Ex Vivo Study of Transepithelial Corneal Cross-linking

PURPOSE: To perform in vitro assessment of different techniques of transepithelial corneal cross-linking (CXL) and to compare the results to deepithelialized CXL. METHODS: Transepithelial CXL was performed after pretreatment with or without penetration enhancers (gum cellulose, 0.44% sodium chloride, and 0.01% benzalkonium chloride) for 15 or 60 minutes. Deepithelialized corneas underwent CXL after pretreatment with riboflavin for 15 minutes, according to the Dresden protocol. All corneas were incubated in 0.3% collagenase A solution and the time to total dissolution was measured. Corneas were also imaged with confocal microscopy to evaluate the corneal epithelium, subbasal nerve plexus, and depth of stromal keratocyte nuclei as a means of measuring the depth of collagen CXL. RESULTS: Deepithelialized CXL corneas with 15 minutes of pretreatment dissolved after 15.4 +/- 3.1 hours, significantly longer (P = .001) than deepithelialized untreated corneas (8.5 +/- 0.6 hours). Transepithelial CXL corneas with 15 minutes of pretreatment with or without penetration enhancers dissolved after 8.3 +/- 2.1 and 7.4 +/- 1.6 hours, respectively. A longer pretreatment of 60 minutes with penetration enhancers resulted in greater resistance to degradation of the transepithelial CXL corneas (14.6 +/- 2.2 hours), which was similar to deepithelialized CXL corneas. The results of the biological assay correlated well with the imaging results obtained by confocal microscopy. CONCLUSIONS: Corneas treated by transepithelial CXL with an extended pretreatment time of 60 minutes and penetration enhancers exhibited similar characteristics as corneas treated by the deepithelialized CXL approach. By confocal imaging, the transepithelial approach with extended pretreatment time demonstrated evidence of epithelial damage, which may have improved the treatment effect of this group.