Abstract PS3-13-16: Cancer systems immunology unravels complexity of reversing immune suppression and predicts beyond RECIST in metastatic breast cancer

Abstract Background: Combination therapy aimed at modulating the tumor microenvironment (TME) is a robust strategy to improve response to checkpoint inhibition and broaden indications for use in patients with metastatic breast cancer. The epigenetic modulator, entinostat when combined with dual checkpoint blockade; nivolumab and ipilimumab demonstrated promising results in our Phase Ib trial (NCI-0944) and in our recent meta-analysis (AACR-3225, 2025). Use of a cancer systems immunology approach accelerated discovery of the complex mechanism of response to treatment. Lastly, we demonstrate the potential for math modeling to predict response similar to RECIST, and inform organ specific response, a powerful tool with potential for use toward a more personalized approach for patients. Methods: To identify mechanisms of response to combination therapy in a preclinical model, we used knowledge-guided subclustering of single-cell RNA-sequencing data and cell circuits analysis to predict salient interactions. Multiparametric flow cytometry and ex-vivo immune suppression assays were used to validate preclinical findings. Imaging mass cytometry and bulk RNA sequencing of patient samples were used to validate findings in patients. Derivation of dynamic mathematical models of tumor-immune modulatory responses were then fit through innovative use of both preclinical (sequencing) and clinical (proteomics) data - to predict that a combination of effects on the TME is necessary for response. Adaptation of this mathematical model was then made to determine its potential for use in prediction of site-specific response in metastatic lesions. We introduced methods employing posterior parameter sampling and simulation to create virtual tumor populations, enabling extrapolation beyond the data to predict probabilities of response in metastatic lesions, even when no data exists at that site. Results: In the mouse lung TME, we identified 39 cell states and salient interactions, of which myeloid, T cell, and B cell subpopulations were most affected by treatment in mice bearing metastatic breast cancer treated with combination therapy. Functional immunologic assays verified inhibition of the ICAM pathway in myeloid cells partially recapitulated treatment effects on CD8+ T cells. We also found evidence that treatment increased anti-tumor IgG production. Analysis of patient biopsies via spatial proteomics corroborated preclinical findings: in responders, we observed increased B cell activation, mature tertiary lymphoid structures (TLS), and increased CD8+ T cell—macrophage distances with treatment. We demonstrated clinical utility of the mathematical model via Bayesian parameter inference with clinical responses measured by RECIST. This revealed that only the immunosuppression parameters were predictive of response; parameters controlling cytotoxicity were uninformative. We also show that the model can predict response at sites that have yet to develop disease—a tool which could be considered for future trials to predict overall response rates based on mechanism of drug response. Conclusions: We conclude that epigenetic modulation via HDACi induces a carefully orchestrated set of changes in plasma cells and CD8 T cells with MDSCs and macrophages to sensitize the TME to checkpoint inhibition. Significant changes in TLS formation and macrophage -T cell interactions in biopsies from patient responders validate findings. More broadly we provide a framework for the discovery of cell-cell interactions that control responses in complex TMEs. We also demonstrate how interdisciplinary data integration fuels this new field of cancer systems immunology to accelerate discovery of mechanisms of successful immunotherapeutic response in previously unresponsive solid tumor types. Citation Format: E. T. Roussos Torres, E. Gonzalez, J. Kreger, Y. Liu, X. Wu, A. Barbetta, A. G. Baugh, B. Al-Zubeidy, J. Jang, S. M. Shin, Z. M. Zhang, V. Stearns, R. M. Connolly, W. Jin Ho, J. Emamaullee, A. L. MacLean. Cancer systems immunology unravels complexity of reversing immune suppression and predicts beyond RECIST in metastatic breast cancer abstract. In: Proceedings of the San Antonio Breast Cancer Symposium 2025; 2025 Dec 9-12; San Antonio, TX. Philadelphia (PA): AACR; Clin Cancer Res 2026;32(4 Suppl):Abstract nr PS3-13-16.