The effect of intravitreally injected povidone-iodine on Staphylococcus epidermidis in rabbit eyes

Purpose: Firstly, the aim of this study was to determine the maximally tolerated dose of intravitreally injected povidone-iodine (PVP-I). A second aim was to test the efficacy of PVP-I on rabbit eyes infected intravitreally with Staphylococcus epidermidis. Methods: Phase I of the study used 16 New Zealand albino rabbits, divided into 4 groups (n = 4 each). Animals were anesthetized and intravitreally injected with 0.1 mL of 50,100, 200, or 400 micrograms (mu g) of PVP-I in 1 eye, and with saline in the other. The animals were examined at days 1, 7, and 14, using indirect ophthalmoscopy and slit-lamp biomicroscopy; electroretinography (ERG) was performed before treatment and prior to euthanization. Histological preparations were examined to determine retinal damage. Phase II of the study divided 20 New Zealand albino rabbits into 4 groups (n = 5 each). Animals were anesthetized and injected with 0.1 mL of S. epidermidis containing 3030 colony forming units (CFU) in 1 eye and saline in the other. Seven (7) h later, animals were treated with 0.1 mL of 20, 50, and 100 mu g of PVP-I, or no treatment. Bacterial concentrations from extracted vitreous were determined 2 days following infection. Results were analyzed for statistical significance, using the Student t test and analysis of variance, and histologic preparations assessed the presence of endophthalmitis. Results: Phase I of the study observed no retinal damage at any of the concentrations studied, as noted by indirect ophthalmoscopy, slit-lamp biomicroscopy, ERG, and histologic exam. Phase II of the study showed no statistical difference in bacterial counts between treatment and control groups. All infected eyes went on to develop endophthalmitis, as observed by indirect ophthalmoscopy and histologic preparations. Conclusions: These results suggest that 400_g of PVP-I can be tolerated intravitreally in rabbit eyes with no noticeable damage over a 14-day period. Results further showed that 100 mu g of intravitreally injected PVP-I has no statistically significant effect on rabbit eyes injected intravitreally with 3030 CFU of S. epidermidis.