Abstract
Purpose: :
To determine the relative cytotoxicities of prostaglandin analog topical medications in an in vitro assay of cultured human corneal epithelial cells, and their effects on corneal wound healing in an ex-vivo model.
Methods: :
Bimatoprost 0.03%, latanoprost 0.005%, travoprost 0.004%, BAK-free travoprost 0.004%, media (control), and methanol (toxic control) were tested in an in vitro cytotoxicity assay. For each agent, cultured microtiter plates of immortalized human corneal epithelial cells grown in confluence (n = 5 per group) were exposed for 25 minutes. Ethidium bromide (EB) staining was performed and fluorescence measured to quantify cellular degradation. Each of the 4 prostaglandin analog formulations, saline (control), and Triton (toxic control) were also topically applied to wounded ex-vivo whole globe porcine eyes. Globes with standardized 5 mm corneoepithelial wounds were incubated for 24 hours in media before topical administration of a test agent (n = 3 per group). Wounds were exposed to the agent for 10 minutes, rinsed with PBS and then with fresh media. The wound healing response and epithelial defects were evaluated using Richardson’s staining solution 48 hours later.
Results: :
In the cytotoxicity assay, mean EB fluorescence was 26.8 with bimatoprost, 27.5 with BAK-free travoprost, 32.7 with travoprost, 60.5 with latanoprost, 12.1 with media, and 51.6 with methanol. Using one-way ANOVA, travoprost, BAK-free travoprost, and bimatoprost all showed significantly less fluorescence than latanoprost (P < .05), indicating less cytotoxicity. For the wound-healing assay, corneoepithelial wounds that had been exposed to saline established a baseline mean healed percentage of 96.89%. The mean healed percentage was 6.06% for Triton, 97.32% for bimatoprost, 84.76% for travoprost, 99.80% for BAK-free travoprost, and 80.51% for latanoprost. The wound sizes for bimatoprost, travoprost, and BAK-free travoprost-treated eyes did not statistically differ from saline-treated eyes (P > .141). Latanoprost-treated wounds were significantly less healed than saline, bimatoprost, or BAK-free travoprost-treated wounds (P <.001).
Conclusions: :
The EB staining model demonstrated that bimatoprost, travoprost, and BAK-free travoprost produce less cellular toxicity than latanoprost or toxic controls. Similarly, bimatoprost and BAK-free travoprost did not impede wound closure, but the effect of other IOP-lowering medications on wound healing should be considered for glaucoma filtration or cataract surgery patients.
Keywords: cornea: epithelium