Corneal distribution of riboflavin prior to collagen cross-linking.

PURPOSE: To evaluate the distribution of riboflavin in the corneal stroma, under varying concentrations and application time.
MATERIALS AND METHODS: In 54 porcine eyes, the central corneal epithelium was removed, and 0.035, 0.1, or 0.2% riboflavin-5-phosphate (in 20% Dextran T-500) was applied for 10, 20, or 30 min (3 x 6 corneas in each of the 3 groups). Trephined corneal buttons were examined using confocal fluorescence microscopy. Stromal riboflavin distribution and concentration was determined by measuring riboflavin fluorescence in optical sections at 10 microm intervals through the entire cornea. The procedure was repeated in 7 human corneal donor grafts using 0.1% riboflavin-5-phosphate for 20 or 30 min.
RESULTS: In porcine corneas, fluorescence intensity peaked within the first 50 microm followed by a steep decline to baseline. Increasing the riboflavin concentration from 0.1 to 0.2% did not increase stromal depth propagation, although a higher concentration in the anterior 200 microm was observed. Reducing the riboflavin application time from 30 to 20 min had no impact on corneal depth propagation or total riboflavin uptake. However, a 10-min further reduction of the application time caused a significantly reduced riboflavin uptake. In all human corneas, fluorescence peaked within the anterior 50 microm, followed by a steep decline to baseline over the next 200 microm; similar to the observations in porcine corneas. The human corneas imbibed more riboflavin compared to the porcine corneas.
CONCLUSIONS: In human and porcine corneas, riboflavin does not appear to fully load the corneal stroma using the current clinical procedure. Instead, the uptake appears to be limited to the anterior approximately 200 microm. Changes in application time and riboflavin concentration have only little influence on stromal depth diffusion.