NONMEM Implementation of Cell Lifespan Models for Hematological Drug Effects
Perez-Ruixo, JJ; Kimko, HC; Chow, A; Piotrovskij, V
Johnson & Johnson Pharmaceutical Research and Development
Different PK/PD models for dealing with the hematological drug effects have been recently published in the literature [1,2]. The models assumed that cells produced by a zero or first order process, survive for a specific duration (cell lifespan) and then are lost. The rate of cell loss must be equal to the production rate but is delayed by the cell lifespan. Drugs can stimulate or inhibit the production rate of cells according to a general Hill function . In order to model the proliferation and maturation stages of cells, different number of precursor pool compartments could be added to the models. In addition, anticancer agents are assumed to act on the precursor pool based on irreversible linear or capacity-limited cytotoxicity .
Due to the fact that delayed differential equations are needed for cell lifespan models, their implementation in NONMEM is not straightforward. The objective is to present the NONMEM implementation of five different cell lifespan models for dealing with the hematological drug effects and to evaluate NONMEM performance to estimate the model parameters.
Model 1 and 2 assume cells are produced by a zero and first order process, respectively. Model 3 and 4 assume a precursor pool indirect model to account for cell proliferation and removal from the circulation. In model 3, every process was quantitated with a specific cell life-span, after a zero order precursor rate production, whereas model 4 assume a zero order precursor rate production followed by a first order proliferation rate. In this model the removal from the circulation occurs after a cell lifespan. Model 5 assume a three-pool indirect model to account for the cell proliferation, maturation and removal from the circulation. Every process was quantitated with a specific cell life-span, after a zero order precursor rate production. Models 1 to 4 assume drug stimulated the precursor production rate. Model 5 mimics the anticancer drug effects and assumes drug effect takes place at the cell proliferation stage following a capacity-limited cytotoxicity.
Simulation followed by estimation were used to evaluate NM performance and the impact of the interindividual and residual variability magnitude on the estimates of the population parameters.
 Krzyzanski W, Ramakrishnan R, Jusko WJ. Basic Pharmacodynamic Models for Agents That Alter Production of Natural Cells. J Pharmacokinet Pharmacodyn 1999; 27: 311-337.
 Krzyzanski W, Jusko WJ. Multiple-Pool Cell Lifespan Model of Hematologic Effects of Anticancer Agents. J Pharmacokinet Pharmacodyn 2002; 29: 467-489.