Identification of Tumor Microenvironment Scoring Scheme Based on Bioinformatics Analysis of Immune Cell Infiltration Pattern of Ovarian Cancer

Abstract

Background. Tumor microenvironment (TME) is the crucial mediator of tumor progression, and the TME model based on immune cell infiltration to characterize ovarian cancer is considered to be a promising strategy. Methods. Sample data of three ovarian cancer cohorts were obtained from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. The scores of 22 kinds of immune cells were calculated based on CIBERSORT, and the TME clusters (TMECs) of ovarian cancer was determined by ConsensusClusterPlus. Genomic subtype was identified by non-negative matrix factorization (NMF). A TME scoring scheme was constructed using k-means algorithm and principal component analysis (PCA) to quantify the TME infiltration pattern of individuals. Results. Four TME subtypes of ovarian cancer samples were defined: TMEC1, TMEC2, TMEC3, and TMEC4. There were also significant differences in overall survival (OS) among the four TMEC, and the OS of TMEC3 was the longest. The difference analysis of TMEC3 and the other three TMECs respectively identified the DEGs and took the intersection, and 585 DEGs were obtained. Two genomic subtypes were identified by NMF based on the expression of 585 genes, which were called GeneC1 and GeneC2. The TME scoring scheme constructed by k-means and PCA algorithm was used to calculate the TME score of ovarian cancer in TCGA. High-TME score was significantly correlated with shorter survival time, older age, lower immunoactivated molecules, and immune checkpoint gene expression. Conclusions. This study highlighted the complexity and diversity of TME infiltration patterns in ovarian cancer and constructed a set of TME scoring scheme to reveal TME infiltration patterns and provided new insights into the landscape of TME.