(M1130-01-02) Multiscale Pharmacokinetic Modeling and Simulation of Local and Systemic Exposure of Intramuscularly Injected LNPs

Purpose: The lipid nanoparticle (LNP) construct is widely utilized for drug delivery, especially achieving great success in stabilizing and delivering mRNAs for intended vaccination and therapeutic purposes. All currently approved mRNA vaccines for COVID-19 adopt LNPs as the delivery vehicle that are given via IM injection. The vaccines have shown strong immune responses. There is however a large knowledge gap in the mechanistic understanding of IM transport and uptake kinetics of injected lipid nanoparticles. In particular, there is little effort in capturing local absorption and presystemic catabolism and subsequently predicting systemic exposure of LNPs, limiting clinical efficacy and safety assessments. Herein, we report a multiscale modeling and simulation approach to predict local and systemic exposure of intramuscularly injected LNPs. The modeling framework combines multiphysics simulation of drug transport and absorption at the injection site with systemic PK modeling of systemic biodistribution.

Methods: A multiphysics simulation model of drug transport and absorption of SC-injected biologics was previously developed by our group (1, 2). We extended the model to simulate IM injection and absorption of LNPs. In the model, IM tissue is treated as a homogenous, poroelastic media, in which hydrodynamics of the interstitial fluid is evaluated according to the poroelasticity and Darcy’s law. The model parameters used in the study are listed in table1. Mass transport of LNPs is described by general convection-diffusion equation (1), with release kinetics (2) and uptake (3) also being considered:
∂(φC_im )/∂t+∇∙(-D_eff·∇C_im )+∇(v ∙C_im )=-P_a·C_im·S_ly/V  (1)
(∂(φC_rel))/∂t=K_rel·C_im  (2)

(∂(φC_cell))/∂t=K_in·C_im-K_out·C_cell  (3)
The terms on the left side of Eq. 1 represent mass accumulation, and diffusive and convective transport, respectively. φ is tissue porosity; C_im is the LNP concentration in the IM tissue; D_eff is the effective diffusion coefficient of the drug in tissue; V is the interstitial flow velocity. The right side of Eq. 1 represents the drug uptake rate by lymphatic vessels. P_a is the apparent permeability of the drug across the lymphatic tissue volume. Vasculature absorption of large biomolecules is generally considered trivial from the injection site and ignored in this study. In addition, the release kinetics and uptake by myocytes in the local tissue are described by Eqs.2 and 3, respectively. C_rel is the disintegrated LNP concentration in muscle tissues; K_rel is the release rate constant; C_cell is the LNP concentration in myocytes; K_in is the uptake rate constant; K_out is the effluent out rate constant. By numerically solving these equations with inputs of physical properties of LNPs and the tissue attributes, the spatiotemporal distribution of the drug concentration, along with other physical quantities, can be calculated.
The overall absorption kinetics of LNP from the lymphatic capillaries at the injection site can then be obtained by integrating over the entire model domain of the porous medium:
m_abs=-∭(φ(C_im-C_cell-C_rel)dV)
k_abs=(dm_abs)/dt
Where m_abs is the absorbed drug amount and k_abs is the local absorption rate. V and t stand for the tissue volume and time correspondingly. The absorption rate is further fitted by an exponential function of time (m_LNP=Dose∙e^(a∙t)) to be integrated into the PK model.

Results: Fig. 1 shows the kinetic profiles of simulated and experimentally measured LNPs amounts remaining in the injection site after IM administration. The experimental data were obtained by measuring deuterated cholesterol labeled LNP for 24 hours (3). The normalized root-mean-square error (NRMSE) between the simulated and observed data is 0.14, suggesting good fitting. Based on the results of the multiphysics simulation of local absorption, we integrated the local absorption kinetics to a whole-body PK model. Fig. 2 shows the comparison of system exposure between our simulation and a literature report of experiment data (4). IV data was used to calculate the system PK parameters, and further used to simulate the IM PK profile. The mean squared error (MSE) of the simulated line is 6.30×10^-6.

Conclusion: In this study, a multiscale simulation and modeling framework was developed to compute the systemic exposure of IM administrated LNPs. The method integrates the multiphysics simulation of drug absorption kinetics at the injection site with the whole-body PK modeling of systemic disposition. Knowledge gained from this study would shed light on the impact of tissue properties, drug characterization on drug PK, further guiding the LNPs design.

References: 1. Zheng F, Hou P, Corpstein CD, Xing L, Li T. Multiphysics Modeling and Simulation of Subcutaneous Injection and Absorption of Biotherapeutics: Model Development. Pharmaceutical Research. 2021;38:607-24.
2. Zheng F, Hou P, Corpstein CD, Xing L, Li T. Multiphysics modeling and simulation of subcutaneous injection and absorption of biotherapeutics: model development. Pharmaceutical research. 2021;38(4):607-24.
3. Di J, Du Z, Wu K, Jin S, Wang X, Li T, et al. Biodistribution and non-linear gene expression of mRNA LNPs affected by delivery route and particle size. Pharmaceutical Research. 2022;39(1):105-14.
4. A Tissue Distribution Study of a [3H]-Labelled Lipid Nanoparticle-mRNA Formulation Containing ALC-0315 and ALC-0159 Following Intramuscular Administration in Wistar Han Rats.2020;1-77


Table 1. Tissue and Drug parameters used in the simulation


Fig. 1 Comparison of simulated distribution with experimental data in muscle tissue. The blue dots (mean ± error) is experimental data of local LNPs amount from our previous work (3). The green line (C_m) indicates the time profile of interstitial LNPs amount. The purple line (C_cell) indicates the time profile of intracellular LNPs amount. The orange line (C_rel) indicates the time profile of disintegrated LNPs amount. The red line (C_total) indicates the total local LNP including interstitial LNPs, intracellular LNPs and disintegrated LNPs.


Fig. 2 Simulated and observed system PK profiles of IV administration (a) and IM administration (b) for LNPs.

Acknowledgements: JD acknowledges the Shanghai Jiao Tong University Zhiyuan Honorary Scholarship. TL thanks Allen Chao Endowment.