy (HR 2.339, 95% CI 1.969-2.779; P < 0.001). Prior ADT duration above the median was associated with longer OS (HR 0.566, 95% CI 0.464-0.690; P < 0.001) and time to discontinuation (HR 0.831, 95% CI 0.699-0.988; P = 0.036) in the antiandrogen therapy cohort and not the chemotherapy cohort, while prior ADT duration below the median was associated with higher prostate specific antigen (PSA) response rate in the chemotherapy vs antiandrogen therapy cohort (61.5% vs 51.1%; P = 0.024). The treatment-free interval after discontinuation was longer after first-line chemotherapy vs antiandrogen therapy (mean 53 vs 39 days; P = 0.030). Conclusion After adjusting for key prognostic factors in this large mCRPC dataset, the OS was similar for first-line chemotherapy vs antiandrogen therapy despite shorter time to discontinuation with chemotherapy and longer treatment-free interval after first-line chemotherapy. These hypothesis-generating data also suggest that duration of prior ADT may assist in the selection of patients for chemotherapy vs antiandrogen therapy.