Abstract B29: Preclinical model to investigate the impact of intratumoral heterogeneity for DNA mismatch repair on colon tumor treatment response

Abstract

Abstract Conventional preclinical mouse models that are typically used to test treatment regimens fail to accurately recapitulate intratumoral heterogeneity in human cancers, often leading to inaccurate predictions of efficacy in the clinic. Though early clinical trials for immunotherapies in metastatic colorectal cancers deficient for DNA mismatch repair (dMMR) have shown promising results, dMMR status alone is not a predictor of successful immunotherapeutic response. If a colon tumor has a mixture of dMMR and proficient MMR (pMMR) tumor cells, some tumor cells might not respond to immunotherapy, leading to resistance. A preclinical model of colon cancer with mixed MMR status is needed to better understand how intratumoral heterogeneity with respect to dMMR impacts tumor response to treatment. We developed a new mouse model containing a mosaic colon consisting of both dMMR cells labeled with green fluorescence and pMMR cells labeled with red fluorescence. To induce colonic tumors, the mutagen azoxymethane was administered. The tumors are easily visualized with bright light or fluorescent colonoscopy to determine tumor size and cellular composition before, during, and after immunotherapy. Surveillance colonoscopy is used to assess tumor development and treatment response in real-time by measuring the change in tumor size and the proportion of green (dMMR) and red (pMMR) tumor cells. Mice develop colon tumors with three different cellular compositions: homotypic dMMR or immunologically ‘hot’ tumors that are predicted to have the best response to immunotherapy; homotypic pMMR or immunologically ‘cold’ tumors that are predicted to have the worst response to immunotherapy; and heterotypic tumors with a mixture of dMMR and pMMR cells or immunologically ‘lukewarm’ tumors that are predicted to have a partial response to immunotherapy. Heterotypic MMR tumors ranged from as few as 5% dMMR (green) cells to as many as 95%. The tumor response to systemic immunotherapy alone, aPD-L1, was dependent on the tumor composition. Large dMMR tumors following aPD-L1 therapy appeared to have large areas of necrosis in wholemount, whereas pMMR tumors following treatment were indistinguishable from pMMR tumors in controls. Mixed MMR tumor response varied depending on percentage of dMMR cells; tumors with a high percentage of dMMR cells remained small after aPD-L1 whereas tumors with a low percentage of dMMR had an outgrowth of pMMR cells after aPD-L1. Our novel mouse model allows for the highly detailed characterization of tumor response, especially considering the clonal composition of a tumor and should greatly facilitate identifying mechanisms of resistance. The capability to investigate complex tumor architecture in regard to tumor response to treatment offers a unique opportunity that would not be possible with current preclinical models in which tumors are homogenous by design. With more accurate preclinical models to recapitulate complex human disease, we can translate treatment regimens into the clinic with greater efficiency. Citation Format: Santina Snow, Rich Halberg, Dawn Albrecht, Maggie Baus, Shane Huebner, Zach Morris, Jamey Weichert, Paul Clark. Preclinical model to investigate the impact of intratumoral heterogeneity for DNA mismatch repair on colon tumor treatment response [abstract]. In: Proceedings of the AACR Special Conference: Tumor Immunology and Immunotherapy; 2022 Oct 21-24; Boston, MA. Philadelphia (PA): AACR; Cancer Immunol Res 2022;10(12 Suppl):Abstract nr B29.