3D Computational Fluid Dynamic Simulation of an Intravitreal Brimonidine Implant in the Rabbit Eye

Abstract

 

Purpose

 

Drug delivery to the posterior segment of the eye is challenging. Traditional methods of systemic and topical delivery are ineffective due to numerous anatomical barriers (e.g. blood-retina barrier, resistance of corneal epithelium, and rapid elimination from aqueous humor). Intravitreal sustained-release implants provide advantages by avoiding these barriers and reprieving the frequent dosing burden. A mathematical model was developed to predict drug distribution following an intravitreal dose of a sustained-release implant.

 

Methods

 

An anatomically accurate 3D rabbit eye was developed to simulate the advection-diffusion of brimonidine from an implant in the vitreous. The diffusion coefficients of the ocular tissues were defined using published ganciclovir data, which have similar physicochemical properties to brimonidine. The similarity was verified by experimental distribution profiles following an intravitreal injection of a bolus mixture of the two compounds into rabbits. The elution profile of brimonidine was used to calculate a flow rate and mass flux across the implant domain. The simulation was compared to vitreous and retina concentrations from a pharmacokinetic rabbit ocular study with a sustained-release polymeric implant containing brimonidine. A parameter sensitivity analysis was conducted to examine the affect of different release rates and implant location on drug distribution.

 

Results

 

The simulated concentration-time profile was in agreement with the measured tissue concentrations using a scaling factor. The concentration magnitude varied significantly between elution profiles and the localized concentration distribution was dependent on the implant location. This suggests that the release rate and implant location are important factors when developing a sustained-release implant.

 

Conclusions

 

Computational fluid dynamic modeling is a valuable tool to predict ocular pharmacokinetics. Parameter sensitivity analysis highlighted the importance of initial location and size of the implant in determining drug distribution.

 

 

Observed and predicted brimonidine concentrations in the retina and vitreous. A scaling factor of 200 was applied to the simulation data.

 

 

Symmetry plane contour plots of the simulated brimonidine concentration (scalar) at t=6 days post dose. For implants with an average release rate of 1.5 μg/day (left), 3 μg/day (middle) and 20 μg/day (right).

 

Keywords: 763 vitreous • 688 retina • 473 computational modeling