Abstract
Purpose:
An in vitro model eye may be of use for determining ocular drug pharmacokinetics and pertinent to drug development. We aimed to examine whether therapeutic protein pharmacokinetics in our in vitro model correlated with current in vivo data.
Methods:
Plastic globes (~4.0 mL internal volume) were fabricated incorporating both a model anterior and posterior segment. These were coated with albumin by incubating in albumin solution (0.1%) at 37°C for 12 hours and then rinsed with distilled water. Phosphate buffered saline (PBS) flowing at a rate of 2.0 μl/min passed from input ports in the anterior segment to sampling ports in the anterior and posterior segment (pH 7.4, 25°C). Anti vascular endothelial growth factor antibody (anti-VEGF) was injected (100 μl) intravitreally into the posterior part of the model and compared to sham control (no protein present). Concentrations at different time points from both posterior and anterior segments were determined by Bradford assay. Data was recorded in triplicate and displayed as mean and standard deviation.
Results:
Both ranibizumab and bevacizumab showed a monoexponential decay in the posterior part of the model, which followed first-order kinetics as drug concentration in the model approached zero. Drug content vs. time plotted semilogarithmically found ranibizumab and bevacizumab to have a half-life of 5.46 days and 6.82 days respectively
Conclusions:
The pharmacokinetic profile of anti-VEGF injections in this model correlates closely with human data (reported half-lifes of ranibizumab and bevacizumab in humans are 7.19 days and 6.7-9.82 days respectively). This has implications for facilitating development of posterior segment protein therapeutics.
Keywords: 497 development •
763 vitreous •
765 wound healing