Comparison between standard and transepithelial corneal crosslinking using a theranostic UV-A device

Purpose To assess corneal concentration of riboflavin in two different corneal crosslinking protocols performed by a novel image-guided therapeutic (or "theranostic") UV-A device. Methods Ten human eye bank donor tissues were used in this work. The tissues underwent corneal cross-linking according to the conventional treatment protocol (n = 5; 30 min of stromal soaking followed by 30 min of 3 mW/cm(2) UV-A irradiance) and the iontophoresis-assisted transepithelial protocol (n = 5; soaking for 5 min at 1 mA/min and 9 min of 10 mW/cm(2) UV-A irradiance) using a theranostic UV-A device (Vision Engineering Italy srl, Italy). The device provided real time assessment of riboflavin concentration by hyperspectral image analysis of the cornea. A 0.1% riboflavin hypotonic solution (Ricrolin+, Sooft Italia Spa, Italy) was used in all cases. Results Manual application of hypotonic riboflavin for 30 min into the stroma achieved greater corneal riboflavin concentration (425 +/- 77 mu g/cm(3)) than transepithelial delivery of riboflavin by corneal iontophoresis (195 +/- 35 mu g/cm(3); P = 0.001). In both UV-A irradiation protocols, corneal riboflavin concentration decreased exponentially with a constant energy rate of 2.3 +/- 0.5 J/cm(2) and 1.8 +/- 0.3 J/cm(2) respectively. At the end of treatment, the average corneal concentration of riboflavin decreased by >= 85%, with values of 54 +/- 29 mu g/cm(3) and 31 +/- 9 mu g/cm(3) (P = 0.11), respectively. Conclusion Manual application of riboflavin onto the stroma achieved almost 50% greater concentration of riboflavin than transepithelial delivery by corneal iontophoresis. The theranostic UV-A device provided a novel approach to estimate corneal concentration of riboflavin non-invasively during treatment.