Identification and validation of the PARP inhibitor-related gene KANK3 for predicting prognosis and immunotherapeutic response in prostate cancer

<b>Background:</b> Prostate cancer (PCa), a prevalent malignant neoplasm in men, has its biochemical recurrence-free survival (BCRFS) serving as a critical determinant for patient prognosis. PARP inhibitors have demonstrated potential therapeutic value in the management of PCa. Nevertheless, the precise influence exerted by their associated genes on BCRFS remains elusive. <b>Methods:</b> We selected the differentially expressed genes after treatment with olaparib and defined them as PARP inhibitor-related genes (PIRGs). Consensus clustering was employed to evaluate the relationships among different PIRGs clusters, prognosis, and the immune microenvironment. Univariate COX regression analysis was used to screen the prognosis-related PIRGs, which were then incorporated into multiple machine learning frameworks. The random forest algorithm with the highest C-index was chosen to construct a BCRFS prediction model. A prognostic nomogram was developed based on the risk score and clinical information, and the predictive performance of the model was assessed. <b>Results:</b> In C4 - 2B and LNCaP cell lines, 230 and 58 genes were differentially expressed, respectively. Consensus clustering results showed distinct survival prognoses and immune - infiltrated microenvironments among different groups. The random forest model had a high average C - index in both the training and validation sets. The prognostic model constructed in this study demonstrated a higher C-index compared to the prognostic models from previous studies. High - risk group patients had a poor immunotherapy response. A nomogram based on risk scores and clinical information accurately predicted PCa patients' BCRFS. Cell experiments revealed that KANK3 was downregulated in PCa and upregulated by olaparib treatment. KANK3 overexpression in PCa cell lines inhibited cell proliferation, migration, and invasion, suggesting its oncogenic role in PCa. <b>Conclusion:</b> Our study has described the correlations between PARP inhibitor-related genes and the immune landscape, recurrence after radical prostatectomy, as well as clinical characteristics. The risk score can improve the existing risk stratification system.