Construction of a new tumor immunity-related signature to assess and classify the prognostic risk of ovarian cancer

Abstract

Ovarian cancer is associated with a high mortality rate. In this study, we established a new immune-related signature that can stratify ovarian cancer patients. First, we obtained immune-related genes through IMMUPORT, and DEGs (Differential Expression Genes) by analyzing the GSE26712 dataset. The APP (Antigen Processing and Presentation) and DEG signatures were established using univariate and multivariate Cox models. Kaplan-Meier analysis revealed the signatures’ prognostic value in training and validation cohorts (HR: 0.379 VS. 0.450; 0.333 VS. 0.327). Nomogram analysis was used to assess the signatures’ ability to predict the 30-month prognosis, which was evaluated using the calibration curve and time-dependent ROC curve (30-month AUC: 0.665 VS. 0.743). Time-dependent ROC, Decision Curve Analysis (DCA) and Integrated discrimination improvement (IDI) was used to compare the new model to previously published gene signatures. 30-month AUC composite variable (0.736) was higher than 9-gene signature (0.657), and composite variable had a larger net benefit and a higher IDI (+2.436%) relative to the 9-gene signature. Tumor immune infiltration and tumor microenvironment scores of the 2 groups separated by APP signature were compared. GSEA was used to identify enriched KEGG pathways. Conclusively, the proposed signature can stratify ovarian cancer patients by risk-score and guide clinical decisions.