Genentech Inc. San Francisco, California, United States
Purpose: T cell receptor (TCR) – engineered T cell therapy is an emerging cancer treatment strategy that has shown evidence of anti-tumor activity in both solid tumors and hematologic cancers [1,2]. Despite initial promise, there remain many challenges in understanding and characterizing the unique cellular kinetics, including trafficking, proliferation, contraction, and persistence of TCR-engineered T cells following infusion into the patient. As TCR-engineered T cell therapies are comprised of live and phenotypically diverse T cells, the pharmacological properties of these therapies differ from molecular therapeutics, and monitoring and predicting the kinetics and resulting pharmacodynamics of these therapies poses distinct challenges. Further, it is not yet known how the cellular phenotype and composition of the infusion product might affect the cellular kinetics and subsequent efficacy and safety. Here, we developed a model based on a published clinical study of cell therapy to allow us to investigate the dynamics of TCR-engineered T cell therapies at different dose levels and explore the impact of altering the phenotypic makeup of the infusion product. Methods: We developed a quantitative systems pharmacology (QSP) model to track T cell kinetics of four different subsets of TCR-engineered T cells (stem-like memory T cells (Tscm), central memory T cells (Tcm), effector memory T cells (Tem), and effector T cells (Teff)) across three physiological compartments, including blood, healthy tissue, and tumor. Each subpopulation undergoes homeostatic proliferation, antigen-driven proliferation and differentiation, apoptosis, margination and trafficking between blood and healthy tissues. In addition, Tem and Teff cell populations can infiltrate the tumor compartment, and Teff can kill tumor cells. To capture the potential competition between immune cells following infusion with TCR-engineered T cells, we incorporate lymphocyte depletion conditioning therapy and subsequent expansion of endogenous T cells (Tendo). The model was calibrated to phase I clinical trial data of TCR-engineered T cells targeting E7 in epithelial cancer patients [1], and implemented as a general platform model adaptable to other cancer antigens and tumor types. The model was developed in Matlab R2020b (MathWorks, Natick MA) using Simbiology and the gQSPsim toolbox [3]. Results: The calibrated QSP model was able to recapitulate the kinetics of TCR-engineered T cells targeting E7 in epithelial cancer patients dosed at three different dose levels (~ 109, ~ 1010, ~ 1011 cells). The model successfully captured the initial margination, subsequent expansion, contraction, and eventual persistence for each of the T cell subsets observed across time for all three dose levels. Following calibration, we sampled parameter space to create a virtual cohort that captured the variability of the clinical data across each dose level. Additionally, we leveraged this model framework to perform a sensitivity analysis to determine the relationship between the persistence of T cells across time and the variability in cellular proliferation, differentiation and trafficking of T cells. This analysis suggests that proliferation rates of Tscm, and Tcm are key drivers that determine the persistence of Tem and Teff across time and these parameters are critical for capturing separate patient profiles (such as those with high levels of cellular expansion or decline). Finally, we examined the impact of variability in dose level and phenotypic makeup of the initial T cell product on the cellular kinetics, distribution, and persistence of the T cell therapy across time. Our simulations suggest that infusion products consisting of a majority of Tscm cells are more likely to have greater numbers of total TCR-engineered T cells remaining within a patient 200 days post-infusion compared to simulations of infusion products with a majority of Teff cells. Future work could leverage this model framework to perform patient-specific analysis during clinical trials with TCR-engineered T cell therapies that target other cancer antigens. Conclusion: The QSP model is able to capture the cellular kinetics in the blood for four different T cell phenotypes across three dose levels. Additionally, the model allows us to investigate determinants of cell expansion in the tumor and healthy tissues. This general modeling framework provides a novel approach for predicting the impact of varying dose and phenotype ratios of TCR-engineered T cell therapies. References: [1] Nagarsheth, N.B., Norberg, S.M., Sinkoe, A.L. et al. TCR-engineered T cells targeting E7 for patients with metastatic HPV-associated epithelial cancers. Nat Med 27, 419–425 (2021). https://doi.org/10.1038/s41591-020-01225-1 [2] Robbins, P. F. et al. A pilot trial using lymphocytes genetically engineered with an NY-ESO-1-reactive T cell receptor: long-term follow-up and correlates with response. Clin. Cancer Res. 21, 1019–1027 (2015). [3] Hosseini I, et al. gQSPSim: A SimBiology-Based GUI for Standardized QSP Model Development and Application. CPT Pharmacometrics Syst Pharmacol. 2020 Mar;9(3):165-176.
