Abstract
Purpose: :
Ocular surgery, including glaucoma filtration surgery (GFS), often results in inflammation and fibrosis. 5- Fluorouracil (5FU) and mitomicin C interfere with the proliferative phase that causes fibrosis while steroids (dexamthasone and triamcinolone) modulate the inflammatory phase. These medicines are usually administered by injection or eye drops. However, the local tissue pharmacokinetics of the drugs are suboptimal due to rapid clearance. We have developed tablets that may be implanted intraoperatively to prolong the exposure time of medicines. As the tablets would be placed under non-sink conditions, we have examined their release profiles using an in vitro and a computational model.
Methods: :
The compounds evaluated included 5FU, dexamethasone (DEX), triamcinolone (TRI), and naproxen (NAP). Tablets (3 mm diameter) were fabricated and the release profile was evaluated using a dispensing pump that supplied flow rates ranging from 2-20 µl/min through either a 50 or 200 µl flow chamber. The released amount of drug obtained at different time points from the out flow of the chamber was determined by HPLC. The correlation between the mathematic model and experimental data was investigated using the Pearson product-moment correlation coefficient.
Results: :
Approximately 8% of TRI was released during a 5 day period while maintaining a concentration of 20-30 µg/ml. Meanwhile less than 5% of DEX was released with a concentration of 10-20 µg/ml. In contrast more than 80% of NAP was released with a concentration of 200-500 µg/ml. With a constant flow rate (2 µl/min), the compounds were found to release faster if the tablets were placed in the 50 µl chamber. A mathematical model based on mass, size, and solubility characteristics of the tablet was developed to correlate with the experimentally derived release profiles. There was a strong, positive correlation (correlation coefficients r2 > 0.8) between the mathematical model and the experimental data.
Conclusions: :
Our computational model has a high correlation with the in vitro model. Both of these models are designed to predict the release profiles of tablets in non-sink conditions. These models may be useful in developing appropriate prolonged release therapies in the eye to prevent scarring.
Keywords: drug toxicity/drug effects • computational modeling • wound healing