Pathogenomics model for personalized medicine in cervical cancer: Cross‐talk of gene expressions and pathological images related to oxidative stress

Jiaqi Li,Siyi Chen,Junsong Wu,Xuefeng Liu,Hejing Liu,Yuedong Liu,Zhuoying Zhu
DOI: https://doi.org/10.1002/tox.23974
IF: 4.109
2023-09-27
Environmental Toxicology
Abstract:An increasing number of studies have shown that oxidative stress plays an important role in the development and progression of cancer. Cervical cancer (CC) is a disease of unique complexity that tends to exhibit high heterogeneity in molecular phenotypes. We aim here to characterize molecular features of cervical cancer by developing a classification system based on oxidative stress‐related gene expression profiles. In this study, we obtained gene expression profiling data for cervical cancer from the TCGA (The Cancer Genome Atlas) and GEO (Gene Expression Omnibus) (GSE44001) databases. Oxidative stress‐related genes used for clustering were obtained from GeneCards. Patients with cervical cancer were divided into two subtypes (C1 and C2) by non‐negative matrix factorization (NMF) classification. By performing Kaplan–Meier survival analysis, differential expression analysis, and gene set enrichment analysis (GSEA) between the two subtypes, we found that subtype C2 had a worse prognosis and was highly enriched for immune‐related pathways as well as epithelial‐mesenchymal transition (EMT) pathways. Subsequently, we performed metabolic pathway analysis, gene mutation landscape analysis, immune microenvironment analysis, immunotherapy response analysis, and drug sensitivity analysis of the two isoforms. The results showed that the isoforms were significantly different between metabolic pathway enrichment and the immune microenvironment, and the chromosomes of subtype C1 were more unstable. In addition, we found that subtype C2 tends to respond to treatment with anti‐CTLA4 agents, a conclusion that coincides with high chromosomal variation in C1, as well as C2 enrichment of immune‐related pathways. Then, we screened 10 agents that were significantly susceptible to C2 subtype. Finally, we constructed pathogenomics models based on pathological features and linked them to molecular subtypes. This study establishes a novel CC classification based on gene expression profiles of oxidative stress‐related genes and elucidates differences between immune microenvironments between CC subtypes, contributing to subtype‐specific immunotherapy and drug therapy.
toxicology,environmental sciences,water resources
What problem does this paper attempt to address?