Sustained-release endotoxin. A model for inducing corneal neovascularization.

Abstract

The rabbit corneal pocket assay is one of the most frequently used systems for the study of angiogenesis. This model particularly is useful to identify stimulators of new blood vessel formation. More recently, however, interest in inhibitors of angiogenesis has grown, and several antiangiogenic agents have been identified. Investigations of angiogenesis inhibitors require a reliable model for the stimulation of neovascularization. One method was modified to produce corneal neovascularization by implanting into the rabbit cornea a sustained-release polymer containing endotoxin (Escherichia coli lipopolysaccharide). The implant was prepared by mixing weighed quantities of endotoxin with ethylene vinyl acetate copolymer (Elvax) and forming 1-mm3 pellets containing 1%, 7.5%, 15%, 20%, 30%, and 40% (w/w) of endotoxin. Pure Elvax pellets were implanted as controls. Intrastromal corneal pockets were created in 92 eyes of male, albino New Zealand rabbits (n = 80), and sterilized endotoxin-copolymer implants were introduced. The growth rate of new vessels was measured by slit-lamp biomicroscopy. Endotoxin loads of 15% (n = 40) produced a strong neovascularization response with minimal stromal edema, with a mean growth rate of 0.21 +/- 0.12 mm/day. Loads of 1%, 7.5%, and 20% yielded 0.1 +/- 0.03 mm/day, 0.27 +/- 0.05 mm/day, 0.30 +/- 0.06 mm/day, respectively (n = 8, each group). Higher loads (30% and 40%; n = 8, each group) produced intense neovascularization, accompanied by severe corneal edema that obscured accurate measurement of the vessels. Corneal pockets that did not contain polymer implants were avascular. When endotoxin-Elvax pellets were removed, the new vessels regressed within 2 weeks.