Impact of topical corticosteroid pretreatment on susceptibility of the injured murine cornea to Pseudomonas aeruginosa colonization and infection.

Research with animal models of Pseudomonas aeruginosa keratitis has shown that use of a topical corticosteroid alone against an established infection can significantly increase the number of colonizing bacteria or worsen clinical disease. Moreover, retrospective analysis has suggested that corticosteroid use in humans is associated with an increased risk of keratitis in eyes with pre-existing disease. Thus, while corticosteroids are often used to reduce ocular inflammation in the absence of infection, the risk of opportunistic infection remains a concern. However, the effect of corticosteroids on the intrinsic barrier function of uninfected corneas is unknown. Here, we tested if short-term topical corticosteroid treatment of an uninfected murine cornea would increase susceptibility to P. aeruginosa colonization or infection after epithelial injury. Topical prednisolone acetate (1%) was administered to one eye of C57BL/6 mice three times a day for 3 days; control eyes were treated with sterile PBS. Prior to inoculation with a cytotoxic P. aeruginosa corneal isolate strain 6206, corneas were subject to superficial-injury by tissue paper blotting, or scratch-injured followed by 12 h of healing. Previously we have shown that blotting renders mouse corneas susceptible to P. aeruginosa adhesion, but not infection, while 12 h healing reduces susceptibility to infection after scratching. Corneas were evaluated at 48 h for bacterial colonization and microbial keratitis (MK). To monitor impact on wound healing, corneal integrity was examined by fluorescein staining immediately after scarification and after 12 h healing. For both the tissue paper blotting and scratch-injury models, there was no significant difference in P. aeruginosa colonization at 48 h between corticosteroid-pretreated eyes and controls. With the blotting model, one case of MK was observed in a control (PBS-pretreated) cornea; none in corticosteroid-pretreated corneas. With the 12 h healing model, MK occurred in 6 of 17 corticosteroid-pretreated eyes versus 2 of 17 controls, a difference not statistically significant. Corticosteroid-pretreated eyes showed greater fluorescein staining 12 h after scarification injury, but this did not coincide with increased colonization or MK. Together, these data show that short-term topical corticosteroid therapy on an uninfected murine cornea does not necessarily enhance its susceptibility to P. aeruginosa colonization or infection after injury, even when it induces fluorescein staining.

A comparison of invasive and cytotoxic Pseudomonas aeruginosa strain-induced corneal disease responses to therapeutics.

PURPOSE: During corneal infection, cytotoxic Pseudomonas aeruginosa strains remain mostly extracellular, while invasive strains can enter corneal cells and replicate within them. We tested the hypothesis that ofloxacin, which easily penetrates host cell membranes, would be more effective than the less cell-permeable antibiotic tobramycin, for treatment of corneal infection by an invasive P. aeruginosa strain. METHODS: A murine model of P. aeruginosa keratitis was used to compare the response to ofloxacin, tobramycin, prednisolone acetate, and non-preserved saline treatment, as well as combination antibiotic-corticosteroid therapy for infection caused by a cytotoxic strain (6206) and an invasive strain (PAO1). Treatment involved hourly eye drop administration for 12 hours. RESULTS: As expected, tobramycin was less effective at eradicating viable bacteria from corneas infected with the invasive strain. Despite rapid sterilization of corneas in other antibiotic treated groups, disease progression occurred during the 12 hour treatment period. Both antibiotics hastened disease resolution over the next 7 days for infections caused by either strain. Corticosteroid use during the 12 hour treatment period was of little added benefit. CONCLUSIONS: Differences between invasive and cytotoxic strain infections in their early response to the different therapeutic regimens did not translate to notable differences after 7 days, but the effects of antibiotics in halting disease progression were delayed for both strain types. These results suggest that successful management might be improved by addressing factors contributing to disease progression during sterilization of the cornea by antibiotics.