An Immune-Related lncRNA Model for Predicting Prognosis, Immune Landscape and Chemotherapeutic Response in Bladder Cancer
Abstract Objective: Long noncoding RNAs (lncRNAs) participate in cancer immunity. Herein, we characterized the clinical significance of immune-related lncRNA model and its associations with immune infiltrations and chemosensitivity in bladder cancer.Methods: Transcriptome data of bladder cancer specimens were employed from The Cancer Genome Atlas. Dysregulated immune-related lncRNAs were screened via Pearson correlation and differential expression analyses, followed by recognition of lncRNA pairs. Then, a LASSO regression model was constructed. Receiver operator characteristic curves of one-, three- and five-year survival were plotted. Akaike information criterion (AIC) value of one-year survival was determined as the cutoff of high- and low-risk subgroups. The differences in survival, clinical features, immune cell infiltrations and chemosensitivity were compared between subgroups.Results: Totally, 90 immune-related lncRNA pairs were selected, 15 of which were put into the prognostic model. The area under the curves of one-, three- and five-year survival were 0.806, 0.825 and 0.828, confirming the favorable predictive performance of this model. According to the AIC value, we clustered subjects into high- and low-risk subgroups. High-risk score indicated unfavorable outcomes. This risk model was in relation to survival status, age, stage and TNM. In comparison to conventional clinicopathological characteristics, the risk model displayed higher predictive efficacy and was an independent predictor. Also, it could well characterize immune cell infiltration landscape and predict immune checkpoint expression and sensitivity to cisplatin and methotrexate.Conclusion: This model conducted by paring immune-related lncRNAs regardless of expressions exhibited a favorable efficacy in predicting prognosis, immune landscape and chemotherapeutic response in bladder cancer.