June 25, 2011
14 min read

Epithelium-on cross-linking growing while epithelium-off seeks FDA approval

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

Comparing the proven way to the latest way has always made for a tough choice. Some would gladly stick with the tried-and-true, while early adopters always push for more and better despite potential risks.

Ophthalmologists are plowing ahead with research about epithelium-on cross-linking, even though epithelium-off methods are not yet approved in the United States. Epithelium-on techniques are controversial even in the European Union, where epithelial-off procedures have been the standard of care for years and are approved for use.

Both modalities are used under regulatory investigational exemptions in the United States, but removing the epithelium has been the standard of care in Europe for 5 years, with 7 years of prior research to back up its approval there.

Yet, epithelium-on procedures promise efficacy with reduced pain, quicker healing and quicker visual acuity improvement, which has led some to engage in clinical trials to prove not only that it works, but also that it can be done with less riboflavin preparation penetrating the stroma and less exposure time of the ultraviolet A (UVA) light source than has been used elsewhere.

A. John Kanellopoulos, MD
With more than a decade of experience with this technique, A. John Kanellopoulos, MD, sees cross-linking as an integral part of customizing corneal refractive procedures.
Image: Kanellopoulos AJ

William B. Trattler, MD, is helping run the CXL-USA Study Group multicenter clinical trial, which has a dozen investigational sites and began looking at epithelium-on procedures starting in May 2010. He said that his results with epithelium-on cross-linking show that visual results are as good as epithelium-off techniques but with much faster visual recovery.

“With epithelial-on cross-linking, patients are more comfortable during the early postoperative period, with a faster visual recovery, lower risk of infection and a faster return to contact lens wear,” he said. In contrast, with epithelium off, the recovery is similar to PRK, and patients need to wait for 1 to 2 weeks to return to RGP contact lenses.

“With epi-on cross-linking, we do not even use a bandage contact lens,” Dr. Trattler, OSN SuperSite Board Member, said.

But approval from the U.S. Food and Drug Administration for cross-linking needs to come first. Once both procedures receive FDA approval, then head-to-head randomized clinical trials could show which method is better.

“People would love to see an epithelium-on technique that works,” said Peter S. Hersh, MD, medical monitor to Avedro Inc., which is currently completing data analysis on multicenter phase 3 studies of epithelium-off corneal collagen cross-linking for the treatment of progressive keratoconus and post-LASIK ectasia.

Peter S. Hersh, MD
Peter S. Hersh

“To date however, there is little real evidence and virtually nothing from the point of view of controlled clinical trials in the literature that looks at the results of epithelial-on cross-linking,” Dr. Hersh, an OSN Refractive Surgery Board Member, said. “In contrast, there is extensive literature on the results of epi-off cross-linking. For instance, in our study, we found stability or topography improvement in 96% of eyes, with an average flattening of maximum keratometry of 1.7 D and average improvement in visual acuity measures. With increased interest in epi-on cross-linking, I’m very sanguine that companies will ultimately get to the methodology and delivery system that will allow us to do it. For instance, Avedro is currently working on a higher-power UV light source that may have advantages in epi-on cross-linking, and Topcon is working with a new epi-on riboflavin formulation.”

Off or on?

Cross-linking was first performed in Europe in 1999, and by September 2006, it had been approved by the European Union. There, per the Dresden protocol, the epithelium is removed, the cornea and the anterior chamber are soaked with dextran-based 0.1% riboflavin solution, and UVA light is delivered via a 3 mW/cm2 source for 30 minutes.

UVA light interacts with the riboflavin, producing reactive oxygen species that create chemical bonds between and within corneal collagen fibrils, making them stiffer and increasing corneal strength and stability by inducing new cross-links between or within collagen fibers.

Many studies have shown the halt of keratoconus progression and even the gains of one or more lines of best corrected visual acuity 1 year postoperatively.

But one issue confounds ophthalmologists, in either epithelium-off or epithelium-on procedures: How exactly does cross-linking work?

A. John Kanellopoulos, MD, was heavily involved in the European experience with cross-linking, with more than a decade of experience with the procedure that has been extensively presented and published. But he noted that while the stiffening of the cornea has been measured through sophisticated mechanics, nobody has shown that this is what happens to collagen on the molecular level. For example, he said, cross-linking’s effect could be related to a shrinking of individual collagen fiber rather than a cross-linking between fibers.

