Analysis of Single-Cell RNA-Seq Data to Investigate Tumor Cell Heterogeneity in Uroepithelial Bladder Cancer and Predict Immunotherapy Response

Background: Numerous studies have suggested a close association between can-cer stem cells (CSCs) and the tumor microenvironment (TME), suggesting that cancer stem-ness might also contribute to ICI resistance. However, the interplay between these physio-logical processes in urothelial bladder cancer (UBC) remains unclear. Method: A meta-analysis was performed using the UBC Single-cell RNA sequencing (scRNA-seq) dataset, and tumor stemness gene sets (Ste.genes) were obtained. The rela-tionship between Ste.genes and ICI response, as well as response to drug therapy, was in-vestigated using Tumour Immune Dysfunction and Exclusion (TIDE) and drug sensitivity analyses. Machine learning based on Ste.genes was also used to predict ICI response. Results: A hypoxia-related tumor subgroup associated with angiogenesis and tumor metas-tasis was identified, and prognostic models were constructed based on hypoxic tumor sub-groups. It was also found that the Ste.genes score was associated with cellular immunity, tumor immunotherapy response, and drug sensitivity. Multiple machine learning models were used to predict ICI response based on Ste.genes, and the AUC was greater than 0.7, indicating that Ste.genes can predict ICI response effectively. Conclusions: In this study, the analysis of UBC scRNA-seq data provided further insight into the role of hypoxic tumor subpopulations in tumor development in UBC, and a prog-nostic model was constructed. Additionally, an association was found between cell stemness and resistance to immunotherapy as well as drug sensitivity in UBC. Ste.genes were ex-tracted and utilized to predict the ICI response.