ECM architecture drives distinct evolutionary patterns of tumor immune escape and elimination in breast cancer

Introduction: Cancer progression poses a clinical hurdle due to tumor heterogeneity that contributes to treatment resistance and immune evasion. While the adaptive immune system can effectively eliminate cancer in certain instances, tumor escape remains a challenge [1-2]. The immune microenvironment plays an important role in this process. One unanswered question is the role of ECM geometry and its effects on tumor immune recognition. In solid malignancies, ECM geometry has been associated with disease stage and T cell infiltration [3-4]. Specifically, empirically observed ECM topologies are frequently categorized based on fiber arrangement: random fibers (TACS1), circumferentially aligned fibers (TACS2), and radially arranged fibers (TACS3) [3]. Even though a clear negative correlation between TACS and patient survival has been established, the specific roles and extent of TACS in T cell-driven cancer evolution remain uncharacterized [4]. The precise mechanisms underlying how TACS influences cell movement are still not fully elucidated, with divergent opinions on whether and how the ECM mediates immune cell infiltration [5]. At present, we currently lack a physical model relating the impact of TACS on the spatial co-evolution between an adaptive immune repertoire and a heterogeneous population of evading cancer cells.

Here, we developed an agent-based model and provided a detailed dynamical description of the role of TACS in tumor evolution when subject to adaptive immune selective pressure. We anticipate that its use can be more broadly applied to understand cancer evolutionary patterns and treatment success or failure in specific cases where observed TACS architecture and phenotypic status are previously defined.

Materials and

Methods: We employ largescale Gillespie based simulation to study tumor and T cell behaviors. Cancer cells in our model undergo division, migration, and express a variable number of tumor-associated antigens (TAAs). Mutations can lead to different cancer clones. Migration occurs along nearby fibers. In epithelial-mesenchymal transition (EMT) models, edge cells divide less, migrate faster, and decrease immunogenicity. CD8+ T cells migrate towards tumor cells, influenced by collagen fiber alignment and chemotaxis. Different T clones possess varying antigen recognition abilities through distinct T cell receptors (TCR). Migrating T cells eliminate recognized tumor cells, divide, and have a finite survival window. The ECM surrounding the tumor is simulated with initially randomly oriented fibers (TACS1). Fiber lengths follow a normal distribution, with density decreasing outward from the tumor. Dividing tumor cells at the edge can alter fiber orientation from TACS1 to TACS2 or TACS3.

Results, Conclusions, and Discussions: Our analysis of tumor-T cell interactions across three TACS types unveils distinct migration patterns (Fig 1A), with TACS3 displaying the highest efficiency, TACS1 intermediate, and TACS2 the lowest (Fig 1B-C). TACS exert a stronger influence on T cells than on tumors, with enhanced chemokine gradients facilitating T cell infiltration (Fig 1D-E). Despite TACS2's limited migration efficiency, it does not entirely impede T cell infiltration. Varied migration efficiencies yield diverse T cell infiltration patterns and levels of immunoediting. Evaluation of T cell infiltration in TACS3- predicts a decline in immunogenic tumor clones due to immune pressure (Fig 1F-H). In our model, while TACS3 showed an overall benefit to immune recognition and higher survival, this contradicted previous findings of lower survival in TACS3+ patients [4]. To reconcile this discrepancy, we introduced EMT. Our results indicated that only after TACS2-TACS3 progression did our model closely match clinically observed outcomes (Fig 2 A-C). Considering mesenchymal tumor cells' ability to upregulate PD-1/PD-L1 expression, we further explore how TACS-specific tumor evolution impacts responses to PD-1/PD-L1 inhibitors and patient survival. We consistently observe higher tumor escape rates or lower survival rates in TACS3 compared to TACS2. In summary, we conclude that TACS3 alone cannot account for the clinically observed lower survival in breast cancer; TACS3-associated late-stage cancer byproducts such as EMT and elevated checkpoint expression significantly decrease patient survival.

Acknowledgements (Optional): JTG is a CPRIT Scholar in Cancer Research. This work was supported by CPRIT (RR210080).

References
1. George, J. T., & Levine, H. (2018). Stochastic modeling of tumor progression and immune evasion. Journal of theoretical biology, 458, 148-155.
2. George, J. T., & Levine, H. (2020). Sustained coevolution in a stochastic model of cancer–immune interaction. Cancer research, 80(4), 811-819.
3. Provenzano, P. P., Eliceiri, K. W., Campbell, J. M., Inman, D. R., White, J. G., & Keely, P. J. (2006). Collagen reorganization at the tumor-stromal interface facilitates local invasion. BMC medicine, 4, 1-15.
4. Conklin, M. W., Eickhoff, J. C., Riching, K. M., Pehlke, C. A., Eliceiri, K. W., Provenzano, P. P., ... & Keely, P. J. (2011). Aligned collagen is a prognostic signature for survival in human breast carcinoma. The American journal of pathology, 178(3), 1221-1232.
5. Li, X., Gruosso, T., Zuo, D., Omeroglu, A., Meterissian, S., Guiot, M. C., ... & Levine, H. (2019). Infiltration of CD8+ T cells into tumor cell clusters in triple-negative breast cancer. Proceedings of the National Academy of Sciences, 116(9), 3678-3687.