Advancing Education, Research, and Quality of Care for the Head and Neck oncology patient.
Introduction: Patient-Derived Xenograft (PDX) is becoming increasingly important in cancer research. PDX involves transplanting a patient's tumor directly into mice, creating an experimental model that closely mimics patient tumors. This allows for the investigations of tumor behavior and response to therapeutic treatments in vivo, closely replicating clinical conditions. Availability of PDX models for head and neck cancers, including rare tumors in this region, remains limited compared to other cancer types. Additionally, addressing chemotherapy resistance—a major obstacle in head and neck cancer treatment, especially adaptive resistance, where tumors adapt to and resist treatment—requires PDX models that can accurately replicate patient tumors. The mechanisms of resistance and overcoming them remain to be fully elucidated, in part because of a lack of suitable experimental models for head and neck cancers.
Methods: We transplanted surgical specimens from head and neck cancer patients into highly immunocompromised mice, generating a library of PDX across various cancer types, including rare cancers. By performing genome profiling on both patients and the established PDX models, we analyzed genetic alterations present in each. Additionally, through pathological analysis, we investigated the histological concordance and tumor microenvironmental components, such as cancer-associated fibroblasts, between patient tumors and PDX tumors. Furthermore, we treated the established PDX models with cisplatin, a key drug in head and neck cancer treatment, to test sensitivity. We isolated RNA from PDX models over time after cisplatin treatment to perform gene expression analyses, elucidating the molecular mechanisms underlying chemotherapy resistance.
Results: We successfully established PDX models at a remarkably high efficiency of approximately 80% (44 of 52 cases), achieving success across various head and neck cancers including oral cavity, hypopharynx, larynx, oropharynx tumors and rare cancer salivary gland carcinoma. No preferences were observed in age, sex, sampling site, recurrence, or cancer stage. During successive passages, the tumor growth rate significantly accelerated, likely due to adaptation to the murine environment or increased malignancy. Pathological analyses demonstrated that key tumor characteristics, such as histological type and composition of cancer-associated fibroblasts, were preserved in PDX models. The genome profiling for original patients identified gene mutations in tumor suppressor genes, such as TP53 (72%) and CDKN2A (67%), without notable driver genes detected. This trend was also conserved in PDX models, but PDX lost the number of mutations from patient tissues and acquired unique mutations during tumor growth. Furthermore, upon treatment with cisplatin, the established PDX models exhibited varying responses, categorizing them as sensitive or insensitive to the drug, which is like the clinical outcome. Temporal gene expression analysis showed that in resistant PDX models, genes related to cytoskeletal reorganization, extracellular matrix formation, and cellular adhesion were upregulated, suggesting their critical roles in acquired therapy resistance.
Conclusion: We have established a highly efficient method for creating PDX models and developed a PDX library that includes rare cancers. Using the research platform, our study also provides the insight into the acquired cisplatin resistance. These PDX models will be valuable for elucidating disease mechanisms, overcoming treatment resistance, and developing new therapeutic approaches.