Read Refractive Surgery Outlook
The mystery of why corneal CXL works
By Linda Roach
Thanks to positive clinical outcomes and apparently minimal risks, corneal cross-linking (CXL) has transformed progressive keratoconus into a treatable condition for thousands of patients around the world. Even though ophthalmologists worldwide have embraced e CXL, both clinicians and ophthalmic researchers still don’t really know why it works.
Short summaries about CXL say that it strengthens the cornea by adding chemical cross-links to the corneal collagen. Mental images of a biochemical skeleton of cross-links between or within collagen fibers are intuitive and irresistible.
But unfortunately, such simple explanations are at best incomplete, researchers say.
“There’s no doubt that this technique works clinically. But ‘cross-linking’ is a buzzword. People don’t really know what it means,” said Henry F. Edelhauser, PhD, a corneal physiologist and professor of ophthalmology at Emory University in Atlanta.
Yes, there are definitely chemical cross-links formed when the cornea is saturated with riboflavin and irradiated with UVA. However, their locations and the specifics of how they add tensile strength to the cornea are still under debate.
And, yes, secondary metrics show that cross-linked corneas are stronger and less elastic, exhibiting central flattening that can remain stable for at least three years and longer in some case studies.
Objective Comparison Impossible
But there are no objective ways to compare protocols, correlate treatment and its effects or gauge how much biomechanical change is “enough.”
Consequently, Dr. Edelhauser and other leading researchers advise caution in interpreting each new page of the unfolding story of corneal cross-linking.
Stephen D. Klyce, PhD, an adjunct professor of ophthalmology at Mount Sinai School of Medicine in New York who is an expert on corneal physiology, said some hypotheses about molecular mechanisms seem to persist in spite of a lack of evidence to support them.
“A lot of people like to think that there are interfibrillar bridges between the collagen fibers that keep them from swelling, but the electron microscopic data are not convincing,” Dr. Klyce said.
Role for Proteoglycans?
Another theory is that cross-linking causes proteoglycans in the extracellular matrix (ECM) to form bridges with collagen fibers, Dr. Klyce continued.
“That’s a hypothesis that I don’t particularly believe. It’s been basically disproven for a long time,” he said.
Michael Mrochen, PhD, senior scientist and the director of the Institute of Refractive and Ophthalmic Surgery in Zürich, Switzerland, adds that it makes no sense to talk about a patient’s CXL-treated cornea being 50 or 100 percent cross-linked because no objective way to measure this exists.
“Today, there is no clinical method out there that has been shown to be useful to determine the outcome of cross-linking by biomechanical measurement,” Dr. Mrochen said.
With these limitations in mind, the three researchers talked with Refractive Surgery Outlook about the mechanisms that they believe or suspect to be likely explanations of why CXL works.
Edelhauser: Three Biochemical Pathways
CXL was well-established as a clinical tool by the time that Dr. Edelhauser and colleagues reported on a series of painstaking experiments that outlined the bare bones of the biochemical pathways that strengthen the cornea with molecular cross-links.1
“First, this is one of the first papers really showing that cross-linking causes an oxidative change in the cornea,” Dr. Edelhauser said.
They confirmed that riboflavin and UVA must be present simultaneously to produce significant cross-linking and that singlet oxygen (1O2) molecules are essential reactants and intermediates in the process. Oxidation correlated directly with the strength of the treated corneal tissue.
“But we don’t necessarily know where the cross-links form – just with collagen alone, or proteoglycans or a combination of both? I think it’s probably a combination of both, but we don’t really know,” he said.
Three Paths to Covalent Bonding
The group’s tests on rabbit and shark corneas suggested three likely biochemical pathways that produce cross-linking, all mediated by singlet oxygen. Dr. Edelhauser and colleagues proposed that riboflavin plus UVA treatment:
- Causes production of imidazolone, an intermediate molecule that quickly attaches itself to other molecules. Commonly, histidine or another amino acid provides the pair of electrons to form the new covalent bond.
- Triggers endogenous populations of carbonyl groups in the ECM (allysine, hydroxyallysine) to form cross-links there.
- Degrades the riboflavin itself, releasing 2,3-butanedione, which also can react strongly with the endogenous carbonyls.
