Development of a PBPK model to predict monoclonal antibody pharmacokinetics and bioavailability following subcutaneous administration
Salih Benamara1,3, Donato Teutonico1, Florence Gattacceca3, Laurent Nguyen1, Antoine Deslandes1, Marylore Chenel2, Erik Sjögren2
(1) Sanofi R&D, Paris, France, (2) Pharmetheus, Uppsala, Sweden, (3) Aix-Marseille University, Marseille, France,
Introduction: Monoclonal antibodies (mAbs) are currently established components of oncology and immunology therapies, typically administered intravenously over extended treatment periods of several years (1). To enhance patient convenience and improve quality of life, there is a growing interest in a less invasive, considerably quicker, and more flexible administration method, typically the subcutaneous (SC) administration (2). Nearly half of the approved mAbs by the Food and Drug Administration (FDA) in recent years involved SC administration (3), highlighting the growing importance of this route of administration (4). Physiologically based pharmacokinetic (PBPK)models are useful approach during drug product development due to their ability to scale pharmacokinetic (PK) prediction between species and populations (5). However, there is a current lack of a suitable model that adequately characterizes absorption after SC administration. The prediction of bioavailability following SC administration presents considerable challenges for therapeutic proteins like mAbs, primarily due to the limited predictive reliability of animal studies (6).
Objective: The aim of this study was to predict the PK of biologics (30 mAbs and 1 fusion protein). To achieve this, the generic comprehensive PBPK model in PK-SIM, originally designed for mAbs administration via IV route, has been expanded. The extension incorporates the modeling and prediction of drug absorption and bioavailability after SC administration.
Methods: The open-source platform Open Systems Pharmacology Suite (OSPS) was used for model development with PK-Sim and MoBi (7). A database containing in vitro drug properties and in vivo PK data of biologics (mainly mAbs) following IV and SC administration was compiled from the literature. Plasma concentrations vs time data of 31 drugs were digitized based on an intensive literature search. The IV plasma concentrations data for each drug were used to estimate the binding to FcRn receptor (FcRn Kd) and build a PBPK model for IV administration. The developed IV models were then expanded by a mechanistic model describing the SC compartment. Briefly, this model is composed by an injection site compartment and linked to the whole-body circulatory system through the SC administration site plasma flow and lymph flow. The top-level model structure for the injection site includes one depot compartment connected to 99 layers in series, representing the volume surrounding the depot). Based on the default PBPK model structure implemented in OSPS for biologics, each layer is composed of five sub-compartments representing: plasma, endosome, interstitial, lymphatic, and intracellular space. Additionally, it includes two compartments representing the local lymph node and the central collecting lymph duct. Model performances were evaluated by visual comparison of the simulated concentration-time profiles to the observed in vivo PK data and predictive errors for AUC and Cmax were also calculated. The bioavailability was determined by calculating AUCSC/AUCIV (normalized by the dose) based on predictions derived from non-compartmental analysis (NCA). The bioavailability value obtained from prediction was compared to the reference bioavailability derived from observed data.
Results: PBPK models after IV administration were successfully established for all included drugs adopting the generic large molecule implementation in PK-Sim. The median of estimated FcRn Kd was 0.93 µM with a range of 0.55 µM to 3.00 µM. These compound specific models described 90% of observed AUCs, across studies and doses, within a 0.80-1.25-fold difference. The SC PBPK model, informed by the estimated FcRn Kd, was able to successfully capture, within a 0.80-1.25-fold difference, observed AUC and Cmax for 60% of the database (18 mAbs), across studies and doses. Furthermore, the model achieved prediction accuracy within 0.50-2.00-fold range for most of the examined cases, 100% for Cmax and 94% for AUC.
Conclusion: The favorable predictive performance achieved by the implemented SC module confirm the potential of the physiologically based components in the OSPS platform, highlighting its usefulness as a valuable tool for enhancing PBPK modeling of mAbs administered SC.
- Kaplon H, Crescioli S, Chenoweth A, Visweswaraiah J, Reichert JM. Antibodies to watch in 2023. MAbs. 2023;15(1):2153410.
- Bittner B, Richter W, Schmidt J. Subcutaneous Administration of Biotherapeutics: An Overview of Current Challenges and Opportunities. BioDrugs. 1 oct 2018;32(5):425‑40.
- Martin KP, Grimaldi C, Grempler R, Hansel S, Kumar S. Trends in industrialization of biotherapeutics: a survey of product characteristics of 89 antibody-based biotherapeutics. mAbs. 31 déc 2023;15(1):2191301.
- Viola M, Sequeira J, Seiça R, Veiga F, Serra J, Santos AC, et al. Subcutaneous delivery of monoclonal antibodies: How do we get there? Journal of Controlled Release. 28 sept 2018;286:301‑14.
- Jones H, Rowland-Yeo K. Basic Concepts in Physiologically Based Pharmacokinetic Modeling in Drug Discovery and Development. CPT Pharmacometrics Syst Pharmacol. août 2013;2(8):e63.
- Dubbelboer IR, Sjögren E. Physiological based pharmacokinetic and biopharmaceutics modelling of subcutaneously administered compounds – An overview of in silico models. International Journal of Pharmaceutics. 10 juin 2022;621:121808.
- Niederalt C, Kuepfer L, Solodenko J, Eissing T, Siegmund HU, Block M, et al. A generic whole body physiologically based pharmacokinetic model for therapeutic proteins in PK-Sim. J Pharmacokinet Pharmacodyn. 2018;45(2):235‑57.