CELL INTELLIGENCE · 肿瘤细胞治疗研究

Novel preclinical models that better mimic the in vivo tumor microenvironment are essential to advance understanding of tumor biology and resistance/response to therapy. Herein, we report development of a novel ex vivo patient-derived three-dimensional lung tumor model (3D-LTM) for use in evaluating response to therapy. With this model system that maintains cell-cell interactions and tissue architecture, we observed heterogeneity of response to immune checkpoint inhibitors (ICI), as noted in non-small cell lung cancer (NSCLC) patients, and defined gene signatures associated with response. Spatial transcriptomics identified positive correlation of CD8 + T cell populations, CD4 + memory T cells, mast cells, NK cells, naive B cells, endothelial cells and non-classical monocytes with response status, and negative correlation of macrophages with response status. Pathway analysis of gene expression showed that chemokine signaling related pathways were activated in responder 3D-LTM tissues, whereas suppression of antigen presentation-related pathways and activation of T reg differentiation-related pathways were associated with non-responder 3D-LTM tissues. Additionally, the abundance of dividing T cells and naive CD8 + T cells differentially correlated with T cell cytotoxicity gene signatures based on response status. Thus, this model may provide utility for rapid testing of therapeutic outcomes and biomarker development.