Identification of immune-associated lncRNAs as a prognostic marker for lung adenocarcinoma.
Identification of immune-associated lncRNAs as a prognostic marker for lung adenocarcinoma.
复制标题
免疫相关lncRNA作为肺腺癌预后标志物的鉴定
DOI：
10.21037/tcr-20-2827
复制
发表时间：
2021-03
影响因子：
0.9
通讯作者：
Zhao X
中科院分区：
文献类型：
--
作者：
He C;Yin H;Zheng J;Tang J;Fu Y;Zhao X
肺腺癌（LUAD）在肺癌患者中所占比例最大，并且在全球范围内发病率和死亡率最高。越来越多的证据表明，免疫相关的长非编码 RNA (lncRNA) 在 LUAD 中发挥作用，尽管它们对免疫治疗和癌症相关死亡的预测价值仍然缺乏研究。基因表达谱和临床数据来自癌症基因组图谱。我们通过单变量和多变量Cox回归以及最小绝对收缩和选择算子回归分析构建了风险模型，随后将每个样本分为低风险或高风险类别。应用生存和受试者工作特征（ROC）分析来评估模型的预后价值。此外，还分析了两个风险组之间的免疫和体细胞突变状态。最后，将该模型应用于胰腺导管腺癌（PDAC）样本，探索该模型在其他癌症中的适用性。我们从 499 名 LUAD 患者中获取数据，并以 7:3 的比例将样本随机分为训练集 (N=351) 和验证集 (N=148)。我们检测到 7 种适用于建立风险特征的免疫相关 lncRNA（AP000695.2、AC026355.2、LINC01843、ITGB1-DT、LINC01150、AL590226.1 和 AC091185.1）。生存分析显示，高风险组患者的总生存期 (OS) 短于低风险组患者。 ROC 分析显示所有数据集中均具有出色的 AUC 值（1 年、3 年和 5 年时 >0.65）。值得注意的是，ESTIMATE算法和PCA、(ss)GSEA和体细胞突变的分析表明，高危组具有更强的免疫抑制状态和更高的肿瘤突变负担(TMB)。此外，由于免疫检查点受体基因和 TLS 相关基因水平较高，低风险组患者对免疫治疗的反应更好。我们使用 7 种免疫相关 lncRNA 的模型对 PDAC 患者显示出类似的适用性。我们构建了基于 7 个免疫相关 lncRNA 的风险特征模型，并显示了其对于识别 LUAD 患者免疫和体细胞突变特征的预后价值，这可能有助于临床治疗计划并阐明 LUAD 免疫的分子机制。
Lung adenocarcinoma (LUAD) accounts for the largest proportion of lung cancer patients and has the highest morbidity and mortality worldwide. Accumulating evidence shows that immune-associated long non-coding RNAs (lncRNAs) play a role in LUAD, although their predictive value for immunotherapy treatment and cancer-related death remains poorly investigated. Gene expression profiles and clinical data were obtained from The Cancer Genome Atlas. We constructed a risk model by univariate and multivariate Cox regression and least absolute shrinkage and selection operator regression analysis and subsequently divided each sample into low- or high-risk category. Survival and receiver operating characteristic (ROC) analyses were applied to assess the prognostic value of the model. Additionally, immune and somatic mutation status were analysed between the two risk groups. Finally, the model was applied to pancreatic ductal adenocarcinoma (PDAC) samples to explore the applicability of the model in other cancers. We obtained data from 499 LUAD patients and randomised the samples into a training set (N=351) and validation set (N=148) at a 7:3 ratio. We detected 7 immune-associated lncRNAs (AP000695.2, AC026355.2, LINC01843, ITGB1-DT, LINC01150, AL590226.1 and AC091185.1) that were applicable for establishing a risk signature. Survival analysis revealed that patients categorised in the high-risk group had shorter overall survival (OS) than those in the low-risk group. ROC analyses showed excellent AUC values in all data sets (>0.65 at 1, 3, and 5 years). Notably, ESTIMATE algorithm and analysis of PCA, (ss)GSEA, and somatic mutations revealed that the high-risk group had a stronger immunosuppressive status and a higher tumour mutation burden (TMB). Moreover, patients in the low-risk group responded better to immunotherapy due to higher levels of immune-checkpoint receptor genes and TLS-related genes. Our model using the 7 immune-associated lncRNAs showed similar applicability for PDAC patients. We constructed a model for risk signatures based on 7 immune-associated lncRNAs and showed its prognostic value for identifying immune and somatic mutation characteristics in LUAD patients, which may assist clinical treatment plans and elucidate molecular mechanisms of LUAD immunity.