A32 Evaluation of the Mutant KRAS-Driven NSCLC Tumor Immune Microenvironment Using Patient-Derived Cell Line Xenografts in a Humanized Mouse Preclinical Model for Development of New Immunotherapy Approaches

Abstract

Tumor cell and tumor microenvironment (TME) features that influence the response to immune checkpoint blockade in lung cancer are incompletely defined. We wanted to develop preclinical models of lung cancer to explore the relationship of non-small cell lung cancer (NSCLC) molecular characteristics (oncogenotype such as mutant KRAS and LKB1/STK11, and mRNA expression profiles) to the human TME. We exploited the use of 8 (all KRAS mutant) molecularly characterized (whole-exome seq mutation and RNAseq analyses) patient-derived NSCLC lines grown as xenografts in vivo in humanized NSG-SGM3 mice in an effort to determine the effect of different NSCLCs on the immune landscape of lung cancer xenografts. The triple transgenic NSG-SGM3 (NSGS) mice express human IL3, GM-CSF, and SCF, which combine the features of the highly immunodeficient NOD scid gamma (NSG) mouse with cytokines that support the stable engraftment of human myeloid lineages and regulatory T-cell populations. Subcutaneous KRAS mutant-driven non-NSCLC xenografts grown in NSG-SGM3 mice “humanized” with CD34+ cord blood cells were subjected to immune landscape analysis through flow cytometry, multiplex immunohistochemistry, and cytokine analysis. The xenografts were also treated with various combinations of checkpoint inhibitors, radiation, and activation of the innate immune pathway with emricasan (pan caspase inhibitor). Our results show: that the 8 KRAS mutant NSCLCs each were associated with a different spectrum of human TME; 3 of the 8 xenografts had only 0.1% of immune cell infiltrates while other xenografts had 20% immune cell infiltrates; that tumor mutation burden (TMB, absolute mutation calls 233 to 2,076) is not predictive of CD8+ T-cell infiltration in KRAS mutant-driven NSCLC xenografts. In NSCLC xenografts with high CD8+ T-cell infiltrate, the CD8+ cells in the TME were not activated, resulting in limited responses to PD-1/PD-L1 therapy. However, stimulation of the cGAS-STING innate immune pathway with emricasan followed by radiation (15 cGy) resulted in dramatic antitumor response. NSCLC xenografts grown in “humanized” mice show great intertumor heterogeneity effects on the TME even within the KRAS mutant subgroup, and it is possible to demonstrate targeted therapy such as emricasan/radiation can lead to changes in improved antitumor responses.