Repeated retinal photocoagulation in monkeys for the optimization of a laser-induced choroidal neovascularization model

Abstract

The laser-induced choroidal neovascularization (CNV) model in nonhuman primates has played a critical role in the development of new therapies for neovascular age-related macular degeneration. The widespread use of this model, however, has been limited by its high costs, mainly due to the lower efficiency of animal use. We optimized the CNV model by administering repeated photocoagulation treatments to the same eye of each animal, and preliminarily evaluated this model using an assessment of the efficacy of an anti-vascular endothelial growth factor (VEGF) agent to address this problem. Seven rhesus monkeys were included and divided into two groups, which were named the laser-only and laser-bevacizumab groups. Each animal underwent 3 retinal photocoagulation sessions in the same eye at 4-week intervals to induce CNV. Three weeks after the first laser treatment, the animals in the laser-bevacizumab group were administered an intravitreal injection of bevacizumab. Fluorescein angiography (FA) was performed in all animals at multiple time points within 12 weeks to assess the severity and development of CNV following each laser treatment. The laser lesions produced in each photocoagulation session were analysed separately using grading and densitometry methods, and CNV severity was represented by the CNV incidence and the mean integrated fluorescence intensity (MIFI), respectively. Our results showed that in the animals in the laser-only group, the average CNV incidence rates were 62.5%, 42% and 50% at 2 weeks after each laser treatment, and the average MIFI values (x105) were 3.83 ± 2.36, 2.66 ± 1.42 and 2.52 ± 0.18, respectively. No significant differences were found among treatments. After week 2, the CNVs progressed or regressed continuously over 2-6 weeks before stabilization, and the time course of CNV development in each animal was generally the same after each photocoagulation session. In the laser-bevacizumab group, however, the average CNV incidence rates of each laser treatment at week 2 were 50%, 0 and 37.5%, respectively, and the average MIFI values were 3.79 ± 0.47, 1.09 ± 0.35 and 2.37 ± 1.35, respectively. The differences between treatments 1 and 2 were statistically significant. Meanwhile, the CNVs induced by laser treatment 1, which progressed during weeks 2-3, were reduced after bevacizumab administration. The average CNV incidence decreased from 50% at week 3 to 4.2% at week 4, and the average MIFI decreased from 4.62 ± 1.15 to 1.76 ± 0.81, both of which were statistically significant. On the other hand, the CNVs induced by treatments 2 and 3 did not show any significant changes over time. Our study demonstrated that repeated retinal photocoagulation in the monkey eye produces relatively consistent CNVs, which can be used to assess the efficacies of anti-angiogenic agents more efficiently.