Establishment of a Large-Scale PDX Library of Head and Neck Cancers for Functional Precision Oncology.

Abstract

BACKGROUND: Precision oncology leverages the molecular and genetic characteristics of tumors to enable accurate diagnosis and effective treatment selection. However, recent clinical trials have highlighted the limitations of current approaches and underscored the need to integrate static molecular profiling with functional analyses using patient-derived xenograft (PDX) models-particularly for cancers such as head and neck cancer (HNC), where driver mutations are rare and prognosis remains poor. METHODS: Here, we aimed to establish a large-scale PDX library for HNC, termed the Fujita Xenograft Library (FXeL), annotated with detailed clinical information. Since 2022, tumor specimens from over 100 surgical cases at Fujita Health University Hospital have been transplanted into immunodeficient mice, resulting in the successful establishment of 62 PDX models. RESULTS: Advanced clinical stage was significantly associated with successful engraftment, and serial passaging led to progressively accelerated tumor growth. Comparative analyses of genomic profiles between patient tumors and PDXs demonstrated that major cancer-related mutations were largely preserved in PDXs, while clonal selection and evolution occurred during engraftment. Histopathological features, including keratinization and nuclear atypia, were retained, whereas stromal components such as cancer-associated fibroblasts exhibited compositional shifts. Furthermore, drug sensitivity assays revealed that PDX responses to cisplatin (CDDP) closely mirrored the clinical outcomes of the corresponding patients. CONCLUSIONS: The FXeL represents a robust and scalable platform for investigating HNC biology and therapeutic response. Despite limitations such as stromal remodeling and the absence of an immune microenvironment, these models provide valuable translational insights and support the advancement of functional precision oncology.