Harnessing clinical knowledge on ligand-targeting drug to develop a new compound targeting the associated receptor : an example of model-based biologics design in pre-clinical development
Helene Karcher, Phil Lowe
Modeling & Simulation, Novartis Pharma AG, Basel, Switzerland
Objectives: The modeling work aims at supporting the design of a new biologics compound (B) against a new cellular-receptor target (R). The model predicts the affinity (Kd), half-life (t1/2) and mechanism of action (non-competitive vs. competitive) that B should have to achieve similar or better human efficacy than the known, marketed drug (D) that binds the ligand (L) associated to the new target R. The efficacious dose, half-life and affinity of the ligand-binding drug (D) are known.
Methods: A 2-compartment PK/PD model was developed to predict the amounts of compound B (or drug D), ligand L, receptors R, and bound species as a function of time following an injection in a human. We assumed bi-directional distribution of B between plasma and interstitium via passive exchange across the vascular endothelium. In addition, we included convective transport of B (or D), and L from the interstitium to the plasma via lymph flow. Clearance from the plasma compartment was varied to yield a half-life for B between 3 and 14 days. We also included molecular binding as follows: B + R -> complex (or D + L -> complex for the marketed drug) and L + R -> complex2, as well as receptor turnover in the 2 compartments. B was considered efficacious when it achieved at steady-state the same receptor occupancy as D acting at its known clinical efficacy. The model was formulated as a system of ordinary differential equations and solved with MATLAB (v.7.4, The Mathworks, Natick, MA).
Results: The mechanism of action that led to the best efficacy was the non-competitive binding mechanism given the desired dose, affinity, half-life and disease conditions. A half-life > 7 days and affinity Kd < 2.5nM for B were predicted to yield greater receptor occupancy than D acting at its clinical efficacy.
Conclusions: Knowledge of dose yielding to clinical efficacy for a marketed ligand-targeting drug was used to deduce associated receptor occupancy using a PKPD model that included ligand and receptor binding, and receptor turnover. The information was used to deduce necessary characteristics (half-life, affinity, mechanism of action) to achieve human efficacy for a new compound targeting the receptor associated to the ligand.
