A systems pharmacokinetic/pharmacodynamic model for concizumab to explore the potential of anti-TFPI recycling antibodies

Abstract

Concizumab is a humanized monoclonal antibody in clinical investigation directed against membrane-bound and soluble tissue factor pathway inhibitor (mTFPI and sTFPI) for treatment of hemophilia. Concizumab displays a non-linear pharmacokinetic (PK) profile due to mTFPI-mediated endocytosis and necessitates a high dose and frequent dosing to suppress the abundant sTFPI, a negative regulator of coagulation. Recycling antibodies that can dissociate bound mTFPI/sTFPI in endosomes for degradation and rescue antibody from degradation have a potential in reducing the dose by extending antibody systemic persistence and sTFPI suppression. We developed a systems PK/pharmacodynamics (PD) model with nested endosome compartments to simulate the effect of decreased antibody binding to mTFPI/sTFPI in endosomes on antibody clearance and sTFPI suppression for exploring the potential of anti-TFPI recycling antibodies in reducing the dose. A dynamic model-building strategy was taken. A reduced PK/PD model without the endosome compartments was developed to optimize unknown target turnover parameters using concizumab PK data. The optimized parameters were then employed in the systems PK/PD model for simulations. The obtained systems PK/PD model adequately described the PK of concizumab in rabbits, monkeys, and humans and the PD in humans. The systems PK/PD model predicted that an anti-TFPI recycling antibody with a 100-fold higher mTFPI/sTFPI dissociation constant in endosomes than concizumab can extend sTFPI suppression from 12 days to 1 month. Thus, the systems PK/PD model provides a quantitative platform for guiding the engineering and translational development of anti-TFPI recycling antibodies.

Keywords: Concizumab; FcRn; Pharmacokinetics modeling; Recycling antibody; Target-mediated drug disposition; Tissue factor pathway inhibitor.

Figure 1.

Model structure of the systems PK/PD model for anti-TFPI antibodies. Model parameters are defined in Appendix I.

Figure 2.

The systems PK/PD model simulations of total antibody and free sTFPI in humans after IV and SC administration of concizumab. The dotted horizontal lines are the reported lower limit of quantification (LLOQ) for concizumab (Agerso et al., 2014) and sTFPI (Duckers et al., 2008).

Figure 3.

The systems PK/PD model simulations of anti-TFPI recycling antibodies in humans at IV dose of 3 mg/kg. Recycling antibodies were simulated by either decreasing the antibody-mTFPI/sTFPI association rate constant (k_eon, left) or increasing the dissociation rate constant (k_eoff, right) in endosomes. The dotted horizontal lines are the LLOQ for sTFPI (Duckers et al., 2008).

Figure 4.

Optimization of antibody nonspecific pinocytosis (CL_up) across species. CL_up was optimized using PK data of linearly cleared antibodies: (A) Bevacizumab and the Fc-engineered variant of bevacizumab, Xtend, in monkeys at IV 4 mg/kg; (B) Concizumab Hz isotype control in rabbits at IV 2 mg/kg. (C) Allometry of CL_up across species based on body weight (BW). CL_up = 0.029BW⁰⁸³ ,R² = 0.98. Symbols are observations, lines are model predictions.

Figure 5.

The systems PK/PD model simulations of concizumab in monkeys. The symbols are observations and lines are model predictions. Dotted horizontal lines are the reported LLOQ for concizumab (Agerso et al., 2014). Group 1, 2, and 3: SC single dose escalation study at two-week intervals, 2-20-80 mg/kg, 20-80-160 mg/kg, and 80-160-200 mg/kg, respectively; Group 4 and 5: IV and SC single dose at 20 mg/kg; Group 6, 7 and 8: SC 91 daily doses at 1 mg/kg, 10 mg/kg, and 50 mg/kg, respectively; Group 9: IV 13 daily doses at 200 mg/kg.

Figure 6.

The systems PK/PD model simulations of concizumab in rabbits after IV administration at 2 and 20 mg/kg. The symbols are observations and the lines are model predictions.

Figure 7.. Local sensitivity analysis of the systems PK/PD model.

Each simulation was done by individually decreasing or increasing the parameter value 10-fold. Outputs are the AUC ratio of total antibody and free sTFPI (AUC_after/AUC_before). The simulations were done at two IV doses: (A) 0.25 mg/kg and (B) 9 mg/kg. The horizontal line indicates AUC ratio = 1.