Objective: Prostate-specific antigen (PSA) decline by 50 % from the baseline to 12 weeks (PSA50w12) is currently used to predict response to treatment and clinical outcome of patients with metastatic castration-resistant prostate cancer (mCRPC). We evaluated the association between PSA changes at 4 weeks and clinical outcome.
Patients and methods: Eligible patients had PSA levels assessed at baseline, and monthly during enzalutamide treatment. Early PSA increase was defined as an increased PSA level at 4 weeks ≥20 % (PSA + 20w4) from baseline. Early PSA decline was defined as a PSA response at 4 weeks ≥30 % (PSA30w4) and ≥50 % (PSA50w4) from baseline. Progression-free survival (PFS), overall survival (OS) and their 95 % confidence intervals (CI) were evaluated by the Kaplan-Meier method and compared with the log-rank test. The impact of early PSA increase and decline on PFS and OS was evaluated by Cox regression analyses.
Results: We assessed 193 patients with median age of 73 years (range 43-91 years). The median follow-up was 11.7 months (range 0.5-27.4 months). PSA + 20w4 predicted both PFS and OS [HR 6.50 (95 % CI 2.63-16.07; p < 0.0001) and HR 10.54 (95 % CI 4.02-27.64; p < 0.0001), respectively], whereas PSA30w4 and PSA50w4 predicted only PFS [HR 0.37 (95 % CI 0.21-0.67; p = 0.0009) and HR 0.34 (95 % CI 0.19-0.60; p = 0.0003), respectively].
Conclusions: An early PSA increase, defined as a PSA level at 4 weeks ≥20 % (PSA + 20w4), could be useful to quickly identify patients unlikely to benefit from enzalutamide. Larger studies are needed to confirm PSA + 20W4 as an early biomarker of primary resistance to enzalutamide.