Choroidal neovascularization (CNVM) represents the most common cause of severe vision loss in patients with age-related macular degeneration (AMD). It is also the most common cause of legal blindness in the United States in patients over the age of 60. Therefore, the study of the diagnosis, prevention and treatment of this condition would have a major impact on the public health. The laboratory has been developing a clinically relevant model of CNVM associated with AMD through the use of an adenovirus overexpression system. Based on evidence that vascular endothelial growth factor (VEGF) is upregulated in eyes with CNVM associated with AMD, we successfully reproduced this finding by overexpressing VEGF in the retina pigment epithelium through the use of subretinal delivery of an E-1deleted adenoviral vector encoding VEGF. Animals receiving this vector developed CNVM in 2 ? 4 weeks with a histologic appearance similar to that seen in humans. The phenotype included penetration of choroidal neovasularization through Bruch?s membrane with sparing of the inner retinal structures. The cellular composition of these lesions which is composed of endothelial cells and pericytes was confirmed by lectin staining, immunohistochemistry and electron microscopy. The location and vascular channel composition was verified by fluorescein angiography and flat mount histology of fluorescein-dextran injected animals. Work has begun on utilizing this model to study the pathogenesis of the condition and to screen for possible treatments. The histology of CNVM associated with AMD is also characterized by an influx of macrophages but an absence of other inflammatory cells. This finding is also seen in vessel intimal neovascularization associated with atherosclerosis and may represent a similar pathogenic altered wound healing mechanism. The laboratory is exploring this concept in our animal model by defining, by immunohistochemistry, the presence of inflammatory cells recruited to the site. Evidence has been collected which demonstrates that there is a marked increase in the number of recruited macrophages at the site of CNVM in these animals but an absence of CD4, CD8 and resident macrophages. These findings were verified by producing CNVM in nude animals in which there is an absence of lymphocytes. Experiments are now underway to determine the relative contribution of macrophages to the development of CNVM. Work has also progressed on screening treatment strategies for CNVM using this model. 2-methoxyestradiol, a potent anti-angiogenic and antiproliferative agent, delivered through a sustained novel rat intravitreal device, appears to prevent the development of CNVM in these animals. In addition, the drug appears to be safe with absence of detectable retinal toxicity as measured by histology and electroretinography in rabbits. These results indicate that this treatment approach may have a clinical application and experiments are continuing to confirm and extend these results with the goal of initiating a Phase I clinical trial in patients with CNVM associated with AMD.
