(T-009) Development of QSP platform model for predicting clinical efficacy and CRS incidence of CD3 bispecifics in STEAP1 Prostate Cancer

Goutam Nair, NA – Senior Scientist, Vantage Research Inc; Bhairav Paleja, NA – Lead Scientist, Vantage Research Inc; Deebarshi Mitra, NA – Scientist, Vantage Research Inc; Seshasai Chakravarthy, NA – Lead Scientist, Vantage Research Inc

Introduction: CD3 bispecifics are engineered antibodies that bind CD3 on T cells and target antigens on cancer cells (TAAs), activating T cells to kill cancerous cells. However, they may induce adverse effects such as CRS (Cytokine Release Syndrome) from on-target off-tumor toxicity, prompting ongoing research to define the optimal therapeutic balance. Research aimed at optimizing CD3 bispecifics for cancer therapy involves integrating mechanistic modeling with experimental data to comprehensively elucidate the drug's mechanism of action (effected via the bispecific – TAA – CD3 trimers) in tumor and normal tissues, emphasizing tumor-associated antigen (TAA) receptor occupancy as pivotal in determining both efficacy and potential adverse events like cytokine release syndrome (CRS).

Objectives: The primary goal of developing the CD3 bispecifics QSP platform model is to understand the physiological drivers of efficacy and adverse events and apply that to recommend an efficacious clinical dose, while proposing dosing strategies to mitigate CRS. Recently we showed a QSP model supporting translational analysis to Recommend an alternate dosing regimen and incremental dose for Xaluritamig (a STEAP TDB for prostate cancer). For the current effort, we focus on recommending a clinical efficacious dose via clinical translation from mouse Tumor Growth Inhibition (TGI) studies.

Methods: In this study, a Quantitative Systems Pharmacology (QSP) platform model was developed for a 2+1 STEAP1 x CD3 bispecific antibody targeting prostate cancer. This platform model was built upon earlier work by Hosseini et al. (2020) [1], which elucidated drug interactions with CD3 and tumor-associated antigens, influencing dimer/trimer formation along with the T cell dynamics. Furthermore, the study highlighted the crucial importance of target receptor occupancy in dictating both drug efficacy and toxicity, providing valuable insights for Non-Hodgkin's lymphoma cancer indication with a 1+1 CD3 bispecific antibody. In recent work, we have further extended this model to a 1+2 bivalent bispecific antibody for Prostate cancer indication to determine the optimal efficacy and dose priming strategy to mitigate CRS. We have used the clinical data for Xaluritamig available from the public source [2] to calibrate and validate this expanded platform model.

Results: The QSP model effectively represented complex PBPK characteristics of drug and T Cells, clinical efficacy and cytokine biomarker data post-Xaluritamig dosing in prostate cancer. It reasonably predicted the efficacious dose observed in clinical settings based on Mouse TGI studies



Conclusions: The model also has capabilities to study the impact of antigen expression levels, binding affinities and baseline T Cell concentration on anti-tumor efficacy and CRS incidence. This platform approach can be used in early clinical drug development stages to inform the therapeutic index.

Citations: [1] Hosseini, Iraj et al. “Mitigating the risk of cytokine release syndrome in a Phase I trial of CD20/CD3 bispecific antibody mosunetuzumab in NHL: impact of translational system modeling.” NPJ systems biology and applications vol. 6,1 28. 28 Aug. 2020, doi:10.1038/s41540-020-00145-7.

[2] Kelly, William K et al. “Xaluritamig, a STEAP1 × CD3 XmAb 2+1 Immune Therapy for Metastatic Castration-Resistant Prostate Cancer: Results from Dose Exploration in a First-in-Human Study.” Cancer discovery vol. 14,1 (2024): 76-89. doi:10.1158/2159-8290.CD-23-0964