A novel lipid metabolism-related lncRNA signature predictive of clinical prognosis in cervical cancer

Background: Cervical cancer (CC) is a serious threat to women populations worldwide. Lipid metabolism is believed to have modulating functions in cancer. Long non-coding RNAs (lncRNAs) are potential biomarkers for the different tumor prognosis. Our work aims at investigating the prognostic value of lipid metabolism-related lncRNAs in CC. Methods: LncRNA expression profiling was conducted in 291 patients from The Cancer Genome Atlas (TCGA). Patient samples were randomly assigned to the training or testing set in a 3:2 ratio. A novel lipid metabolism-related five-lncRNA signature with prognostic value for CC was built through the univariate Cox regression, least absolute contraction and selection operator (LASSO) regression and multivariate Cox regression analyses, and was further evaluated by the Kaplan-Meier methods. Relevant analyses were also applied to identify the independent clinicopathological factors. GO and KEGG analyses were conducted to investigate the biological functions and molecular pathways. Immune infiltration analysis was included to probe the relationship between lncRNA signature and cancer cell microenvironment. Results: The novel lipid metabolism-related five-lncRNA signature was confirmed to be predictive of overall survival (OS) in CC patients. Risk score, cancer stage, pregnancy, and BMI were validated as independent factors with prognostic value. GO and KEGG indicated that lipid metabolism participated in several tumor associated functions and pathways. Moreover, our results suggested that the five-lncRNA expression has potential link with tumor immune microenvironment. Conclusion: In conclusion, we built an innovative prognostic risk signature based upon lipid metabolism-related lncRNAs. The five-lncRNA signature may be beneficial to provide novel potential therapeutic targets and improve personalized treatment strategies for CC patients in future clinical treatments.