OBJECTIVE: To characterize the mechanism by which air bubbles damage the corneal endothelium during phacoemulsification. MATERIALS AND METHODS: A series of experiments was conducted to expose the corneal endothelium of New Zealand white rabbit and human eyes that were obtained from an eye bank to air under different conditions. Phacoemulsification at different power settings and irrigation with and without the introduction of air into the anterior chamber were performed. Corneal endothelial perfusion experiments were conducted with air bubbles that were introduced into the perfusion chamber for 2 seconds to 1 hour. Air was also injected into the anterior chambers of anesthetized rabbits for 2 minutes to 3 hours. Corneas were stained with nitrobenzo-xadiazole-phallacidin and examined with fluorescence microscopy. Selected corneas were also examined with scanning and transmission electron microscopy. RESULTS: Intracameral air bubbles during phacoemulsification, irrigation, and perfusion studies resulted in a severe injury to the corneal endothelium in as little as 20 seconds. Intracameral air bubbles in a living rabbit resulted in a slower injury that was morphologically different from the more rapid injury. CONCLUSIONS: Air bubbles in intraocular fluids with a high surface tension can cause a ring-shaped pattern of damage to the corneal endothelium. The mechanism that caused this pattern of damage appears to be a surface tension phenomenon.
OBJECTIVE: To characterize the mechanism by which air bubbles damage the corneal endothelium during phacoemulsification. MATERIALS AND METHODS: A series of experiments was conducted to expose the corneal endothelium of New Zealand white rabbit and human eyes that were obtained from an eye bank to air under different conditions. Phacoemulsification at different power settings and irrigation with and without the introduction of air into the anterior chamber were performed. Corneal endothelial perfusion experiments were conducted with air bubbles that were introduced into the perfusion chamber for 2 seconds to 1 hour. Air was also injected into the anterior chambers of anesthetized rabbits for 2 minutes to 3 hours. Corneas were stained with nitrobenzo-xadiazole-phallacidin and examined with fluorescence microscopy. Selected corneas were also examined with scanning and transmission electron microscopy. RESULTS: Intracameral air bubbles during phacoemulsification, irrigation, and perfusion studies resulted in a severe injury to the corneal endothelium in as little as 20 seconds. Intracameral air bubbles in a living rabbit resulted in a slower injury that was morphologically different from the more rapid injury. CONCLUSIONS: Air bubbles in intraocular fluids with a high surface tension can cause a ring-shaped pattern of damage to the corneal endothelium. The mechanism that caused this pattern of damage appears to be a surface tension phenomenon.