Objective: Prostate cancer (PCa) is the second disease threatening men's health, and anti-androgen therapy (AAT) is a primary approach for treating this condition. Increasing evidence suggests that long non-coding RNAs (lncRNAs) play crucial roles in the development of PCa and the process of AAT resistance. The objective of this study is to utilize bioinformatics methods to excavate lncRNAs association with AAT resistance and investigate their biological functions.
Methods: AAT resistance-related risk score model (ARR-RSM) was established by multivariate Cox analysis. Paired clinical tissue samples of 36 PCa patients and 42 blood samples from patients with PSA over 4 ng/ml were collected to verify the ARR-RSM. In vitro, RT-qPCR, CCK-8 and clone formation assays were displayed to verify the expression and function of AL354989.1 and AC007405.2.
Results: Pearson correlation analysis identified 996 lncRNAs were associated with AAT resistance (ARR-LncRs). ARR-RSM was established using multivariate Cox regression analysis, and PCa patients were divided into high-risk and low-risk groups. High-risk patients showed increased expression of AL354989.1 and AC007405.2 had poorer prognoses. The high-risk score correlated with advanced T-stage and N-stage. The AUC of ARR-RSM outperformed tPSA in diagnosing PCa. Silencing of AC007405.2 and AL354989.1 inhibited PCa cells proliferation and AAT resistance.
Conclusions: In this study, we have discovered the clinical significance of AC007405.2 and AL354989.1 in predicting the prognosis and diagnosing PCa patients. Furthermore, we have confirmed their correlation with various clinical features. These findings provide potential targets for PCa treatment and a novel diagnostic and predictive indicator for precise PCa diagnosis.
Keywords: AAT resistance; lncRNA; prostate cancer; prostate cancer diagnosis; risk score model.