食管癌（EC）是全球第十常见的癌症，具有较高的发病率和死亡率。其主要亚型包括食管鳞状细胞癌和食管腺癌，通常在晚期被诊断出来。生物缺陷、临床前模型无法完全总结多因素的病因、肿瘤微环境的复杂性和肿瘤的遗传异质性严重限制了EC的临床治疗。EC患者来源的模型不仅保留了原始肿瘤的组织结构、细胞形态学和分化特征，还保留了肿瘤的异质性。因此，与其他临床前模型相比，它们能更好地预测候选药物的疗效，探索新的生物标志物，与临床试验结合，并有效改善患者预后。本综述讨论了建立患者来源模型和基因工程小鼠模型，特别是患者来源异种移植模型的方法和动物。它还讨论了它们作为临床前实验研究工具的优势、应用和局限性，为食管癌的精确个体化治疗提供重要参考，并改善患者预后。©2023年。作者，独家授权给Springer Nature America公司。

Esophageal cancer (EC) is the tenth most common cancer worldwide and has high morbidity and mortality. Its main subtypes include esophageal squamous cell carcinoma and esophageal adenocarcinoma, which are usually diagnosed during their advanced stages. The biological defects and inability of preclinical models to summarize completely the etiology of multiple factors, the complexity of the tumor microenvironment, and the genetic heterogeneity of tumors severely limit the clinical treatment of EC. Patient-derived models of EC not only retain the tissue structure, cell morphology, and differentiation characteristics of the original tumor, they also retain tumor heterogeneity. Therefore, compared with other preclinical models, they can better predict the efficacy of candidate drugs, explore novel biomarkers, combine with clinical trials, and effectively improve patient prognosis. This review discusses the methods and animals used to establish patient-derived models and genetically engineered mouse models, especially patient-derived xenograft models. It also discusses their advantages, applications, and limitations as preclinical experimental research tools to provide an important reference for the precise personalized treatment of EC and improve the prognosis of patients.© 2023. The Author(s), under exclusive licence to Springer Nature America, Inc.