(T-005) A novel PBPK/PD model of preclinical data projects the human pharmacologically active dose of a Biologic that modulates immunosuppressive immune cells in the tumor microenvironment

Eunah Bae, PhD – Scientist, Takeda Pharmaceuticals; Dean Bottino, PhD – Director, Takeda Pharmaceuticals; Cierra Casson, PhD – Director, Takeda Pharmaceuticals; Carla Guarino, PhD – Scientist, Takeda Pharmaceuticals; Tiquella Hatten, MS – Scientist, Takeda Pharmaceuticals; Sarah Hesse, PhD – Scientist, Takeda Pharmaceuticals; Natasha Iartchouk, PhD – Scientist, Takeda Pharmaceuticals; Jessica Riceberg, PhD – Scientist, Takeda Pharmaceuticals; Brandon Wilkinson, MS – Scientist, Takeda Pharmaceuticals; Tomoki Yoneyama, PhD – Director, Takeda Pharmaceuticals; Ting Ting Zhang, PhD – Director, Takeda Pharmaceuticals

Objectives: Certain immune cells in the tumor microenvironment (TME), such as myeloid-derived suppressor cells, regulatory T cells, and type II macrophages, are considered immunosuppressive. Tumors with high levels of infiltration by these cells are associated with poor patient prognosis and resistance to therapy. Thus, modulating one or more of these immunosuppressive immune cells (IICs) in the TME may release an immune brake on cytotoxic effector cells, enabling them to kill cancer cells. A biologic, targeting a surface antigen specific to one of these IIC types, therefore represents a potential therapeutic strategy to exploit this mechanism. We have developed a novel translational model of such a Biologic. The model is calibrated with preclinical data to project the human pharmacologically active dose (PAD) and the optimal biopsy collection time to observe maximal IIC modulation.

Methods: We developed a translational physiologically based pharmacokinetic/pharmacodynamic (PBPK/PD) model that is calibrated to internal preclinical data from mouse, NHP, and human samples to project the human PAD. A minimal PBPK model is utilized to characterize plasma PK of our Biologic, and the Krogh cylinder model is used to predict tumor exposure. The tumor concentration of the Biologic is linked to the PD of IIC modulation, which is then linked to tumor growth inhibition (TGI). The model, calibrated to mouse in-vivo data, is used to determine threshold levels of IIC modulation needed to achieve desired TGI. Human PK is projected from scaling up physiological parameters of the PBPK model calibrated to NHP plasma PK. Human PK to PD relationship is projected from the mouse model, while accounting for a species scaling factor that is derived from mouse and human ex-vivo IIC modulation assays. Finally, human tumor kinetics are taken from literature values that are calculated from patient data.

Results: The PBPK/PD model components captured the totality of the preclinical data, which includes: (i) mouse in-vivo PK, PD, and efficacy; (ii) NHP PK; and (iii) mouse and human ex-vivo PD. The mouse model determined that achieving 60% TGI requires 70% IIC modulation, which was set as the threshold for pharmacological activity. The mouse/NHP model was then translated to predict human PKPD. The model predicted that (i) human PAD is a median value of 0.032 mg/kg, and (ii) optimal biopsy collection time is between 2-7 days post dose.

Conclusions: We have developed a novel PBPK/PD model for a Biologic that modulates an immunosuppressive subset of immune cells with preferential target expression in the TME. The model was calibrated to preclinical data and then translated to humans to project the human PAD and optimal biopsy collection time. These findings will be used to inform our early clinical development plan for the Biologic.

Citations:
[1] Cao, Y, et al. "Second-generation minimal physiologically-based pharmacokinetic model for monoclonal antibodies." Journal of pharmacokinetics and pharmacodynamics 40 (2013): 597-607.
[2] Shah, DK, and Betts, AM. "Towards a platform PBPK model to characterize the plasma and tissue disposition of monoclonal antibodies in preclinical species and human." Journal of pharmacokinetics and pharmacodynamics 39 (2012): 67-86.
[3] Simeoni, M, et al. "Predictive pharmacokinetic-pharmacodynamic modeling of tumor growth kinetics in xenograft models after administration of anticancer agents." Cancer research 64.3 (2004): 1094-1101.
[4] Kay, K, et al. "Estimation of solid tumor doubling times from progression-free survival plots using a novel statistical approach." The AAPS journal 21 (2019): 1-12.