Anoikis-related gene signature as novel prognostic biomarker to predict immunotherapy with bladder urothelial carcinoma

Abstract Objectives Anoikis plays an active role in the metastasis and progression of many tumors and is emerging as a new target for tumor therapy. We aimed to develop an anoikis-related risk model to assess the prognosis of patients with bladder urothelial carcinoma (BLCA) and to explore its potential application value in immunotherapy. Methods Patient expression data and clinical data were obtained from GEO and TCGA database. Lasso regression was used to obtain a risk model and the clinical efficacy of risk model was evaluated with Cox regression, calibration curves, nomogram diagram, and receiver operating characteristics (ROC). Next, GSEA analysis was performed to estimate potential biological pathways for ARGS. The tumor microenvironment (TME) was also assessed, including cancer-associated fibroblast (CAF), CIBERSORT, XCELL, tumor immune exclusion, and tumor-associated macrophage (TAM). Then, ggpubr and ggplot2 packages were utilized to compare immune checkpoint expression discrepancies in different risk groups. Then, we also discussed the survival relevance of ARGS combined with immune checkpoints using survival and survminer packages and evaluated the sensitivity of immunotherapy for ARGS through the cancer immunome atlas (TCIA) and IMvigor210 cohort. Results 15 anoikis genes were identified to construct prognostic ARGS. ARGS can effectively divide BLCA cases into 2 groups with different clinical outcomes and reflect different TME. It was obvious that patients in the high-risk group could not benefit from immunotherapy. Conclusions ARGS can be used to stratify hazards and predict prognosis events in patients with BLCA and give remarkable guidance for personalized and precise immunotherapy.