A Novel Platinum Resistance-related Immune Gene Signature for Overall Survival Prediction in Patients with Ovarian Cancer

Abstract

Abstract Background Ovarian cancer (OV) is a highly heterogeneous gynaecological tumor that makes the prognostic prediction challenging. Resistance to platinum-based chemotherapy is associated with a poor prognosis in OV. There seems to be an overlap between molecular mechanisms responsible for platinum resistance and immunogenicity in OV. However, the predictive role of platinum-resistance-related immune genes for OV prognosis needs to be further explored. Methods In our study, the mRNA expression data of OV patients with corresponding clinical information was collected from the TCGA and ICGC cohort. A multigene signature was constructed for OV patients in the TCGA cohort using the least absolute shrinkage and selection operator (LASSO) Cox regression model, and was validated in the ICGC cohort. Furthermore, we performed functional analysis to explore the immune status between the two risk groups. Results Our data showed that there were 41.1% of the platinum resistance-related genes differentially expressed between immune score low and high OV patients in the TCGA cohort. Univariate Cox regression analysis identified 30 differentially expressed genes (DEGs) associated with overall survival (OS) (P < 0.05). A 14-gene signature was established to classify OV patients into a low- and high-risk group. Patients in the low-risk group showed significantly higher OS than those in the high-risk group (P < 0.0001 in the both TCGA and ICGC cohort), which was associated with different immune status for the two risk groups. Conclusion A novel platinum resistance-related immune model can be used for prognostic prediction in OV. Targeting tumor immunity may be a therapeutic alternative for OV with platinum resistance.