Development and validation of a risk model for effective immune and stromal related signature predicting prognosis of patients with ovarian cancer

Abstract

Abstract The tumor microenvironment (TME) plays a critical role in ovarian cancer (OC) progression, yet the relationship between immune and stromal scores within the TME and prognostic outcomes remains poorly understood. Immune and stromal cell scores were computed using the “estimate” R package, which enabled the assessment of immune and stromal components in OC samples. We then performed univariate and multivariate Cox regression analyses to identify prognostic factors associated with these scores using data from The Cancer Genome Atlas (TCGA). Additionally, LASSO Cox regression were employed to identify key prognostic genes linked to immune infiltration. Our analysis of OC expression data identified 1,667 differentially expressed genes (DEGs) associated with immune and stromal scores. From these, we developed a 6-gene risk model, consisting of ALOX5AP, FCGR1C, GBP2, IL21R, KLRB1, and PIK3CG, which effectively stratified OC patients into high-risk and low-risk groups. Survival analysis and area under the curve (AUC) assessment confirmed the model’s strong predictive accuracy. Furthermore, drug sensitivity predictions indicated that sorafenib was particularly effective in high-risk patients, with this finding validated through in vitro experiments. The 6-gene TME-related risk model offers robust prognostic capabilities for OC and could serve as a valuable tool for clinical stratification and personalized treatment approaches.

Keywords

• Ovarian cancer
• ESTIMATE
• GSEA
• Prognostic signature
• Immune drug response