Development of a novel six-long non-coding RNA signature predicting survival of patients with bladder urothelial carcinoma

Bladder urothelial carcinoma is a malignant tumor with a high incidence in the uropoietic system. Considerable studies have shown that long non-coding RNA (lncRNA) plays an important role in the development and progression of bladder urothelial carcinoma. In this study, the lncRNA expression and clinical data of 377 bladder urothelial carcinoma patients were obtained from The Cancer Genome Atlas (TCGA) database, and differentially expressed lncRNAs in the cancer and normal groups were evaluated. Univariate COX and multivariate COX regression analyses of prognosis were performed on differentially expressed lncRNAs in the training datasets, six prognosis-related lncRNAs (LINC02195, LINC01484, LINC01468, SMC2-AS1, AC011298.1, and PTPRD-AS1) were assessed, and a 6-lncRNA signature was constructed. The predictive capability of this 6-lncRNA signature was validated in the testing datasets and entire datasets. The prognostic ability of 6-lncRNA signature was independent of other clinical elements after multivariate COX regression and stratified analyses of with other clinical elements. We performed functional enrichment analysis with the six prognosis-related lncRNAs. Results of functional enrichment revealed that these prognosis-related lncRNAs might promote the development and metastasis of bladder urothelial carcinoma. In summary, the 6-lncRNA signature that we developed could effectively predict the prognosis of bladder urothelial carcinoma patients. This 6-lncRNA signature might be a novel independent prognostic marker of bladder urothelial carcinoma. Moreover，it also provides novel insights into the mechanism of bladder urothelial carcinoma.

膀胱尿路上皮癌（bladder urothelial carcinoma）是泌尿系统高发的恶性肿瘤。诸多研究证实，长链非编码RNA（long non-coding RNA, lncRNA）在膀胱尿路上皮癌的发生与进展过程中发挥关键调控作用。本研究从癌症基因组图谱（The Cancer Genome Atlas, TCGA）数据库中获取了377例膀胱尿路上皮癌患者的lncRNA表达谱及临床资料，对癌组织与正常组织间的差异表达lncRNA进行筛选与评估。在训练数据集内，针对差异表达lncRNA开展单因素COX及多因素COX预后回归分析，最终筛选出6个预后相关lncRNA（LINC02195、LINC01484、LINC01468、SMC2-AS1、AC011298.1及PTPRD-AS1），并构建了基于该6个lncRNA的预后特征模型。随后在测试数据集及全部数据集内验证了该6-lncRNA特征的预测效能。经多因素COX回归分析及分层临床因素分析后证实，该6-lncRNA特征的预后评估能力独立于其他临床病理因素。本研究还针对这6个预后相关lncRNA开展了功能富集分析，结果显示这些预后相关lncRNA可能通过调控通路促进膀胱尿路上皮癌的发生与转移。综上，本研究构建的6-lncRNA特征可有效预测膀胱尿路上皮癌患者的预后，有望成为膀胱尿路上皮癌新型独立预后标志物，同时也为阐明膀胱尿路上皮癌的发病机制提供了全新的研究视角。