“There’s obviously some validity in higher amount of cross-linking with epithelium-off,” Dr. Kanellopoulos, an OSN Europe Edition Editorial Board Member, said. “The problem is, we are still limited by the fact that we cannot measure how much we are cross-linking. So we are all awaiting a measuring method and potentially measuring unit to quantitate the cross-linking effect of each individual technique.”

Epithelium-on procedure

The modification in the epithelium-on procedure is to pass riboflavin through about 50 µm of epithelium and into the stroma. Dextran is a large molecule, and ophthalmologists in Europe have been dissatisfied with the penetration available, said Parag A. Majmudar, MD, one of the surgeons working on the CXL-USA study of epithelium-on procedures and an OSN SuperSite Board Member. So in the U.S., ophthalmologists dissolve the riboflavin in gum cellulose.

Parag A. Majmudar, MD
Parag A. Majmudar

Surgeons may also use benzalkonium chloride to loosen up tight junctions in the epithelial cells that normally prevent the diffusion of molecules across the epithelium, Dr. Majmudar said. If the cornea is less than 400 µm thick, they will use sterile water or, more commonly, a hypotonic riboflavin solution to swell it to that minimum thickness.

The riboflavin saturation takes a little more time to occur, up to 60 minutes instead of 30 minutes with an epithelium-off procedure, Dr. Majmudar said. He does not use a corneal sponge to instill the riboflavin uniformly, although other surgeons do utilize this.

The UVA exposure is still 3 mW/cm2 for 30 minutes, per the investigational protocol, but there is already interest in intensifying the UVA to 6 mW/cm2 for 15 minutes or even 9 mW/cm2 for 10 minutes.

Dr. Majmudar is awaiting his own 1-year results but said that so far epithelium-on procedures have offered at least as good a result as results in the literature with epithelium-off procedures, with patients generally gaining a line of BCVA with quicker and less painful results.

“I would use a bandage contact lens after the UVA light procedure if there were a significant irregularity to the epithelium, if it looks ‘ratty’ or ‘beat up,’ or if we feel that there’s any concern of a patient having a significant amount of pain,” Dr. Majmudar said. “If the patient is in any discomfort, we’ll put a bandage contact lens on prophylactically. But that’s been less than 5% of the time that I’ve had to do that.”

Hybrid procedure

Eric D. Donnenfeld, MD, an OSN Cornea/External Disease Board Member, was part of the original FDA trial on cross-linking and performed a series of epithelium-off procedures 4 years ago. Concerned about the long healing process that occurs after removing the epithelium, he sought a way to expedite it.

As a modification of the epithelium-off procedure, he designed a device with 80-µm prongs that is pressed onto the cornea without penetrating into the stroma or Bowman’s membrane. (The device is a concept similar to one developed simultaneously and independently by Sheraz Daya, MD.)

Dr. Donnenfeld then applies 2% riboflavin drops every 2 minutes for 30 minutes. If the patient has a thinner cornea, he uses hypotonic riboflavin and then after 30 minutes checks pachymetry and fluorescence in the stroma. He applies UVA light for 30 minutes of 4 mW/cm2. He then applies steroids, antibiotics and a bandage contact lens.

“Most of the research that I’ve looked at, the work that has come out of several labs, has shown that you don’t get the good penetration of riboflavin without removing the epithelium,” Dr. Donnenfeld said. In his youngest patients who have the most aggressive forms of keratoconus, he still performs the traditional epithelium-off procedure. But for the other 95% of cross-linking patients, he performs the hybrid procedure with his device.

He said he will not switch to an epithelium-on procedure.

“Until someone shows me a good epithelium-on study where you get equal amounts of corneal cross-linking, I’ll continue to completely remove the epithelium or partially remove the epithelium,” he said.

Dr. Trattler said that one of the main differences in the epithelium-on technique being used in the CXL-USA study is that the researchers do not stick to the 30-minute riboflavin loading time used in Europe with both epithelium on and epithelium off.

