(T-005) Bench to bedside modelling of mRNA encoding IgG using a multiscale mechanistic PK-PD model: A case study with anti-claudin 18.2

Devam Desai – Center of Pharmacometrics and Systems Pharmacology, University of Florida, Orlando, FL; Rodrigo Cristofoletti – Center of Pharmacometrics and Systems Pharmacology, University of Florida, Orlando, FL

Objectives: Messenger RNA (mRNA) therapeutics encoding monoclonal antibodies offer a novel platform for in vivo protein expression. However, quantitative translation from preclinical to clinical systems remains a major hurdle due to complex multi-compartmental processes including nanoparticle delivery, cellular uptake, mRNA translation, and target binding. This study aimed to develop and apply a multiscale mechanistic pharmacokinetic-pharmacodynamic (PK-PD) model to characterize and predict the in vivo behavior of an mRNA therapeutic encoding an anti-claudin 18.2 IgG, from pre-clinical models to human predictions.

Methods: We developed a multiscale mechanistic model capturing key processes: (i) LNP-mediated delivery and endocytosis via LDLR pathways, (ii) endosomal escape and mRNA release, (iii) cytoplasmic mRNA translation into IgG, (iv) IgG systemic distribution and target binding, and (v) transient increase in cytokines because of exogenous mRNA. Model development was grounded in published in-vitro, mouse and NHP PK-PD data1. Allometric scaling principles and literature-informed differences in LDLR expression and endocytosis kinetics were used for human translation. The model was calibrated using Monolix 2023R1. Sensitivity analysis identified key translational bottlenecks.

Results: The model adequately captured the time course of mRNA, expressed IgG, and transient increase in cytokines in mice and NHP following intravenous mRNA administration. The rate of degradation of IgG was 2.87 1/h (4.28% RSE), The rate of endosomal escape was 231 1/h (10.9% RSE), the rate of degradation of mRNA was 0.114 1/h (6.2% RSE) for mouse and NHP when fitted simultaneously. All the physiological volumes were fixed from literature2. Increase of cytokines in plasma was characterized by using transit compartment model in which Emax was 2390 (2.68%), EC50 was 15 ng/mL (1.85% RSE) and tau was 2.43 1/h (0.93% RSE). The model was able to simulate the %Receptor Occupancy and plasma PK in humans based on the simulations 0.025 mg/kg was selected for the starting dose for first-in-human dose. Sensitivity analysis revealed that degradation of IgG and mRNA were critical determinants of interspecies translation.

Conclusions: This study presents a comprehensive multiscale PK-PD framework to characterize and predict the behavior of mRNA-encoded therapeutic antibodies from preclinical systems to humans. Application to an anti-claudin 18.2 IgG mRNA therapy highlighted species-specific differences in nanoparticle processing and translation kinetics, providing a rational basis for dose selection in early clinical trials. This modeling framework can be extended to other mRNA-based protein therapeutics to improve translational predictability and accelerate clinical development3.

Citations:

References:

1. Mahmud et al. Oncoimmunology. 2023 Oct 16;12(1):2255041
2. Desai DA et al. Front Pharmacol. 2024 Sep 20;15:1454785.
3. Kim SJ et al. Clin Transl Sci. 2024 Dec;17(12):e70101.

Keywords: First-in-human dose, Gastroesophageal cancers, Model informed drug development, Gene therapy