Angiostatin inhibits and regresses corneal neovascularization

Objective: To determine the ability of angiostatin and the angiostatin-producing low-metastatic (LM) clone of Lewis lung carcinoma (LLC) to inhibit and regress corneal neovascularization, as compared with the nonangiostatin-producing high-metastatic (HM) clone. Methods: Three groups of C57BL6/J mice underwent chemical and mechanical denudation of corneal and limbal epithelium. One group remained tumor free while the other 2 were implanted with LLC cells (either the HM or LM clones) subcutaneously the day before, 2 weeks after, or 4 weeks after denudation. Corneas were harvested 2 weeks after tumor implantation (at 2, 4, and 6 weeks after denudation for tumor-free mice). Neovascularization was quantified by CD31 immunostaining. In a second experiment, recombinant angiostatin was delivered continuously for 2 weeks via an osmotic pump in mice with established corneal neovascularization. Results: The mean percentages of neovascularized corneal area in mice 2 weeks after LM-LLC implantation were 4.6%, 3.7%, and 37.0%, at 2, 4, and 6 weeks after scraping, respectively. In contrast, in the mice implanted with HM-LLC, the corresponding values were 45.4% (P=.01), 90.1% (P=.03), and 80.3% (P=.005). For tumor-free mice, the corresponding values were 62.0% (P=.003), 68.9% (P=.03), and 59.3% (P=.06). Mice implanted with angiostatin pumps had a 37.7% neovascularized corneal area 2 weeks after implantation and 4 weeks after scraping while mice implanted with sham pumps had 60.5% (P=.007). Conclusion: Angiostatin inhibits and regresses corneal neovascularization induced by mechanical and alkali corneal injury. Clinical Relevance: This appears to be the first evidence of biologically induced regression of corneal neovascularization, and the first direct demonstration of angiostatin-induced regression of neovascularization in any tissue.