A Double-Blinded Placebo-Controlled Biomarker Stratified Randomized Trial of Apalutamide (APA) and Radiotherapy for Recurrent Prostate Cancer (NRG GU006, BALANCE trial)

Abstract

Purpose/Objective(s)

Currently there are no prospectively validated predictive biomarkers to guide use of hormonal therapy in prostate cancer. NRG GU006, a phase II biomarker stratified randomized trial of patients receiving salvage radiotherapy (SRT) with or without APA, was designed with the hypothesis that transcriptionally defined molecular subtypes would differentially benefit from APA.

Materials/Methods

Patients were enrolled between 4/2018-2/2020 and were required to be status post-radical prostatectomy with a PSA 0.1-1.0 ng/mL without evidence of nodal or distant metastasis and randomized to SRT with placebo or APA 240 mg daily for 6 months. Patients were stratified by PAM50 molecular subtype (luminal B vs non-luminal-B). The primary endpoint was biochemical progression-free survival (bPFS), defined as first occurrence of biochemical, local, regional, distant recurrence, or death from any cause. Key secondary endpoints reported are metastasis-free survival (MFS) and adverse events. The design and analysis involved first testing efficacy within the luminal B subtype (hypothesized to show greater benefit), followed by evaluation of the non-luminal-B group. If the lower limit of the 80% confidence interval (CI) for the hazard ratio (HR [APA/placebo]) in the latter group was >0.77, lack of efficacy in this subgroup would be declared.

Results

A total of 295 eligible patients were enrolled with a median follow-up of 5.0 years. Arms were well balanced, with a median age of 65 years, 50% with positive surgical margins, 51% with pathologic T3 disease, 86% with entry PSA of <0.5 ng/mL; 19% were grade group 4-5, and 43% were luminal B. In luminal B patients APA significantly improved bPFS (HR 0.45, 80%CI 0.29-0.68, one-sided p=0.0062), with 5-year estimated bPFS of 72.4% vs 53.9% in the APA and placebo arms, respectively. In contrast, non-luminal B patients did not demonstrate improvement in bPFS (HR 0.95, 80%CI 0.65-1.41, p=0.44), with 5-year estimated bPFS of 70.2% vs 71.1%, although the lower CI limit was <0.77. MFS was also improved with APA in luminal B patients, HR 0.27, 95%CI 0.07-0.95, p=0.029; 5-year estimates of 94.7% vs 81.8%, but not in non-luminal B patients, HR 1.06, 95%CI 0.41-2.78, p=0.90; 5-year estimates of 89.9% vs 89.3%. In the APA vs placebo arm (regardless of attribution), grade 3+ gastrointestinal toxicity occurred in 5.7% vs 2.6%, and genitourinary toxicity in 3.5% vs 4.5%, respectively. In the APA arm, grade 3+ rash occurred in 5.0% of patients and breast pain in 0.7%.

Conclusion

Patients with transcriptionally defined luminal B tumors derived improvement in clinically meaningful endpoints from the addition of APA to SRT. PAM50 represents the first prospectively validated predictive biomarker for hormone therapy in prostate cancer in a randomized trial.