Das Shaoli, Camphausen Kevin, Shankavaram Uma
Bioinformatics Core Facility, Radiation Oncology Branch, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892, USA.
Cancers (Basel). 2020 Sep 1;12(9):2476. doi: 10.3390/cancers12092476.
To elucidate the role of immune cell infiltration as a prognostic signature in solid tumors, we analyzed immune-function-related genes from four publicly available single-cell RNA-Seq data sets and twenty bulk tumor RNA-Seq data sets from The Cancer Genome Atlas (TCGA). Unsupervised clustering of pan-cancer transcriptomic signature showed two major immune function types: one related to NK-, T-, and B-cell functions and the other related to monocyte, macrophage, dendritic cell, and Toll-like receptor functions. Kaplan-Meier analysis showed differential prognosis of these two groups, dependent on the cancer type. Our analysis of TCGA solid tumors with an elastic net model identified 155 genes associated with disease-free survival in different tumor types with varied influence across different cancer types. With this gene set, we computed cancer-specific prognostic immune score models for individual cancer types that predicted disease-free and overall survival. Validation of our model on available published data of immune checkpoint blockade therapies on melanoma, kidney renal cell carcinoma, non-small cell lung cancer, gastric cancer and bladder cancer confirmed that cancer-specific higher immune scores are associated with response to immunotherapy. Our analysis provides a comprehensive map of cancer-specific immune-related prognostic gene sets that are associated with immunotherapy response.
为了阐明免疫细胞浸润作为实体瘤预后特征的作用，我们分析了来自四个公开可用的单细胞RNA测序数据集以及来自癌症基因组图谱（TCGA）的二十个肿瘤组织RNA测序数据集的免疫功能相关基因。泛癌转录组特征的无监督聚类显示出两种主要的免疫功能类型：一种与自然杀伤细胞、T细胞和B细胞功能相关，另一种与单核细胞、巨噬细胞、树突状细胞和Toll样受体功能相关。Kaplan-Meier分析表明，这两组的预后存在差异，具体取决于癌症类型。我们使用弹性网络模型对TCGA实体瘤进行分析，确定了155个与不同肿瘤类型无病生存期相关的基因，这些基因在不同癌症类型中的影响各不相同。利用这个基因集，我们为每种癌症类型计算了癌症特异性的预后免疫评分模型，该模型可预测无病生存期和总生存期。我们的模型在黑色素瘤、肾细胞癌、非小细胞肺癌、胃癌和膀胱癌的免疫检查点阻断疗法的现有已发表数据上进行验证，证实癌症特异性较高的免疫评分与免疫治疗反应相关。我们的分析提供了与免疫治疗反应相关的癌症特异性免疫相关预后基因集的全面图谱。
