Stefano Zamuner (1), John Toso (2), Sara Brett (3), Maria Feeney (4), Roberto Bizzotto (5), and Daren Austin (1)
(1) Clinical Pharmacology - Modelling and Simulation, GlaxoSmithKline, Stockley Park, UK; (2) Biopharm Translational Medicine, GlaxoSmithKline, Upper Merion, US; (3) Biopharm Translational Medicine, GlaxoSmithKline, Stevenage, UK; (4) Biopharm Discovery Unit, GlaxoSmithKline, Stevenage, UK; (5) Institute of Biomedical Engineering, National Research Council, Padova, Italy
Objectives: A humanised IgG1 monoclonal antibody (mAb) against human Oncostatin M (OSM) has been recently developed for the treatment of rheumatoid arthritis (RA) [1]. The anti-OSM antibody blocks the interaction of OSM with its cell surface signaling receptor component, gp130. Oncostatin M is a member of the interleukin (IL)-6 family of secreted cytokines and is present in the inflamed synovium and blood of patients with RA. Recent negative results in RA indication suggested that the anti-OSM antibody was not sufficiently potent for a cytokine target. In this hypothesis, OSM-antibody complex can act as a carrier protein [2], which prolongs the half life of the cytokine, resulting in accumulation of the complex especially in the joints where it is not efficiently cleared.
Methods: A non-linear mixed effect longitudinal PK-PD analysis on primary efficacy endpoint (DAS28) in RA patient was conducted using a set-point model [3]. Inclusion of covariate effects (i.e. drug exposure and baseline DAS28 on the structure model parameters) has been carried out using SCM procedure in PsN 3.4.2 [4]. In addition, a few target meditated drug disposition (TMDD) models were explored to describe relationship between drug level, free and total serum OSM including investigation of potential carrier protein / agonist effect. Simulations of different scenarios were conducted using Berkeley Madonna software version 8.3.18.
Results: A set point model was successfully estimated and class of exposures was found as a significant covariate. Subjects in the low exposures tertiles showed better response compared with intermediate and high exposures (U-shape curve). Based on these results it has been hypothesised that at high doses, required to fully neutralise the target, there was an increased risk of a protein carrier/ agonist effect in patient. Consequently, TMDD models were developed to account for potential carrier protein effect in the synovial tissue.
Conclusions: Antibodies acting as carrier proteins is not unprecedented and have been documented both preclinically and clinically with mAbs against other soluble cytokines such as IL-2, IL-3, IL-4, IL-6 and IL-7 [5]. This phenomenon is complex to predict and it is related to several factors as target turnover, different antibody and target location (i.e. plasma vs. tissue compartment) and antibody affinity. Quantitative approaches may be of help to identify and describe these potential risks.
References:
[1] MB Baker, M Bendit, AM Campanile, M Feeney, P Hodsman, JF Toso. A Randomised, Single-Blind, Placebo-Controlled Dose Escalation Study to Investigate the Safety, Tolerability and Pharmacokinetics of a Single Intravenous Infusion of GSK315234 in Healthy Volunteers [abstract]. Arthritis Rheum 2009;60 Suppl 10 :427 DOI: 10.1002/art.25510
[2] Finkelman et al. (1993). Anti-Cytokine Antibodies as Carrier Proteins. The Journal of Immunology. 151: 1235-1244.
[3] Austin and Zamuner. A closed-form solution set-point model of treatment response in multiple diseases. PAGE Abstract (2012)
[4] Lindbom L et al. (2005). PsN-Toolkit–a collection of computer intensive statistical methods for non-linear mixed effect modeling using NONMEM. Comput Methods Programs Biomed. 79(3):241-57.
[5] Rehlaender and Cho. (1998). Antibodies as Carrier Proteins. Pharmaceutical Research 15(11): 1652-56
Reference: PAGE 21 (2012) Abstr 2585 [www.page-meeting.org/?abstract=2585]
Poster: Other Drug/Disease Modelling