Biochemical Failure Is Not a Surrogate End Point for Overall Survival in Recurrent Prostate Cancer: Analysis of NRG Oncology/RTOG 9601

Abstract

Purpose: Metastasis-free survival (MFS), but not event-free survival, is a validated surrogate end point for overall survival (OS) in men treated for localized prostate cancer. It remains unknown if this holds true in biochemically recurrent disease after radical prostatectomy. Leveraging NRG/RTOG 9601, we aimed to determine the performance of intermediate clinical end points (ICEs) as surrogate end points for OS in recurrent prostate cancer.

Materials and methods: NRG/RTOG 9601 randomly assigned 760 men with recurrence after prostatectomy to salvage radiation therapy with 2 years of placebo versus bicalutamide 150 mg daily. ICEs assessed were biochemical failure (BF) per NRG/RTOG 9601 (prostate-specific antigen nadir + 0.3-0.5 ng/mL or initiation of salvage hormone therapy; [BF1]) and NRG/RTOG 0534 (prostate-specific antigen nadir+2 ng/mL; [BF2]), distant metastasis (DM), and MFS (DM or death). Surrogacy was assessed by the Prentice criteria and a two-stage meta-analytic approach (condition one assessed at the patient level with Kendall's τ and condition two assessed by randomly dividing the entire trial cohort into 10 pseudo trial centers and calculating the average R² between treatment hazard ratios for ICE and OS).

Results: BF1, BF2, DM, and MFS satisfied the four Prentice criteria. However, with the two-condition meta-analytic approach, there was strong correlation between MFS and OS (τ = 0.86), moderate correlation between DM and OS (τ = 0.66), and weaker correlation between BF1 (τ = 0.25) or BF2 (τ = 0.40) and OS. Similarly, for condition two, the treatment effect of antiandrogen therapy on MFS and OS were correlated (R² = 0.67), but this was not true for BF1 (R² = 0.09), BF2 (R² = 0.12), or DM (R² = 0.18) and OS.

Conclusion: MFS is also a strong surrogate for OS in men receiving salvage radiation therapy for recurrence after prostatectomy. Caution should be used when inferring survival benefit from effects on BF in biochemically recurrent prostate cancer. Lack of comorbidity data did not allow us to assess whether BF in men with no/minimal comorbidity could serve as a surrogate for OS.

Trial registration: ClinicalTrials.gov NCT00002874.

FIG 1.

OS 5-year landmark analysis of intermediate clinical end points. A 5-year landmark analysis was performed, where patients dying or lost to follow-up before 5 years after random assignment were excluded. Time 0 on the x-axis represents 5 years after random assignment. (A) BF per NRG/RTOG 9601, (B) PSA nadir plus 2 ng/mL per NRG/RTOG 0534, and (C) DM/metastasis-free survival. BF, biochemical failure; DM, distant metastasis; OS, overall survival; PSA, prostate-specific antigen.

FIG 2.

Correlation of the treatment effect of antiandrogen therapy (hazard ratio) on OS versus the treatment effect of antiandrogen therapy on intermediate clinical end points. Each circle represents one of the 10 pseudo trial centers with the size of each circle representative of the number of patients randomly assigned to each pseudo trial center. (A) Biochemical failure per NRG/RTOG 9601, (B) PSA nadir plus 2 ng/mL per NRG/RTOG 0534, (C) DM, and (D) MFS. BF, NRG/RTOG 9601 definition of biochemical failure; DM, distant metastasis; MFS, metastasis-free survival; OS, overall survival; PSA, prostate-specific antigen.

FIG A1.

OS 5-year landmark analysis of the American Urological Association Definition of Biochemical Failure. A 5-year landmark analysis was performed, where patients dying or lost to follow-up before 5 years after random assignment were excluded. Time 0 on the x-axis represents 5 years after random assignment. BF, biochemical failure; OS, overall survival.

FIG A2.

Correlation of the treatment effect of antiandrogen therapy (hazard ratio) on OS versus the treatment effect of antiandrogen therapy on the American Urological Association definition of Biochemical Failure. Each circle represents one of the 10 pseudo trial centers with the size of each circle representative of the number of patients randomly assigned to each pseudo trial center. ICE, intermediate clinical end point; OS, overall survival.