Integrative analysis of immune molecular subtypes and microenvironment characteristics of bladder cancer. (24th June 2021)

• Record Type: Journal Article
• Authors: Cao, Jinlong; Li, Jianpeng; Yang, Xin; Li, Pan; Yao, Zhiqiang; Han, Dali; Ying, Lijun; Wang, Lijie; Tian, Junqiang
• Is Part Of: Cancer medicine. Volume 10:Number 15(2021)
• Page Start: 5375
• Page End: 5391
• Publication Date: 2021-06-24
• Subjects: decision tree model -- immune molecular subtypes -- immunotherapy -- random forest -- the cancer genome atlas; 616.994005
• DOI: 10.1002/cam4.4071
• Languages: English
• ISSNs: 2045-7634

Abstract

Abstract: The emergence of immunotherapy has provided an option of treatment methods for bladder cancer (BC). However, the beneficiaries of immunotherapy are still limited to small‐scale patients, and immunotherapy‐related adverse events often occur. It is a major challenge for clinical work to study the immune subtypes of BC and the molecular mechanism of immune escape, and identify the immune responders accurately. Here, we explore the immune molecular subtypes of bladder cancer and potential escape mechanisms. First, we screened the expression profiles of 303 differentially expressed immune‐related genes in BC patients from the Cancer Genome Atlas (TCGA) database, and successfully identified 4 molecular subtypes of BC. By comparing the clinical characteristics, immune cells infiltration, the expression of checkpoint genes, human leukocyte antigen (HLA) genes, and gene mutation status of different subtypes, we identified different clinical and immunological characteristics of 4 subtypes. Among 4 subtypes, Cluster 2 met the general characteristics of immunotherapy responders and responded well to immunotherapy, while Cluster 4 had the highest expression of immune characteristics, and is similar to the immune environment of normal bladder tissue. Then, the weighted gene co‐expression network analysis (WGCNA) of immune‐related genes revealed that brown module was positively correlated with subtypes. Pathway enrichment analysis explored the major pathways associated with subtypes, which are also associated with immune escape mechanisms. Moreover, the decision tree model, which was constructed by the principle of random forest screening factors, was also validated in internal validation set and external validation set from the Gene Expression Omnibus (GEO) cohort (GSE133624 ), and could achieve accurate subtypes prediction for BC patients with high‐throughput sequencing. Taken together, we explored the immune molecular subtypes and their mechanisms of BC, and these results may provide guidance for the development of new BC immunotherapy strategies. Abstract : We successfully identified 4 molecular subtypes of BC via TCGA database. Moreover, the establishment of decision tree model, by the principle of random forest screening factors, can achieve accurate subtype prediction for clinical BC patients with high‐throughput sequencing.