[The establishment of head and neck squamous cell carcinoma PDX models and humanized immune reconstruction]

Objective: To construct patient-derived xenograft (PDX) models from head and neck squamous cell carcinoma (HNSCC) patients, to explore the effect of immune reconstitution timing on the PDX modeling and immune microenvironment in humanized immune system mice (huHSC-NCG-hIL15), and to provide a reliable animal model for research on the mechanisms of head and neck squamous carcinoma and for studies on immune therapy drug interventions. Methods: This study enrolled 28 HNSCC patients (25 laryngeal carcinomas, 3 hypopharyngeal carcinomas). PDX models were established in Balb/c nude (nu) mice, NSG mice, and humanized immune system-reconstituted huHSC-NCG-hIL15 mice. Fresh HNSCC samples were transplanted into Balb/c nu and NSG mice to generate PDX models, with subsequent analysis of success-associated factors. One successfully established PDX tumor was subsequently implanted into humanized immune system-reconstituted huHSC-NCG-hIL15 mice. Tumor transplantation was performed at distinct immune reconstruction timepoints (2 vs. 7 weeks post-reconstitution), and tumor growth patterns were monitored. Flow cytometry and multiplex immunohistochemical staining were utilized to characterize immunological profiles in peripheral lymphoid organs and tumor microenvironments. Hematoxylin-eosin (HE) staining was employed to assess histomorphological concordance between primary patient tumors and PDX model tissues. Results: HNSCC PDX models were successfully established. NSG mice exhibited a higher and more stable tumor take rate compared to Balb/c nu mice (pilot study: 4/10 vs. 3/10 cases; mean take rate 60%-80% vs. 20%-60 %). The PDX success rate in NSG mice was 46.4% (13/28). In the huHSC-NCG-hIL15 mice model with immune reconstitution at 7 weeks, tumors grew significantly faster, and the PDX modeling process was shorter (617 mm³ at day 70 in 7-week cohort vs.280 mm³ in 2-week cohort). Flow cytometry analysis of the immune microenvironment showed that at 7 weeks of immune reconstitution, the proportions of B cells in the spleen and tumor tissues(2-week vs. 7-week: spleen 16.2% vs. 61.7%, tumor 26.0% vs. 38.8%) and myeloid cells in the spleen (2-week vs. 7-week: spleen 47.2% vs. 88.1 %) were significantly higher, while mice at 2 weeks post-reconstitution showed a higher proportion of T cells (2-week vs. 7-week: spleen 13.2% vs. 9.3%, tumor 4.8% vs. 2.5%). HE results demonstrated that the tumor tissues in PDX models maintained a high degree of morphological similarity to the primary tumors in both NSG and huHSC-NCG-hIL15 mouse models. Conclusion: The HNSCC PDX modeling protocol demonstrates operational feasibility and high reproducibility, establishing this model as a robust platform for mechanistic and immunotherapeutic studies.