A Translational Quantitative Systems Pharmacology Model for CD3 Bispecific Molecules: Application to Quantify T Cell-Mediated Tumor Cell Killing by P-Cadherin LP DART®

Abstract

CD3 bispecific antibody constructs recruit cytolytic T cells to kill tumor cells, offering a potent approach to treat cancer. T cell activation is driven by the formation of a trimolecular complex (trimer) between drugs, T cells, and tumor cells, mimicking an immune synapse. A translational quantitative systems pharmacology (QSP) model is proposed for CD3 bispecific molecules capable of predicting trimer concentration and linking it to tumor cell killing. The model was used to quantify the pharmacokinetic (PK)/pharmacodynamic (PD) relationship of a CD3 bispecific targeting P-cadherin (PF-06671008). It describes the disposition of PF-06671008 in the central compartment and tumor in mouse xenograft models, including binding to target and T cells in the tumor to form the trimer. The model incorporates T cell distribution to the tumor, proliferation, and contraction. PK/PD parameters were estimated for PF-06671008 and a tumor stasis concentration (TSC) was calculated as an estimate of minimum efficacious trimer concentration. TSC values ranged from 0.0092 to 0.064 pM across mouse tumor models. The model was translated to the clinic and used to predict the disposition of PF-06671008 in patients, including the impact of binding to soluble P-cadherin. The predicted terminal half-life of PF-06671008 in the clinic was approximately 1 day, and P-cadherin expression and number of T cells in the tumor were shown to be sensitive parameters impacting clinical efficacy. A translational QSP model is presented for CD3 bispecific molecules, which integrates in silico, in vitro and in vivo data in a mechanistic framework, to quantify and predict efficacy across species.

Keywords: CD3 bispecific; PK/PD; immunotherapy; quantitative systems pharmacology; translational modeling.

Fig. 1

a Translational quantitative systems pharmacology model for CD3 bispecific molecules. Parameter descriptions and values are summarized in Tables I to III. The figure represents both mouse and human models, with the following exceptions: binding to sPcad was only included in the human model and T cell proliferation/ exhaustion in the tumor were only included in the mouse model. b Schematic of the bell-shaped concentration relationship which can be observed for CD3 bispecific molecules. Formation of trimers between drugs, T cells, and tumor cells, is required for efficacy. The QSP model predicts trimer concentration and links it to tumor cell killing

Fig. 2

a Serum and b tumor PK profiles of PF-06671008 in PBMC engrafted and non-PBMC engrafted HCT-116 tumor-bearing mice following single-dose intravenous administration at 0.05 and 0.5 mg/kg

Fig. 3

PF-06671008-induced tumor T cell proliferation in mice bearing HCT-116 tumors with human PBMC engraftment. Number of CD3+ cells/mg of the tumor (with standard deviations) is plotted against time following IV administration of control and PF-06671008 at 10 μg/kg, 50 μg/kg, and 500 μg/kg

Fig. 4

Model simulated a serum PK and b tumor trimer concentrations following IV infusion of PF-06671008 at 0.01, 0.1, and 1 μg/kg QW to cancer patients

Fig. 5

Model simulated tumor trimer concentrations at a different P-cadherin receptor expression values (1000–28,706 receptors/cell) and b different E:T ratios (1:1500–1:15) following IV infusion of PF-06671008 at 0.1 μg/kg QW to cancer patients

Fig. 6

QSP model-based strategy for translating preclinical data for CD3 bispecific compounds to the clinic. “Biomeasures” can be defined as system-dependent parameters. TAA is a tumor-associated antigen