Abstract
BACKGROUND: With a lack of specific symptoms, ovarian cancer (OV) is often diagnosed at an advanced stage. This coupled with inadequate prognostic indicators and treatments with limited therapeutic effect make OV the deadliest type of gynecological tumor. Recent research indicates that N6-methyladenosine (m6A) and long-chain non-coding RNA (lncRNA) play important roles in the prognosis of OV and the efficacy of immunotherapy. RESULTS: Using the Cancer Genome Atlas (TCGA) OV-related data set and the expression profiles of 21 m6A-related genes, we identified two m6A subtypes, and the differentially expressed genes between the two. Based on the differentially expressed lncRNAs in the two m6A subtypes and the lncRNAs co-expressed with the 21 m6A-related genes, single-factor cox and LASSO regression were used to further isolate the 13 major lncRNAs. Finally, multi-factor cox regression was used to construct a m6A-related lncRNA risk score model for OV, with good performance in patient prognosis. Using risk score, OV tumor samples are divided into with high- and low-score groups. We explored the differences in clinical characteristics, tumor mutational burden, and tumor immune cell infiltration between the two groups, and evaluated the risk score's ability to predict the benefit of immunotherapy. CONCLUSION: Our m6A-based lncRNA risk model could be used to predict the prognosis and immunotherapy response of future OV patients.