Background
Tumor-bearing murine models are frequently used in head and neck cancer research. Subcutaneous heterotopic tumors are often utilized for their simplicity of implantation and ease of tumor access. However, orthotopic models, implanted into relevant organ-specific environments, are widely considered more translationally relevant. While both models may be used to study head and neck squamous cell carcinomas (HNSCC), few studies have directly compared the tumor immune microenvironments (TIME) of orthotopic and heterotopic tumors. Moreover, tumor draining lymph nodes (tdLNs), sites critical for antigen presentation and T cell activation, are commonly resected in patients with HNSCC. Recent studies have increasingly shown key implications of the loss of tdLN in anti-tumor response. Herein, we investigated locoregional TIME differences in a carcinogen-induced, HPV-negative preclinical oral cancer model, unresponsive to traditional immunotherapy. We hypothesize that there will be no differences in immune cell populations between heterotopic and orthotopic sites in this model.
Methods
ROC1 cells were maintained as published. ROC1 tumors were established in the murine flank and oral cavity, and tumor growth kinetics were assessed at each site. At distinct stages of tumor growth, tumors were harvested, as well as their respective corresponding inguinal and cervical tdLNs. Multi-parameter 28-marker spectral flow cytometry was performed to analyze immune cell populations at each site.
Results
Both flank and oral cavity tumor sites displayed an initial period of delayed tumor growth followed by rapid tumor progression. Comprehensive analyses revealed low T cell infiltration in both tumor models, with greater increases over time in most cell types in the flank compared to the oral cavity. In both models, increases in myeloid cell types over time, specifically non-M1 macrophages and conventional type 2 dendritic cells (DCs) were observed. Inguinal tdLNs had notable increases in regulatory T cells and macrophages, while cervical tdLNs had greater numbers of DCs. Immune checkpoint marker (programmed cell death protein 1, cytotoxic T-lymphocyte associated protein 4, programmed death-ligand 1) expression also increased over time.
Conclusions
Our immunophenotyping results indicate similar ‘cold’ immune profiles between ROC1 TIMEs in the flank and oral cavity, confirming our hypothesis. However, observed differences between sites highlight tissue-specific TIME differences may impact anti-tumor treatment and response. Moreover, differences between corresponding tdLNs indicate changes in immunosuppressive phenotypes, which may further impact response. Our understanding of critical immune microenvironment differences between ROC1 orthotopic and heterotopic models will enable tailoring of future therapeutic strategies, and furthermore, provide insight into model selection and data interpretation from translational studies.