Comprehensive integration of tumor immune microenvironment features to predict prognosis and immunotherapy of hepatocellular carcinoma: results from large-scale and multiple cohorts
Comprehensive integration of tumor immune microenvironment features to predict prognosis and immunotherapy of hepatocellular carcinoma: results from large-scale and multiple cohorts
复制标题
综合整合肿瘤免疫微环境特征来预测肝细胞癌的预后和免疫治疗：大规模和多个队列的结果
作者：
Shuai Zhao;Wangrui Liu;Yichi Zhang;Shuxian Chen;Shiyin Wei;Han Ding;Bing Han;Xiaoling Song;Xiuqin Lu;Lifeng Jiang;Jun Li;Jian Wang
背景资料：肝细胞癌（hepatocellular carcinoma，HCC）的发生发展是一个复杂的生物学过程，肿瘤免疫微环境（tumor immune microenvironment，TIME）和恶性肿瘤表型对HCC的影响尚不清楚。研究方法：基于ImmPort和InnateDB的免疫相关指标，我们首先对样本进行聚类，并使用机器学习方法分析免疫亚型之间的差异表达基因。来自TCGA、LIRI-JP、GSE 76427的共790份HCC样本被招募为发现、检测和验证队列。ICI评分基于Boruta算法构建，用于评估独立的预后和预测性免疫治疗反应影响。我们还在我们中心进行了临床队列验证，以证明ICIscore对肝癌的预测作用。结果如下：首先，根据免疫基因表达水平、免疫细胞丰度、免疫评分和间质评分，我们获得了三个独立的免疫亚型，它们在HCC的生存和异质性差异方面具有显著性。其次，1022个免疫亚型的差异表达基因（DEG）将HCC分为三个免疫基因型，其预后和时间有显著差异。第三，建立了一种新的HCC免疫分型方法ICI评分，并对HCC的性别、分期、进展和DNA变异进行了统计学分析（P<0.05）。此外，ICI评分可以通过趋化因子信号传导、粘着斑和JAK/STAT信号传导途径显著预测对免疫治疗的应答。结论：本研究首先描述了肿瘤微环境和免疫表型簇提高了HCC患者预后的准确性。在HCC中发现的新的独立预后指标突出了肿瘤表型和免疫环境之间的关系。
Background: The occurrence and development of hepatocellular carcinoma (HCC) is a complex biologic process, and the influence of tumor immune microenvironment (TIME) and malignant tumor phenotypes remains unclear..Methods: Based on the immune-related indicators from ImmPort and InnateDB, we first clustered the samples and analyzed the differentially expressed genes among immune subtype using machine-learning methods. A total of 790 HCC samples from TCGA, LIRI-JP, GSE76427 were enrolled as discovering, testing and validation cohorts. ICI score was constructed based on Boruta algorithm and used to evaluate independent prognostic and predictive immunotherapy response implications.We also conducted clinical cohort verification at our center to prove the predictive effect of ICIscore on HCC..Results: First, according to the immune gene expression level, immune cell abundance, immunescore and stromalscore, we obtained three independent immune subtypes with significance in survival and heterogeneity differences of HCC. Secondly, differentially expressed genes (DEGs) among 1022 immune subtypes divide HCC samples into three immune genotyping with significant difference in prognosis and TIME. Third, the ICI score, novel immunophenotyping of HCC, was established and significantly differs gender, stage, progression and DNA variation profiles of HCC (P<0.05). Additionally, ICI score could markedly predict responses to immunotherapies via chemokine signaling, focal adhesion and JAK/STAT signaling pathways..Conclusion: This study first describes that tumor microenvironment and immunophenotyping clusters improve the accuracy of the prognosis of HCC patients. The new independent prognostic indicators discovered in HCC highlight the relationship between tumor phenotype and the immune environment.