(M-100) A Quantitative Systems Pharmacology Model of LNP-mRNA Uptake Through LDLR Binding

CJ Musante, PhD – Head of Pharmacometrics & Systems Pharmacology, Translational Clinical Sciences, Pfizer; Nessy Tania, PhD – Senior Principal Scientist, Translational Clinical Sciences, Pfizer

Objectives: Lipid nanoparticle (LNP)-mRNA therapeutics have had success as a platform for vaccines. However, several hurdles remain in developing LNP-mRNA therapies for systemic delivery, including a lack of understanding of how interindividual variability in underlying mechanistic processes translate to variability observed in clinic. IV delivery of current generation LNPs rapidly accumulate in the liver, where they undergo receptor-mediated endocytosis via binding with low-density lipoprotein receptors (LDLR) [1]. In this work, we develop a quantitative systems pharmacology (QSP) model of the mechanistic processes from LNP-mRNA uptake by hepatocytes through translation of the mRNA into therapeutic protein. By simulating the impact on different aspects of the LDL/LDLR pathway, we can predict the effects of elevated LDL levels, familial hypercholesteremia, and statin use on exposure and efficacy of LNP-mRNA therapies.

Methods: Ordinary differential equations are used to capture LNP, LDL, LDLR, mRNA, and protein dynamics in plasma, extracellular, membrane, and cellular compartments. Each step in the pathway is described using mass-action kinetics. Intracellular kinetics and binding affinities are calibrated to in vitro data. The model is simulated under conditions mimicking healthy controls or individuals with familial hypercholesteremia, elevated plasma LDL, or statin use. In each case, steady-state conditions are found by simulating the system in the absence of LNPs. Using these states as initial conditions, a single IV dose of 30 µM LNP-mRNA is simulated for 24 h. Plasma and cellular LNP, mRNA, and protein concentrations are compared across the tested parameter ranges.

Results: Differences in the internalization rate of LDLR demonstrate the greatest effect on protein concentration. Additionally, LDLR internalization rate has the greatest effect on plasma LNP exposure. While no differences are observed in mRNA or protein concentrations when varying the recycle fraction of LDLR, LDL plasma concentration, and LDLR generation rate, all have a large impact on the plasma exposure. Transient elevation in plasma LDL levels are also observed with LNP-mRNA treatment, however, levels return to baseline within hours of dosing.

Conclusions: A QSP model is presented that captures underlying mechanisms of IV delivery of LNP-mRNA therapeutics. Inclusion of the LNP/LDLR uptake process enables simulation of LNP-mRNA treatment in various clinical subpopulations. Initial simulations suggest that protein concentrations of LNP-mRNA therapies is greatly impacted by the internalization rate of LDLR – a process affected by some classes of familial hypercholesteremia. While other processes in the LNP/LDLR uptake pathway have little impact on efficacy, they do show higher levels of plasma exposure. Future work will focus on calibration of the model to clinical PK and inclusion of a non-specific uptake pathway to account for a second, slower phase of clearance.

Citations: [1] Akinc, A., et al., Targeted delivery of RNAi therapeutics with endogenous and exogenous ligand-based mechanisms. Mol Ther, 2010. 18(7): p. 1357-64.