PURPOSE Androgen receptor (AR) gene alterations, including ligand-binding domain mutations and copy number (CN) gain, have yet to be fully established as predictive markers of resistance to enzalutamide and abiraterone in men with metastatic castration-resistant prostate cancer (mCRPC). The goal of this study was to validate AR gene alterations detected in cell-free DNA (cfDNA) as markers of enzalutamide and abiraterone resistance in patients with mCRPC. METHODS Patients with mCRPC (N = 62) were prospectively enrolled between 2014 and 2018. Blood was collected before therapies-enzalutamide (n = 25), abiraterone (n = 35), or enzalutamide and abiraterone (n = 2)-and at disease progression. We used deep next-generation sequencing to analyze cfDNA for sequence variants and CN status in AR and 45 additional cancer-associated genes. Primary end points were prostatespecific antigen response, progression-free survival (PFS), and overall survival (OS). RESULTS Elevated tumor-specific cfDNA (circulating tumor DNA) was associated with a worse prostate-specific antigen response (hazard ratio [HR], 3.17; 95% CI, 1.11 to 9.05; P = .031), PFS (HR, 1.76; 95% CI, 1.03 to 3.01; P = .039), and OS (HR, 2.92; 95% CI, 1.40 to 6.11; P = .004). AR ligand-binding domain missense mutations (HR, 2.51; 95% CI, 1.15 to 5.72; P = .020) were associated with a shorter PFS in multivariable models. AR CN gain was associated with a shorter PFS; however, significance was lost in multivariable modeling. Genetic alterations in tumor protein p53 (HR, 2.70; 95% CI, 1.27 to 5.72; P = .009) and phosphoinositide 3-kinase pathway defects (HR, 2.62; 95% CI, 1.12 to 6.10; P = .026) were associated with a worse OS in multivariable models. CONCLUSION These findings support the conclusion that high circulating tumor DNA burden is associated with worse outcomes to enzalutamide and abiraterone in men with mCRPC. Tumor protein p53 loss and phosphoinositide 3-kinase pathway defects were associated with worse OS in men with mCRPC. AR status associations with outcomes were not robust, and additional validation is needed.
Bibliographical notePublisher Copyright:
© 2019 American Society of Clinical Oncology.