Abstract
Purpose:
To establish an ex vivo model for post-intravitreal injection endophthalmitis that correlates with in vivo models and adequately simulates all variables.
Methods:
Freshly enucleated (<6 hours) pig eyes were cleaned of extraocular tissue and submerged in 10% povidone iodine for 10 minutes. Suspensions of Staphylococcus epidermidis in phosphate buffered saline (PBS) were prepared in varying concentrations from 9x10 8 to 3x10-1 CFU/ml. Using a 30-gauge needle, 0.1 ml of each suspension was injected through the pars plana to the mid-vitreous cavity. The inoculated eyes were incubated for 24 hours at 37°C in either 2% oxygen to mimic in vivo conditions or normoxia. The external surface of the eyes was then re-sterilized and cultures were taken through a 4mm incision. Vitreous samples were incubated on blood agar in normoxic conditions and checked daily for 3 days. Bacterial growth was quantified as CFU/ml of vitreous. The effect of extraocular bacterial population density was sought by submerging globes in bacterial suspensions of varying concentrations from 1.2x108 to 1.2x10-1 CFU/ml through which the injections were performed. To assess any intrinsic antibacterial effects of pig vitreous, it was incubated on blood agar plates that had been seeded with S. epidermidis.
Results:
Viable bacteria were consistently recovered from vitreous with inoculum loads ≥8x104 bacteria. A maximum yield of 3x106 CFU/ml was reached with an inoculum of 9x106 bacteria. Increasing the inoculum beyond this level limited the its viability. The relationship between intravitreal inoculum and bacterial growth fit a sigmoidal curve with the 50% growth point at an inoculum of 2x106 bacteria. In normoxic conditions, the yield was markedly increased. Cultures were negative in all eyes injected through bacterial suspensions with PBS. Isolated vitreous did not inhibit bacterial growth.
Conclusions:
Bacterial endophthalmitis can be accurately modeled in enucleated pig eyes with resulting bacterial growth consistent with in vivo models showing spontaneous sterilization can occur rapidly with small bacterial loads and there is a maximum concentration of recovered bacteria that cannot be exceeded despite increasing inoculum size. Injection of sterile PBS through contaminated media did not result in consistent recovery of viable bacteria, indicating that insufficient bacteria are introduced by the needle to cause a sustained endophthalmitis.
Keywords: 513 endophthalmitis •
561 injection