Corneal Collagen Cross-linking with Riboflavin/Lumiflavin and Blue Light (450nm) to Reduce Posterior Lamellar Dislocation in Endothelium Keratoplasty

Abstract

Purpose: : To determine the dislocation rate of the posterior lamellar graft in an ex vivo human endothelial keratoplasty model with the use of riboflavin and lumiflavin film activated by blue light to induce corneal collagen cross-linking.

Methods: : Thirty corneoscleral rims unsuitable for corneal transplantation were acquired from Tissue Banks International, Baltimore, MD. An artificial anterior chamber was used to support the 30 corneoscleral rims. Central corneal thickness was measured using an ultrasound pachymeter (DGH, Exton, PA). After this, the cornea was flipped endothelial side up in the artificial anterior chamber in order to create a free lamella using a mechanical microkeratome. A 160-µm head thickness was used in all corneas. The donor buttons (n=6 per group) were assigned into the following treatment groups: 1 (0.1% riboflavin and blue light), 2a (0.5% lumiflavin film and blue light), 2b (0.5% lumiflavin film and no blue light), 3a (only blue light), and 3b (no blue light). Immediately after the free lamellar replacement the blue light (450 nm wavelength, power density of 17.5 mW/cm²) was applied for 20 minutes from epithelial side, in those groups assigned to it. Then, corneas were submitted to a two-step stability test which consisted in applying an external pressure and sudden decompression to verify the graft dislocation rate.

Results: : Group 1 was able to avoid disc dislocation in 100% of cases, and was superior to groups 2b (P=0.002), 3a (P=0.015), and 3b (P=0.015). Group 2a, although it had disc dislocation of 2 in 6 cases, it was not statistically different from group 1 (P=0.394). Group 2a had the second lowest dislocation rate, however was not statistically superior to groups 2b (P=0.065), 3a (P=0.092), and 3b (P=0.092). We could not demonstrate any significant correlation between flap thickness, pre and post-blue light corneal thickness, number of stability tests, and dislocation rate.

Conclusions: : The use of riboflavin and blue light was effective to attach the posterior graft to the stromal bed in this ex vivo endothelial keratoplasty human model. Further testing is required to optimize this procedure before use in the clinical setting.

Keywords: cornea: basic science • cornea: stroma and keratocytes • radiation therapy