(T1430-02-12) Ocular Pharmacokinetic Model for Intravitreal Antibody Injection

Purpose: Intravitreal injection has become a vital route of administration for delivering drugs directly to the posterior segment of the eye. Understanding the pharmacokinetic (PK) profile of drugs following intravitreal injection is crucial for optimizing dosing regimens and ensuring therapeutic efficacy. Here, we present our work on the PK modeling of intravitreal injections.

Methods: We tested an ordinary differential equation (ODE)-based mechanistic model covering essential ocular compartments. The model consists of three ocular compartments, including retina, vitreous humor, and aqueous humor. The model also includes a two-compartmental model for systemic PK. The model has target-mediated drug disposition (TMDD) model components in each compartment to account the target effect. We obtained all parameters from literature and empirical correlations.  We also try calibrating some parameters by fitting to improve the model predictability. We tested publicly available data sets of antibodies. We trained and validated the model using preclinical data, including those of rabbits and cynomolgus monkeys. The data include concentration profiles of ocular and systemic compartments.

Results: Our pre-clinical data are from rabbits and monkeys. The data of rabbit studies usually contains the most comprehensive information, while the data of monkeys are limited. The model generally performs well. Even without fitting to our data sets, the simulation results can capture the trends and well predict the values. The model uses many parameters. We collected parameter values from different sources. One challenge is the physiological parameters, such as ocular compartment volumes. Due to the differences in the test strain, these parameters have a large variability for each species. Our results show that the differences in these parameters can explain the differences in some studies. One critical parameter in the model is the vitreous humor half-life. In most cases, as well as in our model, the vitreous humor is treated as one compartment. The vitreous humor half-life itself describes the slope of the concentration profile. Literature have shown an obvious correlation between the molecular weight and the half-life in each species. The physiology of the eye, such as shape and size, should contribute to the species half-life differences. However, there has been no report to successfully correlate the physiological properties to the half-life. We currently find the empirical correlations works the best in the prediction. If the data are available, the vitreous humor half-life can be obtained by fitting to the data to obtain the best match. The target existence usually does not affect the ocular PK, because the concentrations of common targets are usually much lower than their antibody concentration. The target concentrations are however still important in the calculation of target engagement. TMDD components can be added based on the physiology and the needs.

Conclusion: The model successfully described all data sets. The model shows the potential to become a general framework for ocular and systemic PK. The results also show that a priori estimation based on the MW is possible. The model can provide guidance on PK even without any in vivo data. Any in vivo data can increase the model accuracy, especially of systemic PK. Our modeling practice shows an example of workflow to conduct preclinical PK studies and guide human dose selection.

References: [1] Hutton-Smith, L. A., Gaffney, E. A., Byrne, H. M., Maini, P. K., Gadkar, K., & Mazer, N. A. (2017). Ocular pharmacokinetics of therapeutic antibodies given by intravitreal injection: estimation of retinal permeabilities using a 3-compartment semi-mechanistic model. Molecular pharmaceutics, 14(8), 2690-2696.
[2] Fuchs, H., & Igney, F. (2017). Binding to ocular albumin as a half-life extension principle for intravitreally injected drugs: evidence from mechanistic rat and rabbit studies. Journal of Ocular Pharmacology and Therapeutics, 33(2), 115-122.
[3] Fuchs, H., Chen, L. Z., Low, S., & Yu, H. (2021). Ocular and systemic pharmacokinetics of BI-X, a nanobody targeting VEGF and Ang-2, after intravitreal dosing in cynomolgus monkeys–Evidence for half-life extension by albumin. Experimental Eye Research, 205, 108486.
[4] Caruso, A., Füth, M., Alvarez-Sánchez, R., Belli, S., Diack, C., Maass, K. F., ... & Mazer, N. A. (2019). Ocular half-life of intravitreal biologics in humans and other species: meta-analysis and model-based prediction. Molecular Pharmaceutics, 17(2), 695-709.

Model framework for a general antibody following an IVT injection. There are three ocular compartments and two systemic compartments. Transport directions are shown by arrows. Transport applied for the antibody, the target, and the antibody-target complex, which are not shown individually.

Model components of TMDD. The target may exist in certain but not all ocular and systemic compartments. The synthesis and degradation of the target may occur only in certain compartments. The elimination of the antibody and the complex may also occur only in certain compartments.

Simulation results for Ab1 (26kDa) with all parameters from the literature