(M1530-02-12) Multiphysics Modeling and Experimental Design of Novel Ocular Drug Delivery Systems for Glaucoma Treatment

Purpose: One of the primary reasons for elevated pressure in Glaucoma is a reduction in permeability of Trabecular Meshwork (TM) for Aqueous Humor (AH) drainage. The drug released from Intracameral implants diffuses through AH to TM. The nature of implants’ shape, size, material properties, placements and AH flow structures can heavily influence the reach of the drug to TM. Thus, prior knowledge of AH flow field and concentration field is a must for implant-related design, planning and optimization. In this work, we aim to develop a multiphysics model that predicts intraocular pressure and use it to guide the experimental design of a novel sustained drug release ocular implant that utilizes the haptic region of the intraocular lenses.

Methods: Multiphysics model is developed in COMSOL® for AH flow from the posterior chamber to the anterior chamber of the eye. The axisymmetric, realistic geometry of the chambers is constructed from geometrical information available in the literature [1]. AH is treated as Newtonian fluid and the TM is modeled as a homogeneous porous media. The healthy and Glaucoma conditions are simulated by taking appropriate permeability values of TM. We then simulated the effect of AH hydrodynamics on the drug release from an ocular implant of different sizes (implant-1 and implant-2) and effective drug diffusivities (case 1-5) as shown in Fig. 1a. In order to obtain the model parameters and guide the design of a novel ocular delivery system, we have performed experimental studies on drug-loaded microspheres that can be potentially used in the haptic region of intraocular lenses to attain near zero-order drug release. PLGA microsphere loaded with ∼0.5% (w/v) timolol maleate and Poloxamer hydrogel was prepared using a double-emulsion-solvent evaporation method for possible usage in an ocular implant. The microsphere was evaluated for its microscopic structure, physicochemical properties, and release mechanism.

Results: Our simulations show around one order of variations in concentration within the anterior and posterior chambers. The drug concentration at TM and the average concentration left inside the implant are tracked with time. The drug concentration at TM shows a burst release peak in the initial few hours. The peak concentration value at TM and the corresponding time are shown for five different cases and two implants in Fig. 1(b)-(c). The peak concentration reduces with a decrease in diffusivity. However, the rate of decrease in peak concentration with effective diffusivity is different for both implants. The curves cross-over each other from case 3 (~1E-12 m²s^-1). The implant-1 experiences higher burst release than implant-2 for higher effective diffusivity values inside the polymer matrix and vice-versa. As effective diffusivity decreases the extent of burst release becomes less in implant-1 than implant-2 due to its higher thickness and thus more time is taken by the drug to reach from core to the surface. Similar to therapeutic lower limit, a cutoff value is set as 0.01μg/mL or 3.2E-5 mol/m³ for getting an estimate of total duration for which drug is available in sufficient amounts at TM. Drug availability based on this cutoff varies from 18 hours to 15 days according to case configuration. In the experiments, PLGA microspheres were spherical, with a mean particle size of 10.04 ± 1.89 μm. Scanning electron microscopy analysis showed that the microsphere had a core-shell Di-Depot structure, with multiple TM-loaded hydrogel depots distributed in the core region. The interaction between the hydrogel and PLGA depots resulted in a high encapsulation efficiency and low burst release, allowing for a prolonged administration interval of up to 49 days.

Conclusion: We simulated AH flow in an axisymmetric eye model and successfully validated the IOP values. Next, we studied drug delivery from two implants. Despite the same amount of drug loaded in the implant, the burst release and drug availability (above cutoff value) depend on the effective diffusivity. Minimization of burst release required optimization between drug implant volume and effective diffusivity. We intend to use the vitro experiments to provide the model parameters for our computational in-vivo model, which may help us further optimize implant design.

References: 1. Sánchez, G. M., Del Pozo, C. E., Medina, J. R., Naude, J., & Solorzano, A. B. (2020). Numerical simulation of the aqueous humor flow in the eye drainage system; a healthy and pathological condition comparison. Medical Engineering & Physics, 83, 82-92.
2. Pimenta, A. F., Serro, A. P., Colaço, R., & Chauhan, A. (2018). Drug delivery to the eye anterior chamber by intraocular lenses: An in vivo concentration estimation model. European Journal of Pharmaceutics and Biopharmaceutics, 133, 63-69.

Acknowledgements: This work is funded by the Science and Engineering Research Board (SERB), India grant CRG/2022/003774.

(a) Velocity contour along with streamlines is shown on axisymmetric geometry including anterior chamber, posterior chamber, Trabecular Meshwork and implant. (b) Peak concentration observed at Trabecular Meshwork for all five cases and two implants (c) Time at which peak concentration is occurring at Trabecular Meshwork.