Genetic profiling of hormone-sensitive and castration-resistant prostate cancers and identification of genetic mutations prone to castration-resistant prostate cancer

Background Genetic profiling of patients with prostate cancer could potentially identify mutations prone to castration-resistant prostate cancer (CRPC). Here, we aimed to identify the differences in genetic profiles of patients with hormone-sensitive prostate cancer (HSPC) and CRPC and stratify HSPC patients to identify mutations associated with CRPC progression. Methods A total of 103 samples were collected, including 62 DNA samples from the tumor tissues of 59 HSPC patients and 41 cell-free DNA (cfDNA) samples from prostate cancer patients at different cancer stages. Targeted sequence was conducted on both the tissue DNA and cfDNA. The associations between mutations and clinical outcomes (CRPC-free time) were analyzed using χ 2 test, logistic regression analysis, Kaplan–Meier analysis, and Cox regression analysis. Results By comparing to that of cfDNA sequencing, the results from DNA sequencing of 1-needle (80%) and mixed 12-needle (77.8%) biopsies are highly comparable. FOXA1 (30.5%), CDK12 (23.7%), and TP53 (22.0%) were the top 3 most frequently mutated genes in HSPC patients; 50.8% (30/59) and 44.1% (26/59) HSPC patients had mutations in DDR and HRR pathway, respectively. Mutations in AR and APC as well as the members involved in the regulation of stem cell pluripotency and EMT pathway were often observed in CRPC samples. We established a panel of four genetic mutations (MSH2, CDK12, TP53, and RB1) to predict the risk of CRPC early progression with concordance index = 0.609 and the area under curve of the ROC curve as 0.838. Conclusions In this study, we demonstrated that the cfDNA can be used in genetic profiling in prostate cancer and our newly established panel is capable of predicting which mHSPC patient has a high risk of early CRPC progression.