Key points are not available for this paper at this time.
Background One of the primary reasons for tumor invasion and metastasis is anoikis resistance. Biochemical recurrence (BCR) of prostate cancer (PCa) serves as a harbinger of its distant metastasis. However, the role of anoikis in PCa biochemical recurrence has not been fully elucidated. Methods Differential expression analysis was used to identify anoikis-related genes based on the TCGA and GeneCards databases. Prognostic models were constructed utilizing LASSO regression, univariate and multivariate Cox regression analyses. Moreover, Gene Expression Omnibus datasets (GSE70770 and GSE46602) were applied as validation cohorts. Gene Ontology, KEGG and GSVA were utilized to explore biological pathways and molecular mechanisms. Further, immune profiles were assessed using CIBERSORT, ssGSEA, and TIDE, while anti-cancer drugs sensitivity was analyzed by GDSC database. In addition, gene expressions in the model were examined using online databases (Human Protein Atlas and Tumor Immune Single-Cell Hub). Results 113 differentially expressed anoikis-related genes were found. Four genes (EEF1A2, RET, FOSL1, PCA3) were selected for constructing a prognostic model. Using the findings from the Cox regression analysis, we grouped patients into groups of high and low risk. The high-risk group exhibited a poorer prognosis, with a maximum AUC of 0.897. Moreover, larger percentage of immune infiltration of memory B cells, CD8 Tcells, neutrophils, and M1 macrophages were observed in the high-risk group than those in the low-risk group, whereas the percentage of activated mast cells and dendritic cells in the high-risk group were lower. An increased TIDE score was founded in the high-risk group, suggesting reduced effectiveness of ICI therapy. Additionally, the IC50 results for chemotherapy drugs indicated that the low-risk group was more sensitive to most of the drugs. Finally, the genes EEF1A2, RET, and FOSL1 were expressed in PCa cases based on HPA website. The TISCH database suggested that these four ARGs might contribute to the tumor microenvironment of PCa. Conclusion We created a risk model utilizing four ARGs that effectively predicts the risk of BCR in PCa patients. This study lays the groundwork for risk stratification and predicting survival outcomes in PCa patients with BCR.
Kang et al. (Tue,) studied this question.