Abstract 1825: QSP modeling to inform dose and regimen selection for TAK-280: A bi-specific antibody.

Abstract Background: CD3-targeting T cell engagers (TCE) have been approved as therapy for various hematological and solid tumor oncology indications; however, they can often induce unwanted safety issues, including cytokine release syndrome (CRS) and liver enzyme elevations. TAK-280 is an investigational B7-H3 x CD3ε conditional bispecific redirected activation (COBRA) TCE. In its masked prodrug form, TAK-280 binds to B7-H3, but not to CD3ε. Once in the protease-rich tumor microenvironment, proteases mediate cleavage of the linker resulting in unmasking the prodrug, which allows for the formation of active CD3ε-binding dimers and ultimately CD3 T-cell activation and cytotoxic anti-tumor response against co-engaged B7-H3 expressing cells. We developed a mechanistic PK/PD modeling framework to describe the PK and safety of TAK-280, based on clinical data from a Phase 1 dose escalation study (NCT05220098) investigating TAK-280 for the treatment of patients with advanced, unresectable or metastatic cancer, with a focus on tumor types with enhanced B7-H3 expression. Methods: In the tumor compartment of the model, TAK-280 binds to B7-H3 expressed on tumor cells. Once the dimer is formed, it binds to CD3 to form the desired trimer complex composed of TAK-280, B7-H3 and CD3. It is assumed that this trimer complex is the only species that triggers cytokine release in the tumor compartment. Once cytokines are secreted in the tumor, they are assumed to move to systemic circulation. In the central compartment, cytokines undergo basal production and degradation. Results: A two-compartment model with linear elimination was sufficient to describe the PK. Cytokine dynamics were described using a three-population immune cell pool1 (non-secreting, secreting and refractory) framework. Immune cell dynamics reveal that the cytokine Cmax attenuation over time is a consequence of immune cell de-sensitization by transitioning to the refractory state. This concept of de-sensitization or decreased cytokine release following TCE dosing is clinically well established and mechanistically it is attributed to fewer non-secreting immune cells that are available by the time the top dose is administered (in a lead-in dosing regimen). Conclusion: Using the model, we were able to show that lead-in dosing led to lower cytokine peak concentrations and therefore predicted lower probability for an adverse event. Model simulations demonstrated comparable safety outcomes for both QW and BIW dosing of the bispecific antibody. References: 1.Weddell et. al. CPT Pharmacometrics Syst Pharmacol. 2023; 12:1726-1737 Citation Format: Agnish Dey, Tao Long, Sabrina Collins, Dean Bottino, Jaydeep Srimani, John Gibbs. QSP modeling to inform dose and regimen selection for TAK-280: A bi-specific antibody abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1825.