William B. Trattler, MD
William B. Trattler

“Rather, our experience has been that 60 to 80 minutes of riboflavin loading is required to ensure saturation of the corneal stroma to a similar extent seen with epi-off riboflavin loading,” he said. “With the longer loading time, our patients with epi-on have similar results to our epi-off patients in our study, and the results can be seen looking at the improvement in the corneal shape as seen on difference maps of the preoperative and postoperative cornea.”

“When we look at topography/tomography, it may not be that obvious from one map to another that there’s been a change,” Dr. Trattler said. “A difference map shows that the cornea got flatter in some areas and steeper in other areas. The reshaping of the cornea seen with cross-linking can result in improvement in UCVA and BCVA.”


Haze has been noted in cross-linking. Surgeons have not been able to pinpoint why it occurs or what its clinical effect might be.

“I have seen patients develop haze with epithelium-off procedures,” Dr. Trattler said. “Visually significant haze with epithelial off is fortunately uncommon and responds to treatment with topical steroids, similar to the way that PRK-related haze responds to topical steroid treatment.”

Dr. Hersh described the haze that results in cross-linking as something the physician sees under the slit lamp, not something the patient sees. He has used Scheimpflug densitometry to quantify the degree of haze and has found that it is greatest in the first month, plateaus at 3 months and then significantly decreases over the next 3 months. It then returns to baseline at 1 year in most cases.

“In our clinical trial, haze did not correlate with changes in visual acuity, did not affect visual recovery time and did not affect the 1-year cross-linking topography outcomes,” Dr. Hersh said.

“A significant proportion of patients that have the epithelium-off process will have haze in the cornea, and that is likely a function of the epithelium being removed, not necessarily an indication that cross-linking is working,” Dr. Majmudar said. “The biggest misconception that other surgeons have is that if you don’t see haze, cross-linking did not take effect. But I don’t think that’s true. I’ve seen results in our epithelium-on cases in which there’s been no haze and we’re seeing results that are similar to epithelial on.”

Dr. Hersh said that although the source of haze is associated in some way with the healing process after cross-linking, its ultimate effect on positive or negative differences between epithelium-off and epithelium-on cross-linking remains a mystery. Is the haze secondary to the loss of keratocytes or to their repopulation? Is it related to collagen or changes in lamellar architecture? Is it keratocytic activity or something that involves the entire wound healing process? Or is it simply secondary to the corneal thinning seen after the procedure?

“We only know the rudiments of what collagen cross-linking is doing,” Dr. Hersh said. “This is something that is early in its development, and I think you’re going to see great advances.”

New uses for cross-linking

Francis W. Price Jr., MD, an OSN Cornea/External Disease Board Member, has been using epithelium-off cross-linking under a physician-sponsored investigational device exemption to treat infectious keratitis, primarily bacterial but also including fungal and Acanthamoeba infections.

He limits treatments to smaller areas of obvious infection, taking care to mask clear cornea and the limbus. Epithelium-on treatments do not deliver as much riboflavin into the cornea and decrease the penetration of the UV light with the intensity to do what is necessary for cross-linking, so he sticks with epithelium-off procedures.

“That’s particularly important for fungus and Acanthamoeba, which are full-thickness infections,” he said. Cross-linking treatments appear to get their effect by creating reactive oxygen molecules, or singlet oxygen, which is cytotoxic.

Riboflavin is a flat molecule that slides into the RNA and DNA in cells, disrupting them once the UVA light reaches it. This kills all the keratocytes in the area of treatment.

“The interesting thing is that UVA is used alone to sterilize fluids,” Dr. Price said, comparing it to simple pen-shaped devices that can be dropped into canteens to sterilize water while camping, as well as medical uses to sterilize blood products.

Dr. Kanellopoulos sees a future for cross-linking as a method of prophylaxis for ectasia in LASIK and PRK procedures. By definition, LASIK weakens the cornea, and by definition, cross-linking strengthens it.

“Cross-linking in general was born as a remedy of corneal ectasia,” he said. “In my mind, it is an integral part of customizing corneal refractive procedures.”

Dr. Kanellopoulos has also employed cross-linking for temporary relief in patients with bullous keratopathy, and he has routinely employed cross-linking’s modulation of keratocytes in patients with severe corneal scarring. Specifically, he is addressing several cases of severe corneal blindness with partial topography-guided PRK and cross-linking via the Athens protocol.

