Crosstalk of Redox-Related Subtypes, Establishment of a Prognostic Model and Immune Responses in Endometrial Carcinoma

Simple Summary In order to explore the role of redox as a prognostic indicator in endometrial carcinoma (EC), we detected the expression patterns of 55 redox-related genes (RRGs) in EC cohorts from public databases. Performing consensus cluster algorithm, we determined four molecular subclusters based on RRGs which had significant differences in overall survival (OS) and immune activities of EC patients. Furthermore, we developed a prognostic risk model on the basis of the redox-related subtype by stepwise Cox regression analyses. All EC patients were divided into high-risk and low-risk groups according to the median value of risk score. Our proposed model could accurately assess the clinical outcome and had favorable independent ability in EC cases. Moreover, our signature can serve as a predictor for immune status and chemotherapy sensitivity. Redox plays a central part in the pathogeneses and development of tumors. We comprehensively determined the expression patterns of redox-related genes (RRGs) in endometrial carcinoma (EC) cohorts from public databases and identified four different RRG-related clusters. The prognosis and the characteristics of TME cell infiltration of RRGcluster C patients were worse than those of other RRG clusters. When it comes to the gene cluster, there were great differences in clinicopathology traits and immunocyte infiltration. The RRG score was calculated by Cox analyses, and an RRG-based signature was developed. The risk score performed well in the EC cohort. Samples were separated into two risk subgroups with the standard of the value of the median risk score. Low-risk patients had a better prognosis and higher immunogenicity. In addition, RRG score was closely associated with immunophenoscore, microsatellite instability, tumor mutation burden, tumor stem cell index, copy number variation and chemotherapy sensitivity. The nomogram accurately predicted the prognosis of patients, and our model showed better performance than other published models. In conclusion, we built a prognostic model of RRGs which can help to evaluate clinical outcomes and guide more effective treatment.