Epithelium-on CXL: New frontier for topographic abnormalities

Take-Home Message: The rationale for epithelium-on crosslinking procedures is strong, and may ultimately make customized treatment of topographic abnormalities more attractive for therapeutic and refractive indications.

By Lynda Charters; Reviewed by William J. Dupps Jr., MD, PhD

Cleveland-Though epithelium-on corneal collagen crosslinking (CXL) is emerging as a possible alternative to epithelium-off CXL, a number of important questions remain unanswered.

While the major advantages include less pain and reduced risk of microbial and sterile keratitis, the various techniques have not been directly compared with the epithelium-off gold standard in clinical trials.

The safety rationale for performing epithelium-on CXL is strong, according to William Dupps Jr., MD, PhD.

“Most of the post-treatment symptoms and complications that we see with CXL are associated with epithelial debridement and the ongoing presence of an epithelial defect,” said Dr. Dupps, staff, Ophthalmology, Biomedical Engineering and Transplantation, Cole Eye Institute, Cleveland Clinic, Cleveland.

These complications include delayed epithelial healing, sterile keratitis, microbial keratitis, central stromal scarring, and potential loss of best-corrected visual acuity.

Penetration

The effectiveness of epithelium-on CXL is the primary concern. Riboflavin, ultraviolet light, and oxygen are the essential components of the effectiveness of epithelial-on CXL, but an intact epithelium inhibits their penetration, Dr. Dupps noted.

However, the epithelial-off CXL procedure-considered the most effective approach-demonstrated a failure rate of about 8% in one long-term study because of continued disease progression.

“We expect that the failure rate will be higher for transepithelial CXL approaches if they are less effective in stiffening the cornea,” he said. “Is this an acceptable compromise?”

The goals of treatment, the rate of disease progression, and the impact of complications will likely lead to unique answers for each patient, Dr. Dupps said.

Defining effectiveness

“Exciting” techniques are emerging to measure CXL’s efficacy, according to Dr. Dupps.

One-Brillouin spectroscopy-was used to demonstrate increases in regional corneal stiffness after an epithelial-off procedure, as reported by Scarcelli and colleagues (Invest Ophthalmol Vis Sci. 2013;54:1418-1425).

Another emerging technique is optical coherence tomography (OCT) elastography, which Dr. Dupps and colleagues are using in their laboratory, and which he explained can resolve spatial stiffness properties. In a study using OCT elastography, they compared various transepithelial and epithelium-off CXL techniques performed in rabbit eyes (J Refract Surg. 2013;29:332-341).

After 3 months of follow-up, investigators found the greatest corneal stiffening was found in a group of rabbits treated with a transepithelial approach using benzalkonium chloride and EDTA as permeability enhancers, Dr. Dupps noted.

The results were replicated in a follow-up experiment (Exp Eye Res. 2014;125:114-117).

“We think that this somewhat surprising result may be related to differences in the wound-healing response to CXL and the riboflavin vehicle,” he said. “While the results should not be directly extrapolated to human eyes, the studies did demonstrate the potential for certain transepithelial techniques to produce stiffening effects on par with those of the epithelium-off standard.”

What is the goal?

Is disease stabilization or topographic improvement the primary goal?

“Most would say that disease stabilization is the goal, but CXL also has the potential to improve the corneal topography,” Dr. Dupps said.

Dr. Dupps and co-workers are using computational modeling techniques in their laboratory to gain a better understanding of how the corneal shape and stiffness are related-and how this relationship can be leveraged for better outcomes.

He summarized their early work (Invest Ophthalmol Vis Sci. 2011;52:9174-9187) looking at the differences in the corneal shape response when the corneal properties were changed from weaker to stiffer in simulated myopic LASIK procedures.

“A weaker cornea’s center displaces forward and steepens slightly, leading to a myopic undercorrection,” Dr. Dupps said. “In contrast, a stiffer cornea shows some central flattening and a relative overcorrection of myopia.”

