Abstract 4916: Spatial analysis of local drug induced changes in tumor microenvironment predicts effective treatment combinations in breast cancer

Abstract Anticancer therapeutics primarily designed to target tumor intrinsic mechanisms may also affect components of tumor microenvironment (TME) - immune cells and non-immune stroma. Recent literature reenforces the concept that complex interactions between drugs, neoplastic cells and cells of TME determine the efficacy of anticancer therapies. Systems understanding of these interactions can serve to predict effective treatment combinations simultaneously attacking tumor cell vulnerabilities, enhancing immune surveillance, and mitigating stromal mediators of resistance. The key challenge is to find such TME-modulating combinations in a fast and more informative way. We have developed an integrated analytical platform termed Multiplex Implantable Microdevice Assay (MIMA) to rapidly decompose cancer complexity in drug response and find biomarkers with predictive value for combination therapy efficacy including immunotherapy efficacy. The system deploys a (i) miniaturized implantable microdevice for localized intratumoral drug delivery and (ii) multiplex immunostaining to measure 30+ proteins in single cells at each drug well. Computational analyses of local drug-induced changes provide information about the composition, functional state and spatial cell organization of the tumor and associated TME ultimately bringing new insights into drug mechanisms of action. We used MIMA in genetically engineered mouse models of breast cancer to evaluate effects of five targeted anticancer agents (olaparib, palbociclib, venetoclax, panobinostat, lenvatinib) and two chemotherapies (doxorubicin, paclitaxel) and predicted synergistic antitumor effects with anti-PD-1, anti-CD40, anti-CSF1R immunotherapies and vasculature modulating agents. Some of the most effective combinations were not reported before. A pan-HDAC inhibitor, panobinostat, that synergized with anti-PD-1, induced immunogenic cell death and infiltration of antigen presenting neutrophils. Further longitudinal spatial analyses revealed that mechanisms of resistance co-emerged with these response phenotypes and became prominent over time. We measured fibroblast, protumorigenic macrophage and cytotoxic B cell recruitment associated with heavy deposition of collagen VI, increased immune suppression and emergence of invasive and cancer stem cells. Combination with stroma modulating agent, losartan, improved the efficacy of panobinostat/anti-PD-1 in systemic studies implying that normalization of non-immune stroma might favorably alter aspects of TME for immune and targeted therapy efficacy. All in all, MIMA may represent a new approach to predict effective combination regimens for individual cancer patients. Extended MIMA use and computational modeling of the spatial cell patterns could provide actionable information for development of effective drug doses and schedules. Citation Format: Zuzana Tatarova, Dylan C. Blumberg, Gordon B. Mills, Lisa M. Coussens, Oliver Jonas, Joe W. Gray. Spatial analysis of local drug induced changes in tumor microenvironment predicts effective treatment combinations in breast cancer. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 4916.