Lower Testosterone Level and Metastases-Free Survival in Patients With Nonmetastatic Castration-Resistant Prostate Cancer Treated With Novel Antiandrogens: A Post Hoc Analysis of SPARTAN and ARAMIS.

Purpose: The European Association of Urology guidelines and the latest recommendations from the US Prostate Cancer Conference suggest a castration threshold of 20 ng/dL. However, the current National Comprehensive Cancer Network and AUA guidelines still recommend a castration standard of 50 ng/dL. It remains unknown whether there is a relationship between maintaining lower testosterone levels and the prognosis of patients with nonmetastatic castration-resistant prostate cancer (nmCRPC).

Materials and methods: We conducted a retrospective analysis based on 2 phase 3 clinical trials (SPARTAN-NCT01946204, ARAMIS-NCT02200614). Patients receiving the currently recommended first-line treatment regimens (androgen deprivation therapy + apalutamide in SPARTAN and androgen deprivation therapy + darolutamide in ARAMIS) were classified into 2 groups according to their maintenance levels of serum testosterone during treatment: below 20 ng/dL and above 20 ng/dL. The study end point was metastasis-free survival (MFS). Inverse probability of treatment weighting method was used to balance patients' baseline characteristics. Kaplan-Meier analysis, multivariable Cox regression models, and Cox models that included testosterone levels as a time-dependent covariate were used to investigate the influence of maintenance levels of serum testosterone on MFS.

Results: Baseline characteristics were well balanced between the low testosterone group and the high testosterone group after applying inverse probability of treatment weighting weights. Kaplan-Meier analysis indicated that there was no statistically significant association between serum testosterone levels and MFS in either trial. In both multivariable Cox regression models and time-dependent Cox regression models, the relationship between serum testosterone levels and MFS did not show statistical significance either, using below 20 ng/dL as the reference group (multivariable Cox: SPARTAN HR, 0.68 [95% CI, 0.47-0.98; P < .05], ARAMIS HR, 0.83 [95% CI, 0.57-1.20; P = .320]; time-dependent Cox: SPARTAN HR, 0.84 [95% CI, 0.68-1.04; P = .110], ARAMIS HR, 1.21 [95% CI, 0.71-2.08; P = .480]). The results obtained by setting testosterone levels as a continuous variable were similar.

Conclusions: Among all men with testosterone < 50 ng/dL, maintenance serum testosterone levels ≥ 20 ng/dL were not associated with poorer MFS in the first-line therapy of nmCRPC with novel hormonal agents. The prognostic value of maintaining testosterone levels < 20 ng/dL in patients with nmCRPC is limited, and further treatment intensification is not indicated.

Trial registration: ClinicalTrials.gov Identifier: NCT01946204, NCT02200614.