Abstract: Objective  To investigate side actions of some anti-inflammatory eye drops applied frequently on corneal epithelium and to observe the effects on healing of corneal epithelium using these eye drops routinely. Design Experimental study. Participants 80 New Zealand rabbits. Method Rabbits were divided randomly into two groups (40 rabbits per group), the rabbits of one group had the healthy cornea and were given the eye drops every one hour in their right eyes for 5 days; and the rabbits of the other group had de-epithelium cornea (ethyl alcohol to make a 6 mm epithelium defect in the center) and were given the eye drops four times per day in their right eyes for 7 days. Each group was divided randomly into five subgroups in which the right eye were given topical sodium choloride, pranoprofen, diclofenac sodium, bromfenac sodium and tobramycin-dexamethasone, respectively. The signs were observed with slit lamp at an interval time (before epithelium injury, 12, 24, 48, 72, 96 hours and 5, 6, 7 days after epithelium injury). The histopathologic examination of rabbit corneas was performed for the epithelium thickness at the end of experiment. Main Outcome Measures conjunctival congestion, secretions in the conjunctival sac, the injury and healing time of the corneal epithelium. Results Compared with the sodium choloride subgroup, all the four subgroups with frequent anti-inflammatory eye drops showed none obvious local excitation, but the corneal epithelium defects were existed partly, which was more serious in the diclofenac sodium subgroup (9 points), which was significantly higher than the sodium choloride subgroup (0 point) and the bromfenac sodium subgroup (1 point) (P<0.05), while at the 5th day the epithelial cell layers of cornea in tobramycin-dexamethasone subgroup (3.67±0.52 layers) are less than that of the control group (4.17±0.41 layers)(P<0.05). As for the de-epithelium group, when compared with the sodium choloride subgroup(66.0±11.1 hours), the average corneal healing time in diclofenac sodium, tobramycin+dexamethasone, pranoprofen, and bromfenac sodium subgroups were  (75.0±27.0) hours, (75.0±8.5) hours, (69.0±15.4) hours and (66.0±11.1) hours, respectively (P>0.05), but the epithelial cell layers of cornea in tobramycin-dexamethasone subgroup (2.00±0.00 layers) and diclofenac sodium subgroup (2.50±0.55 layers) are less than that of the control group (5.00±0.00 layers) (P<0.05). Conclusion Frequently using anti-inflammatory eye drops may result in rabbit’s corneal epithelium damage to some extent. Routine topical diclofenac sodium and tobramycin-dexamethasone may interfere rabbit corneal epithelium healing.

Key words:  anti-inflammatory drop, cornea, side actions, epithelial healing, rabbit