Abstract
Purpose :
Corneal collagen crosslinking via photoactivation of riboflavin (RF) is an effective treatment for keratoconus and has increasingly been investigated as a potential corneal refractive procedure. Typically, the epithelium is removed to achieve adequate RF concentration in the corneal stroma. Epithelial removal is less than ideal; as such various methods have been investigated for transepithelial RF delivery. This study examined the effect of benzalkonium chloride (BAC) on transepithelial RF delivery and stromal RF concentration as a function of time and RF concentration. Results were compared against standard and modified epithelium-off Dresden protocols.
Methods :
Experiment 1: To investigate the effect of BAC on stromal RF concentration, cadaver rabbit eyes were split into 4 groups (n≥4). Groups 1-4 received eye drops once every 2 minutes for 30 minutes. Group 1: Epithelium was left intact. Eye drops consisted of 0.1% RF, 0.01% BAK, and 1% methylcellulose. Group 2: Epithelium was left intact. Eye drops consisted of 0.5% RF, 0.01% BAK, and 1% methylcellulose. Group 3: Epithelium was manually removed. Eye drops consisted of 0.1% RF and 20% dextran (Dresden). Group 4: Epithelium was manually removed. Eye drops consisted of 0.5% RF and 20% dextran (modified Dresden). Experiment 2: To investigate the effect of eye drop administration duration on stromal RF concentration, a separate set of eyes were separated into 7 groups of varying treatment times, ranging from 10 to 60 minutes (n≥4). Eyes drops in this experiment consisted of 0.5% RF, 0.01% BAK, and 1% methylcellulose. Following riboflavin administration, stromal concentration (µg/mL) was measured in all groups via RF extraction/fluorometry.
Results :
Experiment 1: Stromal RF concentration for groups 1-4 were, 18.0±3.4, 83.3±45.8, 93.2±51.6, and 215.6±14.4 µg/g, respectively. Group 4 RF concentration was significantly higher than all other groups (p<0.01). Groups 2 and 3 were not significantly different. Experiment 2: RF concentration was found to plateau at 30 minutes.
Conclusions :
0.5% RF with 0.01% BAC can achieve stromal concentrations of RF equal to that of the standard Dresden protocol, suggesting the possibility of transepithelial crosslinking with UVA light. For femtosecond laser mediated collagen crosslinking (NLO-CXL), an RF concentration of 200 µg/g or above is necessary, indicating that BAC alone is insufficient.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.