DataSheet_1_Predicting Prognosis and Distinguishing Cold and Hot Tumors in Bladder Urothelial Carcinoma Based on Necroptosis-Associated lncRNAs.pdf

BackgroundIn reference to previous studies, necroptosis played an important role in cancer development. Our team decided to explore the potential prognostic values of long non-coding RNAs (lncRNAs) associated with necroptosis in bladder urothelial carcinoma (BLCA) and their relationship with the tumor microenvironment (TME) and the immunotherapeutic response for accurate dose.
MethodsTo obtain the required data, bladder urothelial carcinoma transcriptome data were searched from Cancer Genome Atlas (TCGA) (https://portal.gdc.cancer.gov/). We used co-expression analysis, differential expression analysis, and univariate Cox regression to screen out prognostic lncRNAs associated with necroptosis in BLCA. Then the least absolute shrinkage and selection operator (LASSO) was conducted to construct the necroptosis-associated lncRNAs model. Based on this model, we also performed the Kaplan–Meier analysis and time-dependent receiver operating characteristics (ROC) to estimate the prognostic power of risk score. Multivariate and univariate Cox regression analysis were performed to build up a nomogram. Calibration curves, and time-dependent ROC were also conducted to evaluate nomogram. Principal component analysis (PCA) revealed a difference between high- and low-risk groups. In addition, we explored immune analysis, gene set enrichment analyses (GSEA), and evaluation of the half-maximal inhibitory concentration (IC50) in constructed model. Finally, the entire samples were divided into three clusters based on model of necroptosis-associated lncRNAs to further compare immunotherapy in cold and hot tumors.
ResultsA model was built up based on necroptosis-associated lncRNAs. The model revealed good consistence between calibration plots and prognostic prediction. The area of 1-, 3-, and 5-year OS under the ROC curve (AUC) were 0.707, 0.679, and 0.675. Risk groups could be helpful for systemic therapy due to the markedly diverse IC50 between risk groups. To our delight, clusters could effectively identify cold and hot tumors, which would be beneficial to accurate mediation. Clusters 2 and 3 were considered the hot tumor, which was more sensitive to immunotherapeutic drugs.
ConclusionsThe outcomes of our study suggested that necroptosis-associated lncRNAs could effectively predict patients with BLCA prognosis, which may be helpful for distinguishing the cold and hot tumors and improving individual treatment of BLCA.

研究背景
既往研究表明，细胞坏死性凋亡（necroptosis）在癌症发生发展中发挥重要作用。本团队旨在探索与细胞坏死性凋亡相关的长链非编码RNA（long non-coding RNAs, lncRNAs）在膀胱尿路上皮癌（bladder urothelial carcinoma, BLCA）中的潜在预后价值，以及其与肿瘤微环境（tumor microenvironment, TME）和免疫治疗应答的关联，以期实现精准诊疗。

研究方法
为获取所需数据，本研究从癌症基因组图谱（Cancer Genome Atlas, TCGA）数据库（https://portal.gdc.cancer.gov/）检索膀胱尿路上皮癌转录组数据。采用共表达分析、差异表达分析及单变量Cox回归筛选出与膀胱尿路上皮癌中细胞坏死性凋亡相关的预后性长链非编码RNA。随后通过最小绝对收缩和选择算子（least absolute shrinkage and selection operator, LASSO）构建细胞坏死性凋亡相关长链非编码RNA风险模型。基于该模型，采用Kaplan–Meier分析与时序性受试者工作特征（time-dependent receiver operating characteristics, ROC）曲线评估风险评分的预后效能。进一步通过多变量及单变量Cox回归分析构建列线图，并采用校准曲线及时序性ROC曲线评估列线图的预测性能。主成分分析（Principal component analysis, PCA）结果显示高低风险组间存在显著差异。此外，本研究还对构建的风险模型进行了免疫浸润分析、基因集富集分析（gene set enrichment analyses, GSEA）以及半最大效应浓度（half-maximal inhibitory concentration, IC50）评估。最后，基于细胞坏死性凋亡相关长链非编码RNA模型将全部样本分为3个亚型，以进一步对比冷肿瘤与热肿瘤的免疫治疗效果。

研究结果
本研究成功构建了细胞坏死性凋亡相关长链非编码RNA风险模型。校准曲线与预后预测结果具有良好的一致性。该模型的1年、3年、5年总生存期（overall survival, OS）受试者工作特征曲线下面积（area under the ROC curve, AUC）分别为0.707、0.679及0.675。由于不同风险组间半最大效应浓度存在显著差异，风险分组可为系统性治疗提供参考。令人欣喜的是，样本亚型可有效区分冷肿瘤与热肿瘤，这将有助于实现精准诊疗。亚型2与亚型3被归类为热肿瘤，其对免疫治疗药物更为敏感。

研究结论
本研究结果表明，细胞坏死性凋亡相关长链非编码RNA可有效预测膀胱尿路上皮癌患者的预后，有助于区分冷肿瘤与热肿瘤，进而为膀胱尿路上皮癌的个体化治疗提供参考。