Identification Of A Ten Long Non-coding RNA Signature Predicting Survival In Bladder Urothelial Carcinoma

Objective: Bladder urothelial carcinoma(BLCA)is one of the most prevalent cancers with high malignancy and worse prognosis.Patients with BLCA are in urgent need of precise prognosis and treatment.Prognostic biomarkers might improve risk stratification and survival prediction.However,conventional clinicopathologic characteristics,with their insufficient prognosis accuracy,is easy to be blamed as predictive factors.This study aimed to identify a novel lncRNA(long no-coding RNA)signature predicting patients' survival in BLCA.Methods: Following a gene mining strategy,the lncRNA expression data of 520 patients with BLCA from Gene Expression Omnibus(GEO)Database were use to established a lncRNA signature with prognostic significance.Furthermore,several statistical approaches were performed to validate the prognostic value of this signature.Results: Ten lncRNAs were identified to develop a predictive signature and a risk model.Based on the signature,patients in series GSE32894 were divided into high-risk and low-risk subgroups with remarkably different disease-specific survival(HR=1.98,95%CI=1.6-2.4,p<0.001).The predictive ability of this signature was then confirmedin another two independent series by Kaplan–Meier analysis.Multivariate Cox regression analysis revealed the independence of the signature of prevailing clinicopathologic characteristics,including tumor stage and histologic grade.ROC curves showed that this signature exhibited definitely no less predictive accuracy than other known prognostic factors,even some popular immune checkpoints and several potential therapeutic targets.Gene set enrichment analysis identified canonical pathways that were associated with this prognostic signature.Conclusion: The identification this 10-lncRNA signature,through validation,may improve the prognostic prediction and precision medicine in BLCA.