PAGE. Abstracts of the Annual Meeting of the Population Approach Group in Europe.
PAGE 22 (2013) Abstr 2896 [www.page-meeting.org/?abstract=2896]
Poster: Other Drug/Disease Modelling
E. Mezzalana (1), A. MacDonald (2), G. De Nicolao (1), S. Zamuner (2)
(1) Department of Computer Engineering and Systems Science, University of Pavia, Pavia, Italy; (2) Clinical Pharmacology Modelling and Simulation, GlaxoSmithKline, Stockley Park, UK;
Objectives: Otelixizumab is a monoclonal antibody currently being investigated in autoimmunity. It is directed against human CD3ε on T lymphocytes. Its pharmacological effects include 1) down modulation of the CD3/T cell receptor complex on T lymphocytes and 2) a decrease of T cells in blood. The aim of the present work was to integrate a mechanistic target-mediated drug disposition (TMDD) model  to a T lymphocyte pharmacodynamic (PD) model.
Methods: Free drug in serum and CD4+ and CD8+ T lymphocytes counts were measured using immunoassay and flow cytometry, respectively. Free, bound and total receptors were then obtained for both CD4+ and CD8+ T lymphocytes . A QSS-TMDD model accounting for Otelixizumab binding to receptors on both CD4+ and CD8+ cells was implemented . Direct and indirect inhibition models were investigated to describe the observed T cell reduction in blood. Analyses were conducted using NONMEM version 7.2. FOCEI and IMP estimation methods were used and compared. Final models were selected based upon change in OFV, precision estimates, diagnostic plots and visual predictive checks (VPCs).
Results: First, a simple one-compartment PK model with MM elimination was identified using available PK data. This PK model was used to drive a direct or an indirect lymphocyte PD model. Both sequential (IPP) and simultaneous PK-PD (T lymphocytes) analyses were implemented. Based on VPCs, the indirect model provided a slightly better description of lymphocyte time-course and variability. Then, our previous QSS-TMDD model  was improved including a ‘Study' covariate on linear elimination rate constant (K) with a consequent reduction of the estimated high BSV on K. This QSS-TMDD model was used as input for both lymphocyte PD models. Sequential and simultaneous TMDD-lymphocyte analyses were conducted. To overcome problems with OF minimization and covariance step failure with FOCEI the IMP method was used to achieve minimization with covariance step completion. Both TMDD+direct and TMDD+indirect models adequately described the lymphocyte data.
Conclusions: A QSS-TMDD model integrated either to a direct or an indirect inhibitory model was proposed to describe Otelixizumab binding to CD3/TCR on T lymphocytes and subsequent decrease of T lymphocytes in blood. The IMP estimation method proved a useful alternative to FOCEI in case of such complex PK/PD models. Further strategies to improve lymphocytes model integration into TMDD are discussed.