Zhuang Hongkai, Huang Shanzhou, Zhou Zixuan, Ma Zuyi, Zhang Zedan, Zhang Chuanzhao, Hou Baohua
Department of General Surgery, Guangdong Provincial People's Hospital, Guangdong Academy of Medical Sciences, No. 106 Zhongshan Er Road, Guangzhou, 510080 China.
Shantou University of Medical College, Shantou, 515000 Guangdong China.
Cancer Cell Int. 2020 Oct 9;20:493. doi: 10.1186/s12935-020-01588-y. eCollection 2020.
Pancreatic cancer (PC) is one of the most common cancers and the leading cause of cancer-related death worldwide. Exploring novel predictive biomarkers for PC patients' prognosis is in urgent need.
In the present study, we conducted Cox proportional hazards regression to identify critical prognosis-associated lncRNAs (PALncs) in TCGA PC dataset. Based on the results of multivariate analysis, a PALnc-based risk score system was established, and validated in GSE62452 dataset. The validity and reliability of the risk score system for prognosis of PC were evaluated through ROC analysis. And function enrichment analyses for the PALncs were also performed.
In the multivariate analysis, four PALncs (LINC00476, C9orf163, LINC00346 and DSCR9) were screened out to develop a risk score system, which showed a high AUC at 3 and 5 years overall survival (0.785 at 3 year OS, 0.863 at 5 year OS) in TCGA datasets. And the ROC analysis of the risk score system for RFS in TCGA dataset revealed that AUC for RFS was 0.799 at 3 years and 0.909 at 5 years. Further, the AUC for OS in the validation cohort was 0.705 at 3 years and 0.959 at 5 years. Furthermore, the functional enrichment analysis revealed that these PALncs may be involved in various pathways related to cancer, including Ras family activation, autophagy in cancer, MAPK signaling pathway, HIF-1 signaling pathway, PI3K-Akt signaling pathway, etc. And correlation analysis of these tumor infiltrating immune cells and risk score system revealed that the infiltration level of B cell naïve, plasma cells, and CD8+ T cells are negatively correlated to the risk score system, while macrophages M2 positively correlated to the risk score system.
Our study established a four PALncs based risk score system, which reflects immune cell infiltration and predicts patient survival for PC.
胰腺癌（PC）是全球最常见的癌症之一，也是癌症相关死亡的主要原因。迫切需要探索用于预测PC患者预后的新型生物标志物。
在本研究中，我们进行Cox比例风险回归分析，以在TCGA PC数据集中识别关键的预后相关lncRNA（PALnc）。基于多变量分析结果，建立了基于PALnc的风险评分系统，并在GSE62452数据集中进行验证。通过ROC分析评估PC预后风险评分系统的有效性和可靠性。还对PALnc进行了功能富集分析。
在多变量分析中，筛选出四个PALnc（LINC00476、C9orf163、LINC00346和DSCR9）以建立风险评分系统，该系统在TCGA数据集中3年和5年总生存期显示出较高的AUC（3年总生存期时为0.785，5年总生存期时为0.863）。TCGA数据集中风险评分系统对无复发生存期（RFS）的ROC分析显示，3年时RFS的AUC为0.799，5年时为0.909。此外，验证队列中3年总生存期的AUC为0.705，5年时为0.959。此外，功能富集分析表明，这些PALnc可能参与多种与癌症相关的途径，包括Ras家族激活、癌症中的自噬、MAPK信号通路、HIF-1信号通路、PI3K-Akt信号通路等。这些肿瘤浸润免疫细胞与风险评分系统的相关性分析表明，幼稚B细胞、浆细胞和CD8 + T细胞的浸润水平与风险评分系统呈负相关，而M2巨噬细胞与风险评分系统呈正相关。
我们的研究建立了基于四个PALnc的风险评分系统，该系统反映免疫细胞浸润并预测PC患者的生存情况。
