Identification of m(6)A-related long non-coding RNAs for predicting prognosis and immune characterizations in gastric cancer

Background: N6-methyladenosine (m(6)A) mRNA modification triggers malignant behavior in tumor cells, which promotes malignant progression and migration of gastric cancer (GC). Nevertheless, studies on the prognostic value of m(6)A-related long non-coding RNA (MRlncRNA) in GC remain quite restricted. The study aimed to develop a reasonable predictive model to explore the prognostic potential of MRlncRNAs in predicting the prognosis of GC patients and monitoring the efficacy of immunotherapy. Methods: Transcriptomic and clinical data for GC were derived from TCGA. Next, univariate Cox, LASSO and multivariate Cox regression analyses were next used to identify prognostic MRlncRNAs, calculate risk scores and build risk assessment models. The predictive power of the risk models was then validated by Kaplan-Meier analysis, ROC curves, DCA, C-index, and nomogram. We attempted to effectively differentiate between groups in terms of immune cell infiltration status, ICI-related genes, immunotherapy responses, and common anti-tumor drug sensitivity. Results: A risk model based on 11 MRlncRNAs was developed with an AUC of 0.850, and the sensitivity and specificity of this model in predicting survival probability is satisfactory. The Kaplan-Meier analysis revealed that the low-risk group in the model had a significantly higher survival rate, and the model was highly associated with survival status, clinical features, and clinical stage. Furthermore, the model was verified to be an independent prognostic risk factor, and the low-risk group in the model had a remarkable positive correlation with a variety of immune cell infiltrates. The expression levels of ICI-related genes differed significantly between the different groups. Lastly, immunotherapy responses and common anti-tumor drug sensitivity also differed significantly between different groups. Conclusion: The risk model on the basis of 11-MRlncRNAs can serve as independent predictors of GC prognosis and may be useful in developing personalized treatment strategies for patients.