(T-122) A Quantitative Systems Pharmacology Approach to Understand the Variability of Patient Response to T-cell Bi-specifics in Hematological Malignancies

Md Shahinuzzman, PhD – Senior Scientist, Systems Medicine, AstraZenenca; Massimo Lai, PhD – Associate Director, Systems Medicine, AstraZenenca; Chang Gong, PhD – Associate Director, Systems Medicine, AstraZeneca; Holly Kimko, PhD – Executive Director, Systems Medicine, AstraZeneca; Cesar Pichardo, PhD – Senior Director, Systems Medicine, AstraZeneca

Objective: Bispecific T cell engager (TCE) is a promising class of immune-oncology therapies that work by binding to the CD3 subunit of the T cell receptor and to the surface antigen on tumor cell leading to lysis of tumor cells. Quantitative systems pharmacology (QSP) are powerful tools that can be used a-priori to predict how the individual will respond to TCEs and how individual factors (system parameters) might influence treatment success. Despite their usefulness, these models can become intricate due to the inherent complexity that arises from integration of data from various sources and unidentifiable parameter space. To address this, this work aims to conduct a sensitivity analysis on a previously developed bispecific TCE QSP model and identify key parameters that influence the predictions, i.e. the response to treatment.

Methods: We utilized and implemented a QSP framework to evaluate the response of diffuse large B-cell lymphoma, a subtype of non-Hodgkin lymphoma (NHL), to a CD3xCD19 T-cell engaging bispecific antibody. The model implemented in SimBiology® incorporates the dynamics of B cells (pro-B, pre-B, and mature) and T cells (resting, activated, and post-activated) within the peripheral blood, normal tissue, and tumor [1]. We conducted a global sensitivity analysis to compute Sobol’s total and first order sensitivity indices [2][3]. The time-dependent impact of the host factors on the percentage change in tumor volume was quantified with various simulation scenarios to explore dose regiments in patients.

Results: The computed Sobol's indices enabled the identification and subsequent analysis of influential host factors. Initially, the B to T cell ratio within the tumor and production rate of IL6 cytokine played an important role in the percentage change in tumor volume in response to the step-up dosing. As time progressed, the ratio of the baseline concentration of B cells in the lymph nodes compared to that in peripheral blood became more important. This sensitivity analysis allowed us to understand the source of variability observed in clinical studies, suggesting potential changes in the dose regimen for some patients to improve efficacy.

Conclusions: Sensitivity analysis is an important process of QSP model development and evaluation which helps identify critical input factors that affect treatment outcomes. By conducting a global sensitivity analysis, we provide a quantitative understanding of the factors that affect the treatment response over time. Future work aims to guide the further mechanistic understanding of the treatment response and improvement in compound design to maximize the treatment outcome.

Citations: [1] Hosseini I, Gadkar K 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 Syst Biol Appl. 2020 Aug 28;6(1):28;
[2] Saltelli, Andrea, Paola Annon et al. “Variance Based Sensitivity Analysis of Model Output. Design and Estimator for the Total Sensitivity Index.” Computer Physics Communications 181, no. 2 (February 2010);
[3] Florian Augustin (2024). Global Sensitivity Analysis for SimBiology Retrieved April 29, 2024.