The process of closure of experimental retinal holes in rabbit eyes

Background: To investigate the healing process of retinal holes, including the identification of the cell types which play an important role in the pr ocess, we created experimental retinal holes with minimal damage to retinal pigment epithelium (RPE) in rabbit eyes. Methods: Pars plana vitrectomy wa s performed in the rabbit eye. A dome-shaped retinal detachment (bleb; diam eter 1.5 mm) was made by injecting balanced salt solution into the subretin al space, followed by making a retinal hole (diameter 0.5 mm) in the center of the bleb with a silicone-tipped extrusion needle. In one group of rabbi ts, fluid-air exchange was performed and sulfur hexafluoride gas was inject ed into the vitreous cavity postoperatively. In another group, gas tamponad e was not performed. The operated eyes were examined ophthalmoscopically an d enucleated at 1, 4, 7, 14, 30, and 90 days after surgery. Tissues were pr epared in 5-mu m sections for hematoxylin-eosin staining and immunohistoche mistry with antibodies to cytokeratin 18 and glial fibrillary acidic protei n (GFAP) and examined by light microscopy. Results: In the gas-injected eye s, the retinal holes were ophthalmoscopically closed by 7 days after the su rgery. Microscopic examination revealed that the sensory retina around the retinal hole was reattached, and the area of retinal defect was covered wit h cells which were positive for cytokeratin 18 and GFAP by 7 days after the surgery. In the eyes without gas tamponade, the retinal holes did not clos e during the observation period. Conclusions: These findings suggest that e arly attachment between the sensory retina and RPE could be essential for c losure of a retinal hole, where glial and RPE cells might play an important role. This model seems to be useful to investigate the process of closure of retinal holes.