Identification and verification of anoikis-related gene markers to predict the prognosis of patients with bladder cancer and assist in the diagnosis and treatment of bladder cancer.

BACKGROUND

The recurrence and mortality rates of bladder cancer are extremely high, and its diagnosis and treatment are global concerns. The mechanism of anoikis is closely related to tumor metastasis.

METHODS

First, we obtained all the data needed for this study from a public database through a formal operational process. The data were then analyzed by bioinformatics technology. Through the limma package, we screened and obtained 313 anoikis-related genes [false discovery rate (FDR) <0.05, |log fold change (FC) | >0.585]. Then, through univariate independent prognostic analysis, we further screened 146 genes (P<0.05) related to the prognosis of bladder cancer from 313 differential genes. These 146 prognostically relevant differential genes were used for least absolute shrinkage and selection operator (LASSO) regression for further screening to obtain model-related genes and output model formulas. Through the nomogram, we can calculate the survival rate of patients more accurately. The accuracy of the nomogram was also confirmed by calibration curves, independent prognostic analysis, receiver operating characteristic (ROC) curves, decision curve analysis (DCA) curves. We then analysed the sensitivity of immunotherapy in bladder cancer patients with different risk scores via Tumor Immune Dysfunction and Exclusion (TIDE).

RESULTS

Through bioinformatics technology and public databases, a prognostic model including 9 anoikis-related genes (, , , , , , , , ) was obtained. Integrating risk scores with clinical information, we obtained a nomogram that can accurately predict patient survival. By querying the immunohistochemical results of the Human Protein Atlas database, two of the nine model-related genes (, ) have the value of further research and are expected to become new biomarkers to assist the diagnosis and treatment of bladder cancer. Through immune-related analysis, we found that patients in the low-risk group appeared to be more suitable for immunotherapy, while drug sensitivity analysis showed that bladder cancer patients in the high-risk group were more sensitive to common chemotherapy drugs.

CONCLUSIONS

In this study, a prognostic model that can accurately predict the prognosis of patients with bladder cancer was constructed. and are expected to become a new biomarker for the diagnosis and treatment of bladder cancer.