Comprehensive analysis of tumor immune-related gene signature for predicting prognosis, immunotherapy, and drug sensitivity in bladder urothelial carcinoma

BACKGROUND
Bladder urothelial carcinoma is a prevalent cancer recognized worldwide. Its treatment poses ongoing challenges due to significant illness, high death rates, postoperative complications affecting quality of life, and a lack of specific molecular targets for therapy. The aim of this study was to develop a predictive model to assess the prognostic significance of genes, evaluate responses to immunotherapy, and determine susceptibility to various drugs in patients with bladder urothelial carcinoma.

METHODS
Clinical details and gene expression data related to immune function were obtained from The Cancer Genome Atlas datasets for patients with bladder urothelial carcinoma. These data were analyzed using the R programming language and related software packages. The analysis included identifying genes with different expression levels, examining their biological roles and associated pathways, constructing gene co-expression networks, performing statistical regression analyses to identify prognostic factors, developing and evaluating a prognostic model, conducting gene set variation analysis to understand biological processes, analyzing immune function and checkpoint molecules, predicting responses to immunotherapy, and forecasting drug sensitivity.

RESULTS
A set of 11 immune-related genes with different expression levels was selected to create a gene signature associated with immune response in bladder carcinoma for predicting patient outcomes. In both the initial group of patients used to train the model and a separate group used to test it, patients classified as high-risk by the gene signature had lower overall survival rates compared to those classified as low-risk. The model’s accuracy in predicting survival, as measured by the area under the receiver operating characteristic curve, was 0.712 in the training group and 0.631 in the testing group, indicating its predictive ability. This finding was further validated in two independent datasets, GSE39281 and IMvigor210, where the high-risk groups also showed significantly lower overall survival, with accuracy values of 0.609 and 0.563, respectively. Patients in the training group were divided into low- and high-risk categories based on the median risk score calculated from the bladder carcinoma immune gene signature. Gene set variation analysis revealed 21 biological pathways that were positively correlated with the risk scores generated by the model. The high-risk group exhibited higher levels of stromal cell presence, immune cell infiltration, an overall estimate of tumor purity, and a score indicating exclusion of T cells from the tumor microenvironment. Conversely, the low-risk group showed increased expression of cytotoxic T-lymphocyte antigen 4, suggesting a potentially better response to immune checkpoint inhibitors. Notably, significant differences were observed in immune subtypes, immune-related functions, and survival outcomes related to immune response between the two risk groups defined by the model. The accuracy of our model in predicting survival was 0.765 and 0.660 in two comparisons, respectively, which was better than other existing models, such as the tumor inflammation signature, tumor immune dysfunction and exclusion, and various clinical factors. A visual representation of the predictive factors, called a nomogram, was also presented, showing good accuracy in predicting 1-, 2-, and 3-year survival, with values of 0.727, 0.772, and 0.765, respectively, indicating the signature’s strong predictive power. Finally, 20 small molecule drugs were identified as potentially effective, with the drug TW.37 showing the most significant difference in the concentration required to inhibit 50% of cell growth between the high- and low-risk patient groups, suggesting its potential as a treatment option.

CONCLUSIONS
The immune-related gene signature model developed in this study can predict the prognosis of patients with bladder urothelial TW-37 carcinoma and has the potential to guide personalized treatment decisions regarding immunotherapy and chemotherapy drugs.