Dr. Kanellopoulos has shown that placing riboflavin in a corneal pocket created by a femtosecond laser and concentrating cross-linking in that area to treat early stages of keratoconus appears to be equally effective as the Dresden protocol. On the horizon, he foresees wider applications of the techniques he has introduced and has experimented with UVA light fluences of 5 mW/cm2, 6 mW/cm2, 7 mW/cm2 and 12 mW/cm2, which have created quite a buzz in the past 6 months, he said.

Epithelium-on potential

But again, the new modalities of cross-linking are unproven, compared with the epithelium-off Dresden protocol, “the standard of care in that it is the only protocol that carries all the basic science with it.” Dr. Kanellopoulos said.

Dr. Hersh said, “I most certainly embrace doing epithelium-on cross-linking within the confines of a controlled clinical trial. It would make life much easier. It clearly would be easier for the patient, because you’d be able to get the patient into a contact lens more quickly. There’s less of a potential risk of infection because there’s an epithelial barrier that’s been broken.”

At the crux of the matter is whether epithelium-on techniques provide as much efficacy as epithelium-off techniques. Epithelium-on methods allow less riboflavin into the cornea and absorb more UVA light before it reaches the stroma. But, Dr. Hersh said, does that mean there is less clinical effect?

“I can’t tell you how frustrating it was for all these years. Patients would come in with keratoconus, and I would tell them there’s nothing we can do aside from a good contact lens fitting: ‘When you need to have a transplant, come see me,’” Dr. Majmudar said. “Even if you don’t do anything else, we can stabilize and improve in the majority of cases, and later on, if you have to add additional treatments, such as Intacs (Addition Technology) or topo-guided PRK, you can have additive treatment on top of the building block of cross-linking.”

“I have no doubt that the science will get better and better,” Dr. Majmudar said, “and some of these preconceptions that people have of epithelium-on vs. epithelium-off will diminish once people see what kinds of results are possible. And some of the results that we’re seeing are just phenomenal.” – by Ryan DuBosar

What should be the goal for a standard of care in cross-linking?

Lindstrom's Perspective
Collagen cross-linking a great advancement for surgeons, patients


