Stemness-Relevant Gene Signature for Chemotherapeutic Response and Prognosis Prediction in Ovarian Cancer

Background: Ovarian cancer (OC) stands as the leading cause of cancer-related deaths among women, globally, owing to metastasis and acquired chemoresistance. Cancer stem cells (CSCs) are accountable for tumor initiation and exhibit resistance to chemotherapy and radiotherapy. Identifying stemness-related biomarkers that can aid in the stratification of risk and the response to chemotherapy for OC is feasible and critical. Methods: Gene expression and clinical data of patients with OC were downloaded from The Cancer Genome Atlas (TCGA) and the Gene Expression Omnibus (GEO) database. Four thousand three hundred seventeen stemness-related genes (SRGs) were acquired from the StemChecker database. TCGA was used as the training dataset, while GSE30161 served as validation dataset. Univariate Cox regression analysis was used to identify overall survival (OS)-related SRGs, and multivariate Cox regression analysis and random survival forest analysis were used for generating stemness-relevant prognostic model. Kaplan-Meier plots were used to visualize survival functions. Receiver operating characteristic (ROC) curves were used to assess the prognostic predictive ability of SRG-based features. Associations between signature score, tumor immune phenotype, and response to chemotherapy were analyzed via TIMER 2.0 and oncoPredict R package, respectively. A cohort of Shanghai Cancer Center was employed to verify the predictive robustness of the signature with respect to chemotherapy response. Results: Seven SRGs (actin-binding Rho activating C-terminal like (ABRACL), growth factor receptor bound protein 7 (GRB7), Lin-28 homolog B (LIN28B), lipolysis stimulated lipoprotein receptor (LSR), neuromedin U (NMU), Solute Carrier Family 4 Member 11 (SLC4A11), and thymocyte selection associated family member 2 (THEMIS2)) were found to have excellent predictive potential for patient survival. Patients in the high stemness risk group presented a poorer prognosis (p  < 0.0001), and patients with lower stemness scores were more likely to benefit from chemotherapy. Several tumorigenesis pathways, such as mitotic spindle and glycolysis, were enriched in the high stemness risk group. Tumor with high-risk scores tended to be in a status of relatively high tumor infiltration of CD4+ T cells, neutrophils, and macrophages, while tumor with low-risk scores tended to be in a status of relatively high tumor infiltration of CD8+ T cells. Conclusions: The stemness-relevant prognostic gene signature has the potential to serve as a clinically helpful biomarker for guiding the management of OC patients.