心脏毒性是抗肿瘤药物治疗失败的主因之一。预测肿瘤患者的心脏毒性风险并据此优化用药方案是控制药物心脏毒性的关键。人诱导多能干细胞分化心肌细胞的出现为肿瘤患者提供了一种心肌细胞药物敏感性的体外测试方法，使药物心脏毒性的精准预测成为可能。但该方法的应用仍面临巨大挑战：基于离体细胞实验结果如何预测在体心脏毒性风险？根据本课题组的工作基础，TD模型可对体外研究中药物浓度与心肌细胞损伤的关系进行定量，PBPK模型能预测机体心肌细胞药物浓度，两者结合可评估药物对心肌组织的损伤，心肌组织损伤后在心脏代偿作用调节下的心功能改变可用QSP模型描述。因此我们提出QSP-PBPK-TD模型可解决心脏毒性的体外-体内转化问题。本课题以多柔比星等4种不同作用机制的心脏毒性药物作为工具药，基于细胞、动物以及虚拟人群实验和建模仿真的方法开展抗肿瘤药物心脏毒性的体内外转化研究，期望籍此建立抗肿瘤药物心脏毒性的精准预测方法。

Despite the successes of anticancer drugs, drug-induced cardiotoxicity emerged as a new challenge for cancer-directed therapies. To prevent drug-induced cardiotoxicity, health care providers are recommended to evaluate the risk of cardiac dysfunction before beginning therapy and then adjust the therapeutic strategy. Thus, precise prediction of cardiotoxicity appeared as an urgent task for cardio-oncology, a new filed to prevent the cardiovascular side effect of anticancer therapy. Recently, the development of human pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs) creates the opportunity for a “personalized” approach to assess the sensitivity of cardiomyocytes to the cardiotoxic drug. But there is still a scientific gap on how to translate the in vitro data to human clinical outcomes. According to our previous studies, toxicodynamic (TD) model could be used to estimate the concentration-toxicity relationship based on in vitro data and population physiologically based pharmacokinetic (PBPK) model facilitate the prediction of drug concentration in cardiomyocytes. The combination of PBPK model and TD model could help to predict the myocardial injury. The structural, neurohumoral, cellular, and molecular mechanisms could be activated by cardiac injury to maintain the cardiac function, which is defined as compensatory mechanism. To precisely predict the cardiac outcome of anticancer therapy, the compensatory effect should be considered beyond cardiac injury. A quantitative systems pharmacology (QSP) model has been developed to describe the compensatory mechanism in our previous study. Thus, we believe the QSP-PBPK-TD model is a potential approach for the in vitro-in vivo translation of cardiotoxicity. In this project, we will investigate the cardiotoxicity of anticancer drugs with different toxic mechanism (doxorubicin, 5-fluorouracil, sunitinib and trastuzumab) in in vitro, animal and virtual population levels. Data from these studies was used to develop and validate the QSP-PBPK-TD model for in vitro-in vivo translation of drug-induced cardiotoxicity. Finally, the ultimate goal of this project is to develop a precision approach for the prediction of anticancer-induced cardiotoxicity based on QSP-PBPK-TD model and hiPSC-CMs.