Quantitative Pharmacology and Individualized Therapy Strategies in Development of Therapeutic Proteins for Immune-Mediated Inflammatory Diseases
by Honghui Zhou, Diane R. Mould
8 Tutorial: Numerical (NONMEM) Implementation of the Target‐Mediated Drug Disposition Model
8.1 Introduction
The target‐mediated drug disposition (TMDD) model is a nonlinear system of differential equations, with multiple fixed and random effect parameters that describe complex biological processes. Differential equations of the TMDD model describe processes with very different characteristic timescales, from few minutes for binding processes to several weeks for the elimination of the drug. Numerical methods for solving these equations are numerically unstable, unless the step size is taken to be extremely small. In mathematics such systems are called “stiff.” System parameters are often poorly identifiable either due to stiffness of the differential equations (for the full TMDD model) or due to limitations of the available data. This results in long computational run times and numerical difficulties, requiring careful attention to the selection of software, estimation methods, and its parameters. This chapter is intended to describe how to apply the NONMEM software application [1] and NONMEM estimation methods to the TMDD equations. NONMEM is the NONlinear Mixed Effect Modeling package originally developed by Stuart Beal, Lewis Sheiner, and Alison Boeckmann and is now being developed by Robert Bauer. While several other software packages for nonlinear mixed effect modeling are available, NONMEM remains the standard in the pharmaceutical industry. The chapter should be studied ...