  • Boxer Wachler BS, Pinelli R, Ertan A, Chan CC. Safety and efficacy of transepithelial crosslinking (C3-R/CXL). J Cataract Refract Surg. 2010;36(1):186-188; author reply 188-189.
  • Greenstein SA, Fry KL, Bhatt J, Hersh PS. Natural history of corneal haze after collagen crosslinking for keratoconus and corneal ectasia: Scheimpflug and biomicroscopic analysis. J Cataract Refract Surg. 2010;36(12):2105-2114.
  • Greenstein SA, Shah VP, Fry KL, Hersh PS. Corneal thickness changes after corneal collagen cross-linking for keratoconus and corneal ectasia: one-year results. J Cataract Refract Surg. 2011;37(4):691-700.
  • Hersh PS, Greenstein SA, Fry KL. Corneal collagen cross-linking for keratoconus and corneal ectasia: One-year results. J Cataract Refract Surg. 2011;37(1):149-160.
  • Kanellopoulos AJ. Collagen cross-linking in early keratoconus with riboflavin in a femtosecond laser-created pocket: initial clinical results. J Refract Surg. 2009;25(11):1034-1037.
  • Kanellopoulos AJ. Comparison of sequential vs. same-day simultaneous collagen cross-linking and topography-guided PRK for treatment of keratoconus. J Refract Surg. 2009;25(9):S812-S818.
  • Kanellopoulos AJ. Cross-linking plus topography-guided PRK for post-LASIK ectasia management. In: Garg A, Alió JL, Lin JT, et al, eds. Mastering the Advanced Surface Ablation Techniques. New Delhi, India: Jaypee Brothers Medical Publishers; 2008:204-214.
  • Kanellopoulos AJ. Cross-linking plus topography guided PRK for post-LASIK ectasia management. In: Garg A, Pinelli R, Kanellopoulos AJ, O’Brart D, Lovisolo CF, eds. Mastering Corneal Collagen Cross Linking Techniques (C3-R/CCL/CxL). New Delhi, India: Jaypee Brothers Medical Publishers; 2009:69-80.
  • Kanellopoulos AJ. Cross-linking plus topography-guided PRK for post-LASIK ectasia management. In: Garg A, Rosen E, Goes FJ, et al, eds. Instant Clinical Diagnosis in Ophthalmology: Refractive Surgery. New Delhi, India: Jaypee Brothers Medical Publishers; 2009:258-269.
  • Kanellopoulos AJ. IntraLase-assisted LASIK, Video V; Comparison of topography guided to standard LASIK for hyperopia, Video VI; Limited topoguided PRK followed by collagen cross linking for keratoconus, Video VII. In: Garg A, Fine IH, Pallikaris IG, et al, eds. Video Atlas of Ophthalmic Surgery. New Delhi, India: Jaypee Brothers Medical Publishers; 2008.
  • Kanellopoulos AJ. Post-LASIK ectasia. Ophthalmology. 2007;114(6):1230.
  • Kanellopoulos AJ. PRK and C3-R. In: Boxer Wachler BS, ed. Modern Management of Keratoconus. New Delhi, India: Jaypee Brothers Medical Publishers; 2008:219-228.
  • Kanellopoulos AJ, Binder PS. Collagen cross-linking (CCL) with sequential topography-guided PRK: A temporizing alternative for keratoconus to penetrating keratoplasty. Cornea. 2007;26(7):891-895.
  • Kanellopoulos AJ, Binder PS. Management of corneal ectasia after LASIK with combined, same-day, topography-guided partial transepithelial PRK and collagen cross-linking: The Athens protocol. J Refract Surg. 2011;27(5):323-331.
  • Krueger RR, Kanellopoulos AJ. Stability of simultaneous topography-guided photorefractive keratectomy and riboflavin/UVA cross-linking for progressive keratoconus: case reports. J Refract Surg. 2010;26(10):S827-S832.
  • Krueger RR, Ramos-Esteban JC, Kanellopoulos AJ. Staged intrastromal delivery of riboflavin with UVA cross-linking in advanced bullous keratopathy: laboratory investigation and first clinical case. J Refract Surg. 2008;24(7):S730-S736.
  • Raiskup-Wolf F, Hoyer A, Spoerl E, Pillunat LE. Collagen crosslinking with riboflavin and ultraviolet-A light in keratoconus: long-term results. J Cataract Refract Surg. 2008;34(5):796-801.
  • Vinciguerra P, Albè E, Trazza S, Seiler T, Epstein D. Intraoperative and postoperative effects of corneal collagen crosslinking on progressive keratoconus. Arch Ophthalmol. 2009;127(10):1258-1265.
  • Wollensak G, Spoerl E, Seiler T. Riboflavin/ultraviolet-A-induced collagen crosslinking for the treatment of keratoconus. Am J Ophthalmol. 2003;135(5):620-627.

  • Eric D. Donnenfeld, MD, can be reached at OCLI, 2000 North Village Ave., Rockville Centre, NY 11570; 516-766-2519; email: eddoph@aol.com.
  • Peter S. Hersh, MD, FACS, can be reached at The Cornea and Laser Eye Institute, 300 Frank W. Burr Blvd., Suite 71, Teaneck, NJ 07666; 201-883-0505; email: phersh@vision-institute.com.
  • A. John Kanellopoulos, MD, can be reached at 115 East 61st St., New York, NY 10065; 917-770-0586; email: ajk@brilliantvision.com.
  • Parag A. Majmudar, MD, can be reached at Chicago Cornea Consultants, 1585 N. Barrington Road, Doctors Building 2, Suite 502, Hoffman Estates, IL 60169; 847-882-5909; email pamajmudar@chicagocornea.com.
  • Francis W. Price Jr., MD, can be reached at Price Vision Group, 9002 N Meridian St., Suite 100, Indianapolis, IN 46260; 317-814-2990; email: fprice@pricevisiongroup.net.
  • William B. Trattler, MD, can be reached at the Center for Excellence in Eye Care, 8940 N. Kendall Drive, Suite 400E, Miami, FL 33176; 305-598-2020; email: wtrattler@gmail.com.
  • Disclosures: Drs. Donnenfeld, Majmudar and Price have no relevant financial disclosures. Dr. Hersh is medical monitor to Avedro Inc. Dr. Kanellopoulos is on the speakers bureau for Priavision and Seros Medical. Dr. Trattler is a consultant to CXL-USA.