These very different shape responses lead to different refractive outcomes and are explained purely by a difference in the corneal stiffness.

“Our conclusion was that the corneal biomechanical properties can be an important driver of refractive outcomes,” Dr. Dupps said.

This relationship is even more important in keratoconus.

“If we envision corneal ectasia as a disease that is fundamentally a regional weakening of the cornea, we can generate topographic progression of the disease in a model, even without thinning the cornea,” he said. “The lower the corneal elastic modulus, the higher the corneal curvature.”

However, this relationship between corneal strength and curvature is nonlinear, Dr. Dupps said.

Investigators found that in the eye modeled as part of their 2011 publication (Invest Ophthalmol Vis Sci.), the maximum K value increased modestly in response to decreases in the corneal elastic modulus up to 30%, then increased precipitously.

“Furthermore, every patient has a different corneal geometry, and therefore, a different tipping point regarding disease progression,” he said.

Dr. Dupps and his team are working to translate patient-specific simulation into a clinical tool for predicting the risk of ectasia or its progression.

“The hope is that at-risk patients can be identified prior to procedures that would precipitate structural failure and steered instead toward tissue-sparing and/or corneal stiffening procedures,” he said. “For patients who are known to have keratoconus, disease progression simulations may be helpful in estimating risk of progression and determining the optimal timing of CXL procedures.”

Beyond stabilizing keratoconus

Another finding from the modeling studies has important implications for patients who have already lost vision due to topographic disease progression.

“Focal collagen crosslinking has the potential to produce greater reductions in cone steepness than the pan-corneal treatments most patients currently receive,” Dr. Dupps said.

In a modeling study of standard 9-mm treatment zones, Dr. Dupps observed the typical 1- to 2-D regional flattening effect. More focal treatments with smaller diameters decentered toward the cone much greater reductions in cone steepness and marked reductions in coma. Technologies are in the pipeline to allow spatial treatments and higher intensity treatments.

“This experiment provided a structural rationale for customizing the approach to CXL and may be important in the debate about transepithelial CXL,” he said. “The stiffness of the cone region relative to its surroundings is a major driver of response; enhancing the stiffening effect in the weakest area of the cornea relative to other areas will produce more ‘leverage’ for local flattening by also allowing favorable compensatory steepening away from the cone.”

Building on this, Dr. Dupps explained that existing evidence for a thinner epithelial covering in the region of the cone could favor outcomes with epithelium-on CXL in some cases.

“If thinner epithelium favors greater crosslinking effect where it most needed and less where it is not, the net curvature response could still be very favorable even if absolute stiffening effect is lower overall than epithelium-off CXL,” he said.

Transepithelial CXL techniques are attractive because of potential benefits related to less postoperative pain, lower risk of sterile and microbial keratitis, and faster visual recovery. Questions about relative effectiveness require more clinical investigation, and the results are likely to depend on technique differences and peculiarities of the epithelial thickness profile seen in keratoconus.

A hybrid solution combining advantages of epi-off and epi-on treatment could be a localized treatment with or without focal corneal debridement that offers less corneal exposure to ultraviolet light, a smaller epithelial defect, and potentially greater topographic improvement, according to Dr. Dupps.

William J. Dupps Jr., MD, PhD

E: bjdupps@sbcglobal.net

This article was adapted from Dr. Dupps’ presentation during Cornea Subspecialty Day at the 2014 meeting of the American Academy of Ophthalmology. Dr. Dupps does research for and is on the medical advisory board of Avedro. He is also founder of OptoQuest, a Cleveland Clinic company with a commercial interest in computational modeling of ocular surgery. Support for the work described above was provided by Avedro, the National Institutes of Health (R01 EY02338), the National Keratoconus Foundation/Discovery Eye Foundation, an Ohio Third Frontier Innovation Platform Award to the Cleveland Clinic Cole Eye Institute, and an Unrestricted Grant and Career Development Award from Research to Prevent Blindness.