David Ternant, Hervé Watier, Céline Desvignes, Nicolas Azzopardi, Alain Le Pape, Valérie Gouilleux-Gruart, Gilles Paintaud
Université de Tours, Tours, France
Introduction: Immunisation against monoclonal antibodies (mAbs) lead to the development of anti-drug antibodies (ADA) that are responsible of an increase in mAb clearance by approximately 1.5 fold [1]. Immunization against mAbs represents a big challenge in human therapy. Only a few previous works assessed the joint kinetics of both mAbs and anti-mAb ADA and suggested a nonlinear elimination of mAbs [2], as well as a mechanism of mAb elimination similar to target-mediated drug disposition (TMDD) [3].
Objectives: This study aimed at describing the kinetics of immunisation and interactions of rituximab, an anti-CD20 mAb, and anti-rituximab antibodies (ARA) in mice using TMDD-derived models, defined as ADA-mediated drug disposition (ADAMDD).
Methods: In this study, 129 male mice of 9 different lines were administered with a single 50 μg dose of rituximab in the tail vein. A total of 829 blood samples were collected at hours 1, 6 and 24, and days 3, 7, 15 and 25 after administration. Rituximab and ARA concentrations were respectively measured at each sampling time and at day 25 using in-house ELISA techniques. Rituximab pharmacokinetics was described on the basis of a two-compartment model. Rituximab ARA-mediated elimination was described using Michaelis-Menten, constant total target (RTOT), irreversible binding (IB) or quasi-steady-state (QSS) approximations [4-5]. The kinetics of ARA input was described using 4 transit compartments, and zero and first-order rate constants for ARA production and transit, respecrtively. First-order transit and elimination rate constants were described using a single parameter. The kinetics of ARA was then tested as a two-compartment model, assuming that both rituximab and ARA share the same values for central volume of distribution, and transfer and elimination rate constants. Rituximab ARA-mediated elimination was assumed as occuring in both central and peripheral compartments. Levels of ARA were tested as a covariate on kinetic parameters. Model parameters were estimated using nonlinear mixed-effect modeling using Monolix suite (Lixoft®, Antony, France). Models were compared using difference in Akaike’s information criterion (AIC).
Results: A nonlinear elimination shape of rituximab concentrations was observed in two thirds of mice. Compared to Michaelis-Menten model, rituximab concentration-time data in mice was described using IB (ΔAIC=-173.58) than RTOT (ΔAIC=-163.70) or QSS (ΔAIC=-171.78) approximations. The performance of IB model was improved while considering two-compartment ARA kinetics (ΔAIC=-4.47) compared to IB model. This two-compartment ARA IB model was even more refined using FcRn-mediated elimination model (ΔAIC=-4.88). Model parameter estimates (estimate, interindividual standard deviation) were: central volume of distribution (VC=1.8 mL, 0.28), first-order central-to-peripheral (k12=2.3 d-1, –) and peripheral-to-central (k21=1.2 d-1, –) transfer rate constants, zero-order ARA input rate constant (kin=11.4 nM/d-1, 3.3), first-order transit and elimination rate constant (ktr=0.24 d-1, –) and antibody fraction unbound to FcRn and eliminated (FU=0.091, 0.20). The kinetics of ARA input revealed an onset of ARA-mediated elimination starting 7 to 10 days after rituximab injection. Levels of ARA were strongly associated with kin (ΔAIC=-114.79).
Conclusions: Since rituximab does not bind mouse CD20, nonlinear elimination may be imputable to immunization against rituximab only. Rituximab concentration-time data was satisfactorily described using our ADAMDD model. This suggests that interactions of mAbs and ADA are similar to mAb-target interactions and might be described using TMDD modeling strategies. Main issues may be the description of the onset of ADA production, which is hardly predictible in human therapy.
References:
[1] Bensalem A, Ternant D. Pharmacokinetic Variability of Therapeutic Antibodies in Humans: A Comprehensive Review of Population Pharmacokinetic Modeling Publications. Clin Pharmacokinet. 2020;59:857-874.
[2] Ng CM, Loyet KM, Iyer S, Fielder PJ, Deng R. Modeling approach to investigate the effect of neonatal Fc receptor binding affinity and anti-therapeutic antibody on the pharmacokinetic of humanized monoclonal anti-tumor necrosis factor-alpha IgG antibody in cynomolgus monkey.Eur J Pharm Sci. 2014;51:51-8.
[3] Liao KH, Udata C, Yin D, Sewell KL, Kantaridis C, Alvarez DF, Meng X. A mechanistic pharmacokinetic model with drug and antidrug antibody interplay, and its application for assessing the impact of immunogenicity response on bioequivalence testing. Br J Clin Pharmacol. 2020;86:2182-2191.
[4] Target-mediated drug disposition model: approximations, identifiability of model parameters and applications to the population pharmacokinetic-pharmacodynamic modeling of biologics. Gibiansky L, Gibiansky E. Expert Opin Drug Metab Toxicol. 2009;5:803-12.
[5] Dua P, Hawkins E, van der Graaf PH. A Tutorial on Target-Mediated Drug Disposition (TMDD) Models. CPT Pharmacometrics Syst Pharmacol. 2015;4:324-37.
Reference: PAGE 30 (2022) Abstr 9951 [www.page-meeting.org/?abstract=9951]
Poster: Drug/Disease Modelling - Oncology