Abstract
Purpose :
Therapeutic effectiveness of intravitreally administered drugs relies on their distribution around and elimination from the intraocular space. Unfortunately, currently used models fail to adequately predict human intravitreal pharmacokinetics. This is in part due to large inter-individual variations in vitreous structure and an inability to control physiological factors such as vitreous liquefaction, intraocular flow and saccade in these models. In this study, an in vitro set up was designed to model human intravitreal pharmacokinetics. Different physiological conditions were built into the model to evaluate their impact on model drug distribution and elimination from the set up.
Methods :
A Perspex mould was designed using dimensions of the human eye. Synthetic vitreous was prepared using various combinations of hyaluronic acid (HA) and agar to achieve a range of viscosities. Intraocular flow through the model was controlled with an Ismatec pump and saccade was simulated by placing the model in a rocking hybridisation oven at 37 °C and 50 rpm. Pharmacokinetic parameters of the model were assessed using fluorescein as a model drug. Following intravitreal injection of the dye under different conditions, aliquots were withdrawn from the set up at different time points and assayed using microscopic or spectroscopic assays. Diffusion and elimination were compared to data attained using ex vivo porcine eyes.
Results :
A 2 µl/min intraocular flow rate rapidly distributed dye throughout the model irrespective of vitreous viscosity. In absence of simulated flow, vitreous with the highest concentrations of HA (5 mg/ml) and agar (4 mg/ml) demonstrably slowed dye diffusion; however, no differences were seen in diffusion properties of any of the remaining preparations (HA 0.7-2.5 mg/ml; agar 0.95-2 mg/ml). Simulated saccade did not influence dye migration when using synthetic vitreous; however, migration was accelerated when phosphate buffered saline was used in place of vitreous. Dye distribution was significantly more rapid through synthetic vitreous in the phantom eye than ex vivo porcine eyes.
Conclusions :
Intraocular flow is the primary driving force for small hydrophilic molecule migration around our eye model, and is able to override the impact of both vitreous viscosity and saccade. Further studies will assess pharmacokinetics of larger molecules using our model, and will also compare model behaviour to obtained human data.
This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.