肝癌免疫治疗少，且临床试验效果不显著，个体差异大，副作用不明确。究其原因是缺少有效的临床前动物模型。 申请人通过原代肝癌异种移植人源化小鼠模型发现，肝癌PD-L1等免疫抑制分子的表达与GPC3嵌合抗原受体T细胞的杀伤效应成负相关，提示：肿瘤组织的免疫抑制环境，可能是免疫治疗的临床前与临床试验结果存在差异的主要因素。为了验证该假设，本项目拟在高度免疫缺陷小鼠NSI中，构建同一肝癌病人肿瘤与免疫共存的人源化模型；并对比该模型与原代肝癌组织形态、肿瘤抗原、转录表达、浸润免疫细胞亚群等的异同点，探索该模型是否可重现人体内肝癌对免疫细胞的抑制作用，有效模拟免疫治疗在人体内的疗效；申请人还将利用自主研发的肝脏自发损伤的NSIF小鼠构建病人肝脏人源化模型，以评估免疫治疗的个性化肝毒性。本项目同时建立肝癌样本及对应人源化模型的大数据库，推进肝癌病人免疫治疗的研发、筛选，亦为临床精准治疗方案提供了科学依据。

The lack of proper animal models for evaluating the efficacy and safety of immunotherapies discourages the clinical application of immunotherapy for treating hepatocellular carcinoma (HCC). Using HCC patient-derived xenografts (PDX), I revealed that the killing capacity of anti-GPC3 chimeric antigen receptor (CAR) T cells is negatively correlated to the expression level of inhibitory immune molecules in HCC cells, such as PD-L1. This result suggests that immune regulators may play a role to affect immunotherapeutic efficacy. In this project, I propose to establish HCC and immune system co-existed xenografts from the same patient in immunodeficiency mice and characterize this novel humanized model by comparing tissue morphology, tumor antigen expression, transcriptome, and immune cell infiltration of the tumors from the xenografts and the original tumor samples from the patient. I will then use this humanized mouse model to study how HCC suppresses human T cells and assess efficacies of CAR T cells for treating HCC. To evaluate the safety of immunotherapies, I will use the NOD/SCID/IL2rg-/-Fah-/- strain that was generated in our laboratory to establish xenografts, in which mouse hepatocytes are replaced with human normal hepatocytes. I will also establish a bank of HCC PDX models with clinical information of patients that can be shared with research institutions and pharmaceutical companies to develop novel drugs and immunotherapies for treating HCC in the future.