Purpose: Early PSA response has been found to be prognostic of outcomes in metastatic hormone sensitive prostate cancer. We performed a secondary analysis of the TITAN trial to determine if early PSA response was predictive of treatment efficacy in metastatic hormone sensitive prostate cancer patients.
Materials and methods: Early PSA response was defined as achieving a PSA level of ≤ 0.2 ng/mL by 6 months of random assignment. A Cox proportional hazard model was constructed in a landmark population with an interaction term between the treatment and early PSA response to determine differential treatment effect on overall survival (OS). We applied multivariable Cox proportional hazard regression model with time to early PSA response fitted with restricted cubic spline to determine the association of time to early PSA response with OS.
Results: Approximately 24% (124/524) of patients in the androgen deprivation therapy (ADT) alone group and 61% (321/524) in the apalutamide group had PSA response ≤ 0.2 ng/mL by 6 months. Longer time to early PSA response was associated with significantly superior OS in the apalutamide group. There was a significant difference in treatment effect from apalutamide on OS (P = .03 for interaction) among 6-month PSA responders (HR: 0.66; 95% CI: 0.44-1.00) vs nonresponders (HR: 1.14; 95% CI: 0.89-1.46). This difference in treatment effect was not statistically significant at 3 months (P = .17 for interaction). Among 6-month PSA responders, 3-year confounder-adjusted OS was 84% (80%-88%) for the apalutamide group and 74% (66%-82%) for the ADT alone group. Among nonresponders, 3-year adjusted OS for the 2 treatment arms were 58% (52%-65%) and 56% (51%-60%), respectively.
Conclusions: Early PSA response by 6 months was a predictor of treatment efficacy from ADT plus apalutamide on OS. Longer time to early PSA response was associated with superior OS in the apalutamide arm.
Keywords: androgen receptor signaling inhibitors; early PSA response; metastatic hormone sensitive prostate cancer.