Development of an in vitro pharmacokinetic model of the human eye

Abstract

Purpose: Pharmacokinetics remains a major challenge in drug development for the posterior segment. Accurate evaluation and optimisation of vitreous drug distribution is hindered by differences between humans and animal platforms. Furthermore, measurements at serial time points necessitate use of many animals, which has both cost and ethical implications. We aimed to design an in vitro flow model that would mimic posterior segment fluid dynamics, and aid the determination of vitreous drug distribution. Similar to the design of simulated gastric solutions to mimic dissolution within the intestinal tract, we are also focused on developing a simulated vitreous solution that can be used in early preclinical studies.

Methods: Several prototype chambers (~4.2 mL internal volume ) were fabricated incorporating both a model anterior and posterior segment. These were filled with simulated vitreous (a polymeric combination of hyaluronic acid and agar) replicating the viscosity and rheology of human vitreous. Aqueous flow input ports and an output sampling port were used to provide fluid flow by introducing a 2.0 μl/min flow rate with phosphate buffered saline (pH 7.4) from the anterior segment at room temperature (25°C). Dexamethasone sodium phosphate (1.0 mg) was solubilised in distilled water (1.0 ml) and this solution (50 μl) was injected into the model. Samples (100 μl) were taken over 5 days at 24 hourly intervals from different anatomical regions. Analysis was performed using High Performance Liquid Chromatography (HPLC) at 240 nm.

Results: The viscosity properties of the hyaluronic acid-agar combination were found to be similar to that reported for human eyes. This solution is thought to be a reliable simulation of in vivo viscosity that can occur. A concentration of 1.03 mg/ml (±0.06 mg/ml) dexamethasone sodium phosphate was injected into the model and sampled for 5 days. It was sampled at various points from the model and release kinetics showed it was within the therapeutic range (10-100 uM).

Conclusions: This simple dynamic posterior segment model may aid the pharmacokinetic study of ocular drugs in early preclinical development. Further studies to incorporate saccadic movements, different forms of dexamethasone, temperature conditions and in vivo-in vitro correlations are required including testing the release kinetics of other anti-inflammatory and anti-fibrotic drugs.

Keywords: 763 vitreous • 765 wound healing • 497 development