The rise in checkpoint inhibitor (CPI) therapies has revolutionized oncology and informed on the mechanisms regulating natural antitumor immune responses in patients. However, because immunotherapy often results in immune-related adverse events (irAEs), CPIs have also revealed the key role that immune checkpoint receptors play in the active maintenance of immunologic tolerance towards self antigens. irAEs target a variety of organ systems and can have long-term, potentially fatal consequences for treated patients. Little is known about why irAEs occur, how to predict irAEs, or whether it is possible to block irAEs while sustaining CPI-induced antitumor effects. One of the challenges for therapeutically investigating these open questions has been the lack of suitable animal models for systematically studying irAEs in different organs. To address this, we have genetically engineered an animal model called NINJA (iNversion Inducible Joined neoAntigen), which bypasses T-cell tolerance mechanisms and allows us to induce the expression of known neoantigens de novo. Induction of neoantigens in healthy organs with NINJA is achieved by tissue-specific Cre recombinase expression and subsequent administration of doxycycline and tamoxifen (dox/tam), which expresses neoantigens in the Cre+ cells. This allows us to use either tissue-specific infection with Cre-expressing adenoviruses or tissue-specific promoters expressing Cre to control the organ that will express neoantigen after dox/tam treatment. Moreover, NINJA allows us to compare how T cells respond against antigens when presented in different contexts, from self antigens in a peripheral tissue, to neoantigens expressed during a viral infection or developing cancer. In the context of viral infection, the function of neoantigen-specific effector CD8 T cells is highly dependent on the organ expressing the neoantigen, with robust, functional effector T cells observed when antigen is induced in skin or muscle and weak, low-functioning PD-1+ T cells when antigen is induced in liver or lungs. Similarly, in absence of inflammation, neoantigen induction in hepatocytes results in rapid and dominant systemic CD8 T-cell tolerance, while induction in skin leads to effector CD8 T-cell responses. Crucially, in the latter context, CPI treatment results in severe skin irAEs. Finally, because NINJA is on a non-autoimmune-prone C57Bl/6 background, we are testing whether one can block CPI-induced irAEs while allowing CPI-induced antitumor responses. In summary, the NINJA model is demonstrating that tissues have a strong bearing on T-cell function regardless of whether antigens are encountered in the context of infection or as self antigens. Furthermore, there appears to be a hierarchy, where certain tissues (i.e., skin) are programmed to provide stronger immune responses than in others (i.e., liver). This might explain why certain tissues are prone to the development of T cell-mediated irAEs after CPI treatment.

This abstract is also being presented as Poster B1.

Citation Format: Martina Damo, Ivana William, Nikhil S. Joshi. Investigating the mechanisms of immune-related adverse events using engineered animal models [abstract]. In: Proceedings of the AACR Special Conference on Tumor Immunology and Immunotherapy; 2019 Nov 17-20; Boston, MA. Philadelphia (PA): AACR; Cancer Immunol Res 2020;8(3 Suppl):Abstract nr PR5.