Abstract 5634: The single mouse trial format predicts the sensitivity towards checkpoint inhibitor treatment in NSCLC PDX models in HuCD34NCG/CRL mice, an innovative triple immune deficient mouse strain

Abstract

Abstract Immunotherapy is revolutionizing cancer treatment. As a consequence, there is an urgent need for predictive preclinical models to support compound development with the aim of bringing optimal drug candidates into clinical trials. Patient-derived xenografts (PDX) are the gold-standard in cancer drug development. One major drawback of PDX is the lack of an immunological competent host. To overcome this hurdle triple immunodeficient mice are humanized with hematopoietic stem cells (HSC) to enable the interaction between human immune cells and human tumor cells in a rodent host. The current study uses HSC humanized NCG (NOD-Prkdcem26Cd52Il2rgem26Cd22/NjuCrl) mice, HuCD34NCG, to evaluate the sensitivity of five NSCLC PDX models towards anti CTLA-4 (Ipilimumab), anti-PD-1 (Nivolumab) and the combination thereof. We identified one predicted responder and one non-responder and in a single mouse trial approach (SMT) and confirmed the results in a conventional experimental set-up with n=5 per treatment arm. Infiltration of human immune cells was detected by flow cytometry in tumor tissue (TIL) and hematopoietic organs. Serum of tumor bearing animals was analyzed for human and mouse cytokine secretion pre- and post-treatment. The NCG mouse strain was genetically characterized by whole exome sequencing (WES) and the mutational profile compared to other common mouse strains. The mutational profile showed the highest overlap with the genome data from other triple immune compromised mice and with NOD/SCID indicating their close hereditary relationship. Non-tumor bearing HuCD34NCG mice depicted engraftment of human CD45+ cells in peripheral blood of >25% and >60% in bone marrow and spleen after 8-12 wks post injection. Thus, they fulfilled enrollment criteria for conducting a drug screen for immune modulating compounds. The five NSCLC PDX models showed a distinct sensitivity profile ranging from partial remission (tumor growth inhibition, TGI, of 87.5% compared to untreated control) to progression (no TGI). In all models the combination treatment was the most active therapy (TGI: 60%), followed by Ipilimumab (TGI: 55%) and Nivolumab (TGI: 48%). Nevertheless, in selected models anti-CTLA-4 treatment was the most efficacious therapy. The two models examined in more detail LXFE 397 (=responder, TGI 55% in both formats) and LXFA 400 (non-responder, TGI 2%SMTvs 5%, n=5/group) displayed the same sensitivity pattern in the conventional set-up as compared to the SMT screening approach. The TIL analysis revealed a pronounced rise in the CD3+ subset and increased CD4+/CD8+ ratio in the most effective treatment arms. Subsequent serum analyses will help identify possible biomarkers predicting drug response towards checkpoint inhibition. Taken together our study proves that the HuCD34NCG mouse is a fully functional drug development tool. In combination with highly predictive PDX tumor models this in vivo platform provides a further step to support the development of new drugs targeting the host immune response. Citation Format: Julia Schüler, Jenny Rowe, Eva Oswald, Stephen Festin. The single mouse trial format predicts the sensitivity towards checkpoint inhibitor treatment in NSCLC PDX models in HuCD34NCG/CRL mice, an innovative triple immune deficient mouse strain [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 5634.