Ruyue Pan

Background Cuproptosis and oxidative stress (COS) are emerging regulators in cancer biology. However, their link to long non-coding RNAs (lncRNAs) and clinical outcomes in ovarian cancer remains unclear. Objective This study aimed to construct and validate a prognostic signature based on COS-related lncRNAs to improve risk stratification and provide insights into therapeutic strategies for ovarian cancer. Methods RNA-seq and clinical data from TCGA and GEO were analyzed to identify COS-related lncRNAs via WGCNA and Pearson correlation. Prognostic lncRNAs were screened using Cox regression and modeled using multiple machine learning algorithms. Immune profiles, mutation patterns, drug sensitivity, and experimental validation were performed. Results A 12-lncRNA signature was established that stratified patients into high- and low-risk groups with significant survival differences (HR = 22.6145, p < 0.001). The model showed strong predictive performance (AUCs: 0.91/0.967/0.974) and was validated externally. High-risk patients exhibited greater mutation burden, altered immune pathways (interferon, TGF-β), and differential predicted sensitivity to agents like Axitinib and Pazopanib. RT-qPCR confirmed the expression patterns of 11 out of 12 lncRNAs. Conclusion This study proposes a robust 12-lncRNA signature linking cuproptosis and oxidative stress with prognosis and therapy response in ovarian cancer, offering preliminary insights for personalized treatment guidance.