A Mechanistic Approach to Predicting Human Pharmacokinetics of Monoclonal Antibodies from Preclinical Data: A Case Example
Kenneth T. Luu (1), Eugenia Kraynov (1)
(1) Department of Pharmacokinetics Dynamics and Metabolism, Pfizer Global Research and Development, La Jolla, California, USA
Objectives: The objective of this study was to develop a mechanistic approach to predicting human pharmacokinetics (PK) from in-vitro mechanistic data and in-vivo nonhuman primate PK data based on the construct of a target-mediated drug disposition (TMDD) model (1).
Methods: A fully human IgG2 monoclonal antibody (mAbX), currently under development in the oncology therapeutic area, directed against a membrane-bound target was evaluated in this case study. A two-compartment, full TMDD model (in contrast with quasi-steady state TMDD models) was implemented to describe the nonlinear PK profiles observed in monkeys. Mechanistic parameters were obtained from in-vitro experiments (kon and koff binding kinetics and the mAbX-target complex internalization rate) and from the literature (degradation rate of the free target). Unique in this approach is the estimation of the in-vivo receptor abundance as a parameter which was intended to be scalable to human. Human PK was predicted using a hybrid approach: allometric scaling of physiological parameters such as kel, and k12 and k21, but mechanistic parameters specific for human targets were used. The mean predicted human PK profiles were compared against the currently available clinical PK data of patients receiving mAbX from 0.5 up to 6.75 mg/kg.
Results:. The two-compartment TMDD model described the nonlinear single-dose and multiple-dose PK profiles of mAbX in monkeys and estimated all parameters with reasonable precisions. The model predicted the human PK to follow TMDD which was confirmed by the clinical data. By visual inspection, the mean predicted PK profiles reasonably overlapped with the spread of the observed individual patient PK profiles. Furthermore, noncompartmental analysis of the mean prediction and the observed data indicated that predicted and observed CLs, and AUCs, were within 1.5 fold of each other.
Conclusion: The PK of monoclonal antibodies, especially those directed against membrane-bound targets that are of significant abundance, is often greatly affected by the binding and target kinetics. Thus, incorporating target kinetics into a mechanistic PK model to be used for interspecies scaling is a sensible approach for human PK prediction. Based on the mAbX case example, such an approach successfully predicted the human PK based on the preclinical data including in-vitro target kinetic data and in-vivo non-human primate PK data.
. Mager, D.E. & Jusko, W.J. General pharmacokinetic model for drugs exhibiting target-mediated drug disposition. J Pharmacokinet Pharmacodyn 28, 507-32 (2001).