Department of Integrated Traditional Chinese & Western Medicine, Second Xiangya Hospital, Central South University, Changsha, Hunan 410011, P.R.China.
Department of General Surgery, The Second Xiangya Hospital, Central South University, No.139 Middle Renmin Road, Changsha, Hunan410011, P.R. China.
Int J Med Sci. 2021 Jan 1;18(1):88-98. doi: 10.7150/ijms.49449. eCollection 2021.
Colorectal cancer (CRC) is a commonly occurring tumour with poor prognosis. Autophagy-related long non-coding RNAs (lncRNAs) have received much attention as biomarkers for cancer prognosis and diagnosis. However, few studies have focused on their prognostic predictive value specifically in CRC. This research aimed to construct a robust autophagy-related lncRNA prognostic signature for CRC. Autophagy-related lncRNAs from The Cancer Genome Atlas database were screened using univariate Cox, LASSO, and multivariate Cox regression analyses, and the resulting key lncRNAs were used to establish a prognostic risk score model. Furthermore, quantitative real-time polymerase chain reaction (qRT-PCR) analysis was performed to detect the expression of several lncRNAs in cancer tissues from CRC patients and in normal tissues adjacent to the cancer tissues. A prognostic signature comprising lncRNAs AC125603.2, LINC00909, AC016876.1, MIR210HG, AC009237.14, and LINC01063 was identified in patients with CRC. A graphical nomogram based on the autophagy-related lncRNA signature was developed to predict CRC patients' 1-, 3-, and 5-year survival. Overall survival in patients with low risk scores was significantly better than in those with high risk scores (P < 0.0001); a similar result was obtained in an internal validation sample. The nomogram was shown to be suitable for clinical use and gave correct predictions. The 1- and 3-year values of the area under the receiver operating characteristic curve were 0.797 and 0.771 in the model sample, and 0.656 and 0.642 in the internal validation sample, respectively. The C-index values for the verification samples and training samples were 0.756 (95% CI = 0.668-0.762) and 0.715 (95% CI = 0.683-0.829), respectively. Gene set enrichment analysis showed that the six autophagy-related lncRNAs were greatly enriched in CRC-related signalling pathways, including p53 and VEGF signalling. The qRT-PCR results showed that the expression of lncRNAs in CRC was higher than that in adjacent tissues, consistent with the expression trends of lncRNAs in the CRC data set. In summary, we established a signature of six autophagy-related lncRNAs that could effectively guide clinical prediction of prognosis in patients with CRC. This lncRNA signature has significant clinical implications for improving the prediction of outcomes and, with further prospective validation, could be used to guide tailored therapy for CRC patients.
结直肠癌（CRC）是一种预后不良的常见肿瘤。自噬相关长非编码 RNA（lncRNA）作为癌症预后和诊断的生物标志物受到了广泛关注。然而，很少有研究专门关注它们在 CRC 中的预后预测价值。本研究旨在构建一个稳健的自噬相关 lncRNA 预后签名用于 CRC。使用单因素 Cox、LASSO 和多因素 Cox 回归分析从癌症基因组图谱数据库筛选自噬相关 lncRNA，并使用筛选出的关键 lncRNA 建立预后风险评分模型。此外，通过定量实时聚合酶链反应（qRT-PCR）分析检测了来自 CRC 患者的癌症组织和癌症组织相邻正常组织中几种 lncRNA 的表达。在 CRC 患者中确定了由 lncRNA AC125603.2、LINC00909、AC016876.1、MIR210HG、AC009237.14 和 LINC01063 组成的预后特征。基于自噬相关 lncRNA 特征开发了一个图形列线图，用于预测 CRC 患者 1、3 和 5 年的生存情况。低风险评分患者的总生存率明显优于高风险评分患者（P < 0.0001）；在内部验证样本中也得到了类似的结果。该列线图适用于临床应用并给出了正确的预测。在模型样本中，1 年和 3 年的受试者工作特征曲线下面积分别为 0.797 和 0.771，内部验证样本中分别为 0.656 和 0.642。验证样本和训练样本的 C 指数值分别为 0.756（95%CI=0.668-0.762）和 0.715（95%CI=0.683-0.829）。基因集富集分析表明，这六个自噬相关 lncRNA 在包括 p53 和 VEGF 信号通路在内的 CRC 相关信号通路中高度富集。qRT-PCR 结果表明，lncRNA 在 CRC 中的表达高于相邻组织，与 CRC 数据集的 lncRNA 表达趋势一致。总之，我们建立了一个由六个自噬相关 lncRNA 组成的特征，可以有效地指导 CRC 患者预后的临床预测。该 lncRNA 特征对改善结局预测具有重要的临床意义，并可通过进一步的前瞻性验证，用于指导 CRC 患者的个体化治疗。
