Collagen crosslinking: Epithelium on versus off?

Chicago-The debate over the need to remove the corneal epithelium when performing corneal collagen crosslinking (CXL) will continue until long-term data from peer-reviewed studies become available to answer the question of whether a procedure done with the epithelium on results in sufficient crosslinking for long-term arrest of ectasia progression, said Farhad Hafezi, MD, PhD, during refractive surgery subspecialty day at the annual meeting of the American Academy of Ophthalmology (AAO).

In addition, results from experimental models investigating such issues as riboflavin penetration and concentration in the corneal stroma and the effects of the CXL procedure favor epithelial removal, he said.

Parameters show potential

"Further study is needed, and the only conclusion that can be reached for now is that the CXL treatment parameters show great potential for optimization," Dr. Hafezi said. "Certainly, the way the procedure will be performed 5 years from now will be different [from] the technique used today."

The conventional method for CXL involves epithelial removal in order to increase riboflavin penetration into the corneal stroma, which is poor when the epithelium is intact because riboflavin is a high-molecular-weight macromolecule. Riboflavin is needed in the stroma for UVA-induced generation of oxygen free radicals that induce the crosslinking reaction.

"The less riboflavin present in the stroma, the less oxygen radicals are generated and the less crosslinking that is achieved," Dr. Hafezi said. "Furthermore, the epithelium has a tendency to block the short-wavelength UVA light used for CXL, and while this barrier effect is not that great, it still reduces the amount of energy that reaches the corneal stroma."

As reviewed by Dr. Hafezi, available data from clinical trials and experimental models clearly show a benefit of the epithelium-off method for enhancing riboflavin bioavailability at its target site and increasing the depth and magnitude of the crosslinking effect.

For example, in an ex vivo study of human corneas, HPLC analysis showed a 40-fold lower concentration of riboflavin in the stroma when the riboflavin solution was applied with the intact epithelium on versus off.

In addition, findings from confocal microscopy and slit-lamp examination show that with the epithelium-off method, crosslinking occurs to a depth of 270 to 330 µm.

In contrast, findings from an animal model using confocal microscopy indicated the depth of crosslinking was only 50 to 80 µm when the procedure was performed with the intact epithelium on, and stress-strain measurements from a rabbit eye study showed the biomechanical effect of the epithelium-on method was only 20% of that achieved with conventional CXL.

"These data indicate that crosslinking occurs with the epithelium-on method, but the effect is shallower and less than when the epithelium is removed," Dr. Hafezi explained.

According to his recent research, the concentrations of benzalkonium chloride, dextran, and sodium chloride in the riboflavin solution play important roles in the passage of the vitamin through the epithelium. In particular, dextran blocks the paracellular transport of substances through the epithelium.

"These findings indicate that the currently commercially available riboflavin solutions being used for transepithelial CXL may not be optimally formulated," Dr. Hafezi explained.