In vitro and in vivo evaluation in rabbits of a controlled release 5-fluorouracil subconjunctival implant based on poly(D,L-lactide-co-glycolide)

Purpose. To design a controlled release 5-fluorouracil (5-FU) implant to provide prolonged antifibroblast concentrations of 5-FU in the subconjunctival tissues but low concentrations of 5-FU in other ocular tissues. Method. Implants (5 mg; 2.5 mm diameter x 1.2 mm thickness) of 5-FU or 9:1, 8:2, 7:3 5-FU to polymer mass ratios were made by compression. Poly(D,L-lactide-co-glycolide) polymers with 50:50 and 75:25 lactide to glycolide ratios were used. In vitro release characteristics of four types of implants were studied: 5-FU alone (CT), 5-FU/polymer matrices (MT), coated 5-FU/polymer matrices with a central hole drilled through the matrix and coating (CM1), and with a central hole in the coating (CM2). MT and CM1 (9:1 drug/polymer) were selected for subconjunctival implantation in rabbits. C-14-5-FU levels in various ocular tissues and retrieved pellets were monitored. Results. First-order release was observed from CT, MT and CM1 implants. Zero-order release profiles were observed from CM2 implants. Drug release was retarded by formulating 5-FU in a matrix comprising poly(D,L-lactide-co-glycolide) which in turn could be modified by the lactide/glycolide ratio of the polymer, the drug to polymer ratio, coating, and hole dimensions: First-order release kinetics were observed for MT and CM1 implants in the in vivo study in rabbits. A correlation between in vitro and in vivo release was established which allowed in vivo release to be predicted from in vitro release data. For CM1, therapeutic tissue concentrations of 35.2 mu g/g (conjunctiva) and 17.7 mu g/g (sclera) were found at the implantation site up to 200 hours post-implantation. Tracer levels were undetectable in other ocular tissues. Conclusions. The CM1 implant maintained steady antifibroblast levels in target tissues and minimized levels in nontarget tissues.