(595c) Evaluating the Role of Antibody-Drug Conjugate Immune System Effects and Payload Sensitivity in Low HER2+ Solid Tumors Via Agent-Based Modeling

Antibody-drug conjugates (ADCs) have had dramatic success in the clinic, including the HER2-targeting ADC trastuzumab deruxtecan (T-DXd) that was just approved for any HER2+ tumor. Despite this progress, the relationship between the multiple mechanisms of action of ADCs, including payload killing and activation of tumor immunity, is poorly understood. Clinical studies are often limited by small sample sizes of matched biopsies before and after treatment, so far limiting conclusions. For in vivo studies, ADC efficacy is typically measured in immunodeficient mice, which is unable to capture immune effects that may be occurring in the clinic. Additionally, many ADC payloads show lower potency in mouse cells relative to human cells, underestimating payload cell killing in immunocompetent mouse models. Previous studies with immunocompetent mice showed an increase in ADC efficacy versus immunodeficient mice, but these have not been able to discern the contributions of different mechanisms of action, particularly in low expression levels. Yet, this information is critical for designing the next generation of ADCs for lower expression targets. Using a validated hybrid agent-based computational model SimADC, we added immune cell functionality, specifically CD8+ T cell killing, to quantify the relationship between payload efficacy and immune effects. Although tumors grow more slowly in immunocompetent mice, we find that T cell activity increases above baseline levels with ADC administration, indicating a synergistic effect between ADC treatment and immune cell killing rather than additive effect. By training the model on experimental data sets using different payloads, we also find that this activity is different between microtubule-inhibiting payloads and DNA-alkylating payloads. Notably, the immune effects are larger in magnitude than direct payload killing once the target HER2 expression drops below 10,000 HER2/cell, which is considered low expression. Simulation tools like SimADC can help address questions currently intractable with animal experiments and reduce the cost and time of development to design more efficacious ADCs and dosing regimens for cancer patients.