Collagen gene signature in the tumor microenvironment predicts survival and guides prognosis in bladder cancer

Collagen, within the tumor microenvironment (TME), assumes a crucial function in the development of cancer. However, the expression of collagen genes in bladder cancer (BCa) remains inadequately comprehended. The aim of this study is to examine the collagen genes expression in the BCa TME, and construct a nomogram to predict the overall survival for patients diagnosed with BCa. The Cancer Genome Atlas (TCGA) database (N = 401) and the Gene Expression Omnibus (GEO) database (N = 165) were employed for training and validation cohorts. The correlation between the collagen genes assay and overall survival was investigated using Cox regression analysis. Model construction employed the least absolute shrinkage and selection operator (LASSO) Cox regression algorithm. The model's performance was thoroughly assessed, including its discrimination, calibration, and clinical utility. The nomogram, incorporating the P3H4, C1QTNF6, COLGALT1, COL4A1, COL14A1, RGCC, PPARG, SCX and age by utilizing least absolute shrinkage and selection operator Cox regression algorithm, exhibits the favorable predictive capability in the area under the receiver operator characteristic curve, the calibration curve and decision curve analysis. Then, we investigated that C1QTNF6, COL4A1, COL14A1, RGCC and P3H4 were significantly associated with lymph node-positive BCa patients. Additionally, the correlation of collagen genes with tumor mutation burden and immune characteristic was elucidated. We developed a favorable prognostic model using collagen genes, which are potential biomarkers for forecasting the prognosis of BCa.