Abstract
BACKGROUND: Endometrial carcinoma (EC) is a malignant cancer spreading worldwide and in the fourth position among all other types of cancer in women. The purpose of this paper is to explore the prognostic value of the immune-autophagy gene in endometrial carcinoma (EC) based on bioinformatics, construct an immune-autophagy prognostic model of endometrial carcinoma, search for independent prognostic markers, and finally predict the potential therapeutic drugs of TCGA subgroup. METHODS: The Cancer Genome Atlas (TCGA) database was used to extract transcriptome sequencing data of patients suffering from EC; 28 kinds of immune cells were scored by ssGSEA, and the immune subtypes were grouped by consistency cluster analysis. The accuracy and effectiveness of the grouping were verified by the analysis of differential gene expression and survival rate of immune checkpoints in the two groups to provide the premise and basis for the establishment of independent prognostic factors. The expression of different genes in high and low immune groups was analyzed. The analysis of various genes' expression in immune groups (high and low) has been performed. Go function annotation and KEGG pathway enrichment analysis were used to evaluate the difference of immune infiltration between high and low immune groups. The immune and autophagy genes were crossed, the key (hub) genes were selected, the risk was scored, the prognosis model was constructed, and the independent prognostic markers were established. CAMP and CTRP 2.0 were used to test the drug sensitivity. RESULTS: According to the level of immune cell enrichment, the results have been subcategorized into two immune subtypes: high immunity group_ H and low immunity group_ L. Two immune subtypes, CD274, PDCD1, and CTLA4, were detected in the immune system_ H and immunity_L. A significant difference was detected between these two groups in the expression and survival rate. Few more differences were also detected between the two groups through the evaluation of immune infiltration, which proved the grouping's accuracy and effectiveness. Differential gene expression analysis showed that there were 721 DEGs and 3 hub genes. DEGs are mainly involved in lymphocyte activation, proliferation, differentiation, leukocyte proliferation, and other biological processes, mediate chemokines' activities, chemokine receptor binding, and other molecular functions, and are enriched in the outer plasma membrane, endoplasmic reticulum, and T cell receptor complex. The enriched pathways are allograft, complex, inflammatory, interferon-alpha, interferon-gamma, E2F, G2M, mitotic, etc. CONCLUSION: Through bioinformatics analysis, we successfully constructed the immuno-autophagy prognosis model of endometrial cancer and identified three high-risk immunoautophagy genes, including VEGFA, CCL2, and Ifng. Four potential therapeutic drugs were predicted as sildenafil, sunitinib, TPCA-1, and etoposide.