“The trouble with this type of study is that animal tissue is not going to give you all the answers,” Dr. Edelhauser said.
For instance, OCT images show a demarcation line about 300 microns deep in the stroma after CXL that has been interpreted by Dr. Mrochen and others as the boundary between living and dead keratocytes.
Dr. Edelhauser disagrees. “You can look at the light scatter there all day, but until you can use electron microscopy to examine the ultrastructure in cross-linked human corneas, we’re not going to know what’s really going on,” he said.
For LASIK, it took more than a decade to collect enough cadaver corneas for such studies, he said.
Klyce: Riboflavin and Proteoglycans
“If you could understand the molecular mechanisms behind cross-linking, you might be able to find more robust methods to do the procedure,” Dr. Klyce said.
His favorite theory does not directly involve collagen fibrils. Instead, he postulates that a stronger cornea comes from cross-links between riboflavin and long proteoglycan molecules found in the extracellular matrix. Covalent bonding would thin the extracellular spaces and enable collagen fibrils to pack together more closely.
“Proteoglycans normally have fixed negative charges along the chain, which makes them repel one another in the ECM,” Dr. Klyce explained. “Riboflavin might interact with the fixed negative charges to covalently bind to them and reduce the density of negative charges along the molecules.”
Less Space, Thinner Cornea
It has been shown in other experimental contexts that, with their negative charges bound, proteoglycan molecules move closer to each other and “practically precipitate,” Dr. Klyce said.
In the cornea, this would mean less space between collagen fibrils and a stronger cornea. The narrower space also would explain the common, puzzling observation that the cornea thins slightly in the early post-CXL period, he said.
Evidence for or against Dr. Klyce’s theory will require basic research in animal models.
“To get at the molecular basis of what’s going on, you have to do animal work – and something more than that same old slide of how riboflavin stiffens the rabbit cornea,” Dr. Klyce said. “You need better data than that.”
Mrochen: Where’s the Riboflavin?
In Zurich, the chairman of the institute that Dr. Mrochen directs is Theo Seiler, MD, leader of the team that developed and validated the standard (Dresden) protocol for corneal cross-linking.
“All factors we have looked at all indicate that there is cross-linking happening with this protocol,” Dr. Mrochen said.
It typically gives patients a better corneal shape and improved BCVA, with about 1 percent of patients losing more than two lines of BCVA, according to Dr. Mrochen’s informal review of studies published in 2009 and 2010, which included 1,040 subjects. Treatment failure, defined as keratoconus progression, was slightly less than 6 percent, he said, referring to unpublished data.
Doubts about Epi-On Approach
Recently, he has devoted more attention to doing mechanistic analyses of why epithelium-on CXL cannot work as well as epi-off procedures.
“The primary driver in this discussion is the concentration of riboflavin in the cornea. How much riboflavin can get into the cornea through the epithelium?” he asked.
A study published in 2009 reported that cross-linking effects in rabbit corneas were substantially lower when the epithelium was left in place. The study attributed this to poorer riboflavin penetration.2
However, increasing the riboflavin solution’s concentration won’t solve that problem, said Dr. Mrochen, who has submitted a paper to a journal on this subject.
“With an 0.1% riboflavin solution, we get a good cross-linking effect down to about 300 microns,” he said. “When you increase the concentration, you get more riboflavin into the stroma, but more light is absorbed at a shallower depth.”
1. McCall AS, Kraft S, Edelhauser HF, et al. Mechanisms of corneal tissue cross-linking in response to treatment with topical riboflavin and long-wavelength ultraviolet radiation (UVA). Invest Ophthalmol Vis Sci. 2010;51(1):129-138. PubMed abstract
2. Wollensak G, Iomdina E. Biomechanical and histological changes after corneal crosslinking with and without epithelial debridement. J Cataract Refract Surg. 2009;35(3):540-546. PubMed abstract
Dr. Edelhauser has been a consultant to and received research funds from Alcon Laboratories.
Dr. Klyce is a consultant with Clinical Research Consultants, which is conducting the Topcon cross-linking clinical trial.
Dr. Mrochen has a financial interest in products related to corneal cross-linking.