Objective To evaluate the safety repeated intravitreal injection of bevacizumab (Avastin) with different dosage in rabbitsprime;eyes. Methods Fourteen chinchilla rabbits were randomly divided into 3 groups, including both eyes of 2 rabbits in the control group,the right eyes of the other 12 rabbits in the experimental group,and the left eyes of the 12 rabbits in the experimental control group. The eyes in the experimental group underwent intravitreal injection of bevacizumab with the dosage of 2.5 mg/0.1 ml and 5.0 mg/0.2 ml of bevacizumab; the eyes in the experimental control group underwent intravitreal injection of normal saline with the same dosage as in the experimental group. Injections were performed every two weeks and lasted six weeks. Clinical observation and retinal function tests were performed before and two days after every injection. The eyes were sacrificed 1 week and 4 weeks after last intravitreal injection respectively.Electron and optical microscope and TUNEL were performed.Results After intravitreal injection,no obvious anterior chamber flare, abnormal change of the ocular fundus, or vitreous opacity and hemorrhage was observed in all of the eyes.No change was found by indirect ophthalmoscope,Bultrasonic inspection, ultrasound biomicroscopy and optical coherence tomography. The number of anterior chamber flare before and after the injection with the dosage of 2.5 and 5.0 mg, the difference among the 3 groups didnprime;t differ much from each other (Pgt;0.05).Amplitude and pattern of ERG responses and flash VEP were similar between the control and experimental groups (Pgt;0.05). Some inflammatory cells were found in the some bevacizumabinjected eyes 1 week after injection, and vanished 3 weeks later. The histological configuration of the retina didnprime;t change in both experimental control and the control group. Electron microscopy showed that plasma cells were presented and vacuolelike change was observed in part of the photoreceptor cells in 5.0 mg experimental group 1 week after injections.Cellular apoptosis was observed in the photoreceptor cell layer. The number of apoptotic cells was more in 5.0 mg experimental group than that in the control and experimental control group 1 week after injections (Plt;0.01). Conclusion Multiintravitreal injection with 5.0 mg bevacizumab may have mild toxicity to the retina in the rabbits.
Objective To observe the inhibitory effects and characteristics of intravitreal injection with bevacizumab on laser induced choroidal neovascularization (CNV).Methods Twelve male brown norway（BN）rats were divided into the bevacizumab group and control group with six rats in each group. One eye of rats were received a series of 8 diode laser esions around optic disc to induce CNV,then the rats in bevacizumab group and control group underwent intravitreal injection with 2 mu;l bevacizumab and ringer's lactate.On days 7,14,and 21，the morphology and leakage of CNV were observed by fundus fluorescein angiography (FFA) and indocyanine green angiography (ICGA).On day 21 after photocoagulation,the photocoagulated eyes were enucleated and processed for histopathologic examination, including hematoxylin and eosin (Hamp;E) staining and immunohistochemistry staining for vascular endothelial growth factor（VEGF）.Results On day 7 after photocoagulation,ICGA showed that CNV developed in the bevacizumab group and the control group. FFA showed that leakage intensity in the bevacizumab group was significantly lower than that in the control group，but the bevacizumab group gradually increased over time. The mean thickness of CNV significantly decreased in the bevacizumab group.The CNV in the bevacizumab group were negative for VEGF according to the result of immmuohistochemistry staining.Conclusions Early intravitreal injection with 2 mu;l bevacizumab can reduce the thickness of CNV and inhibit the leakage of CNV. However, bevacizumab could neither block the formation of CNV, nor suppress the permeability permanently. Combined other therapies with bevacizumab may be more potential to treat CNV effectively.