An Experimental Study On Choroidal Neovascularization Inhibited By Pigment Epithelium-derived Factor

Objective: To study the effect of adenoviral vectored pigmentepithelium-derived factor (AdPEDF) gene transfer on establishedneovascularization in a model of choroidal neovascularization (CNV), observethe trend of the regression and change of CNV, and ascertain the better methodbetween intravitreal injection and periocular injection.Methods: Sixty-eight female BN rats with weight of 160 to 180g were usedfor this study. Four rats (8 eyes) were randomly selected as normal controlgroup and another 64 rats (128 eyes) were as experimental group. Theexperimental group (64 rats) was randomly divided into 4 groups with 16 rats ineach group. Intravitreal injection with AdPEDF 1μl in group A; intravitrealinjection with control vector (AdNull) 1μl in group B; periocular injection withAdPEDF 1μl in group C; periocular injection with AdNull 1μl in group D. Thenthe each group was randomly divided into 4 sub-groups with 4 rats in eachsub-group according to the different follow-up time. All the rats wereanesthetized and the pupils were dilated. Eight laser photocoagulation spots bykrypton laser, with the spot size of 50gin, duration of 0.05 second, intensity of350～380mW and wave length of 647.1nm, were applied to each eye at the areaof 2 PD away from and around the optic disc. The rats received theexaminations of fundus fluorescein angiography and were sacrified with nooperation in normal control group just for comparing with the experimentalgroups after 14 days. At the same time, the experimental groups received theexaminations of FFA. Then they received an intravitreal or periocular injection of AdPEDF or AdNull. Finally, the rats were sacrified on days 3, 7, 14, 28 afterinjection. We observed the histopathology and TUNEL staining after injection ata given time.Results: After injection in the choriodal neovascularization model, thedevelopment of CNV in the group A and group C was significantly inhibited.1. The leakage appeared in eyes injected with AdPEDF was lighter than in eyes injected with AdNull on days 7, 14, 28 after injection (P＜0.01).2. Histologically, there was no difference between the experimental groups andthe control group under microscope on days 3 after injection. CNVdecreased and vascular channels attenuated in group A and group C on days7, 14, 28 after injection; meanwhile, CNV kept fibrovascular proliferation,and large collections of vascular channels were seen in group B and groupD.3. The FVP membrane was observed beneath the retina, and the meanthickness of the recovered lesions (n=25) was measured. The thickness ofCNV in group A and group C was significantly less than in normal controlgroup on days 7, 14, 28 after injection(P＜0.01).The thickness of CNV ineyes injected with AdPEDF was significantly less than in eyes injected withAdNull on days 3, 7, 14, 28 after injection (P＜0.01). On days 3, 14, 28 afterinjection, there was a statistical difference in the thickness of CNV betweenin eyes with intravitreal injection and in eyes with periocular injection(P＜0.05,P＜0.01,P＜0.05).4. There was TUNEL staining with choroidal neovascular lesions in theexperimental group. On days 7, 14, 28 after injection, eyes given aintravitreal or periocular injection of AdPEDF had TUNEL-positive cells inthe areas of choroidal neovascularization, but none in the areas of choroidal neovascularization with eyes given a intravitreal or periocular injection ofAdNull. Furthermore, TUNEL-positive cells were not observed in the areasof choroidal vascularization in all rats.5. Complication: Except complicated cataract in few rats after intravitrealinjection, no other complications were observed.Conclusion: These data indicate that increased expression of PEDF causesregression of CNV by promoting apoptosis of neovascular endothelial cells. Thethickness of CNV reduces from day 7, reaches the peak on day 14, and keepsstabilization to day 28 after injection. The effect of eyes with periocularinjection is quick, and the function of eyes with intravitreal injection is strong.The method of periocular injection is easy and repeated. The method ofintravitreal injection is complex and induces much complication. At last, 1μlAdPEDF of 1×10¹⁰ pfu/ml can inhibit CNV, and no toxic reaction to retina.