The bioavailability of a monoclonal antibody (mAb) or another therapeutic protein after subcutaneous (SC) dosing is challenging to predict from first principles, even if the impact of injection site physiology and drug properties on mAb bioavailability is generally understood. We used a physiologically based pharmacokinetic model to predict pre-systemic clearance after SC administration mechanistically by incorporating the FcRn salvage pathway in antigen-presenting cells (APCs) in peripheral lymph nodes, draining the injection site. Clinically observed data of the removal rate of IgG from the arm as well as its plasma concentration after SC dosing were mostly predicted within the 95% confidence interval. The bioavailability of IgG was predicted to be 70%, which mechanistically relates to macropinocytosis in the draining lymph nodes and transient local dose-dependent partial saturation of the FcRn receptor in the APCs, resulting in higher catabolism and consequently less drug reaching the systemic circulation. The predicted free FcRn concentration was reduced to 40–45%, reaching the minimum 1–2 days after the SC administration of IgG, and returned to baseline after 8–12 days, depending on the site of injection. The model predicted the uptake into APCs, the binding affinity to FcRn, and the dose to be important factors impacting the bioavailability of a mAb.

Background: Understanding interactions between cytokine antagonists and drugs is essential for effective medication management in inflammatory conditions. Recent regulatory authority guidelines emphasise a systematic, risk-based approach to evaluating these interactions, underscoring the need for mechanistic insight. Proinflammatory cytokines, such as interleukin-6 (IL-6), modulate cytochrome P450 (CYP) enzymes, reducing the metabolism of CYP substrates. Cytokine antagonists (such as IL-6 receptor antagonists) can counteract this effect, restoring CYP activity and increasing drug clearance. However, quantitative prediction of cytokine-mediated drug interactions remains challenging, as existing models often lack the mechanistic detail needed to capture the dynamic relationship between cytokine signalling, receptor engagement, and downstream modulation of drug metabolism. Methods: A physiologically based pharmacokinetic (PBPK) framework incorporating cytokine–receptor binding, subsequent downregulation of CYP expression, and blockade of the cytokine signalling by a therapeutic protein antagonist was developed to simulate and investigate cytokine antagonist-drug interactions. Tocilizumab, a humanised IL-6 receptor antagonist used to treat several inflammatory conditions associated with elevated IL-6 levels, was selected as a model drug to demonstrate the utility of the framework. Results: The developed PBPK model accurately predicted the pharmacokinetics profiles of tocilizumab and captured clinically observed dynamic changes in simvastatin exposure before and after tocilizumab treatment in rheumatoid arthritis (RA) patients. Simulated IL-6 dynamics aligned with observed clinical profiles, showing transient elevation following receptor blockade and associated restoration of CYP3A4 activity. Prospective simulations with commonly co-administered CYP substrates (celecoxib, chloroquine, cyclosporine, ibuprofen, prednisone, simvastatin, and theophylline) in RA patients revealed dose regimen- and drug-dependent differences in interaction magnitude. Conclusions: This study demonstrated the utility of PBPK models in providing a mechanistic understanding of cytokine antagonist-drug interactions, supporting enhanced therapeutic decision-making and optimising patient care in inflammatory conditions.

Background/Objectives: Physiologically based pharmacokinetic (PBPK) modeling is an important tool in biologic drug development. However, a standardized modeling strategy is currently missing. A cross-industry collaboration developed PBPK models for seven case studies, including monoclonal antibodies, antibody–drug conjugates, and bispecific T-cell engagers, to identify key parameters and establish a workflow to simulate biologic drugs in monkeys and in humans. Methods: PBPK models were developed in the monkey with limited data, including the molecular weight, the binding affinity to FcRn, and the additional systemic clearance of IgG, which is 20% of the total clearance. The binding affinity was only available for human FcRn and corrected for the known species-dependent differences in IgG binding. The strategy of monkey simulations was evaluated with an additional 14 studies published in the literature. Three different scenarios were simulated in humans afterwards: without, with allometrically scaled, and with optimized additional systemic clearance. Results: The plasma peak concentration and the area under the curve were predicted within 50% of the observed data for all studied case examples in the monkey, which demonstrates that sparse input parameters are sufficient for successful predictions in the monkey. Simulations in humans demonstrated the need for additional systemic clearance, because drug exposure was highly overpredicted without an additional systemic clearance term. Allometric scaling improved the predictions, but optimization led to the best fit, which is currently a limitation in the translation from animals to humans. Conclusions: This work highlights the importance of understanding the general mechanisms of drug uptake in different tissue types and cells in both target-dependent and -independent processes.