Identification of platinum resistance-related gene signature for prognosis and immune analysis in bladder cancer.

Currently, there is limited knowledge about platinum resistance-related long non-coding RNAs (lncRNAs) in bladder cancer. We aim to identify platinum resistance-related lncRNAs and construct a risk model for accurate prognostic prediction of bladder cancer. Transcriptomic and clinical data were extracted from The Cancer Genome Atlas (TCGA) database, and platinum resistance-related genes were obtained from HGSOC-Platinum. The platinum resistance-related lncRNAs were obtained by the Spearman correlation analysis. Then, we constructed a risk score model through Cox regression analysis and the LASSO algorithm. The model was verified by analyzing the median risk score, Kaplan-Meier curve, receiver operating characteristic (ROC) curve, and heatmap. We also developed a nomogram and examined the relationship between the risk score model, immune landscape, and drug sensitivity. Lastly, we assessed the differential expression of PRR-lncRNAs in the cisplatin-resistant bladder cancer cell line and the normal bladder cancer cell line using qRT-PCR. We developed and validated an eight-platinum resistance-related lncRNA risk model for bladder cancer. The risk model showed independent prognostic significance in univariate and multivariate Cox analyses. Based on multivariate analysis, we developed a nomogram. The modified model is both good predictive and clinically relevant after evaluation. Furthermore, immune-related and drug-sensitivity analyses also showed significant differential expression between high and low-risk groups. The qRT-PCR demonstrated that most of the lncRNAs were upregulated in cisplatin-resistance cancerous tissues than in control tissues. We have developed a predictive model based on eight platinum resistance-related lncRNAs, which could add meaningful information to clinical decision-making.