The safety and efficacy of corneal cross-linking is leading to new methods and indications for the procedure.

Do you remember your dad’s old car? Or perhaps your grandfather’s car would be more appropriate? What did it look like? What did it sound like? How comfortable were those old bench seats? How safe were those seat belts (if it even had seat belts)? Now, imagine the newer models of vehicles that we see on the road today, along with all of the changes and advances that have taken place in the automobile industry.

We see that evolution taking place in almost every facet of both health care and eye care. Corneal collagen cross-linking (CXL) embodies that transformative illustration.

What we think of today as CXL was first developed in Dresden, Germany in the late 1990s.1 The groundbreaking procedure revealed a way to take a diseased cornea, such as one that has keratoconus, and make it stronger, reducing or potentially eliminating the need for a corneal transplant. Numerous studies quickly followed to test the procedure’s safety and efficacy in treating various types of corneal disorders including keratoconus,2 bullous keratopathy,3 and infectious keratitis. CXL quickly proved itself to be safe, easy to perform, and efficacious in many conditions; it quickly gained acceptance as the standard of care for keratoconus and post-excimer laser keratoectasia across Europe. In the European Union, CXL has been available in every country for more than a decade, and it has been available in Canada since 2008. U.S. Food and Drug Administration (FDA) clinical trials were initiated in April 2008, and the procedure was finally approved in the United States in April 2016.

However, that is only the beginning of the story; the technology continues to evolve, and new iterations and developments are underway that can continue to provide important benefits for patients and meet several needs. We will take a look at the history as well as the future of CXL.


CXL uses ultraviolet (UV) light at a specific wavelength (365nm to 370nm) to interact with riboflavin (vitamin B2) in the corneal stroma. CXL depends on having both of these constituents present in the stroma. The stroma has an absorption rate of UV-A light of 32%; however, in the presence of riboflavin, this increases to 95%.4 Riboflavin is a photoactive molecule, and in the presence of UV-A light, it generates free radicals in the form of reactive oxygen species. These free radicals then interact with the adjacent collagen fibrils to form covalent bonds between and within stromal collagen strands.

The additional scaffolding provided by these covalent bonds adds considerable strength to the cornea, causing it to be more resistant to ectasia. Early porcine, rabbit, and human models demonstrated that corneal rigidity increased between 70%5 and 300%6 in biomechanical stress-strain measurements. In addition, it has been noted that the post-cross-linked stromal collagen is more resistant to degradation via collagenase, pepsin, and trypsin.7 This enhancement of the corneal strength and structure sets the stage for CXL’s use in several areas of corneal disease.


It is well established that UV light can contribute to cataract formation as well as to corneal endothelial cell damage and dropout. Studies in rabbits revealed that the cytotoxic threshold of endothelial cells in the presence of riboflavin and UV-A light was 0.36mW/cm2 at the level of the endothelium.8 Understanding the absorption coefficient of the riboflavin-UV-A complex helped to determine the maximum irradiance level at the corneal epithelium of 3.0mW/cm2 for corneas thinner than 400 microns.

The resulting general guideline of utilizing this irradiance level in corneas that are at least 400 microns in thickness has helped to ensure the safety of the CXL procedure, resulting in only a few cases of endothelial damage since its inception. Because of the high level of absorption of the UV-A light by the riboflavin-saturated cornea during the irradiation phase of the treatment, the amount of UV-A reaching the anterior surface of the lens does not reach cytotoxic levels and thus does not lead to cataract formation. The potential risk of cataract has not born out to be a significant issue across two decades of CXL utilization.


The original Dresden protocol involved several steps.9 First, the cornea was denuded of epithelium to approximately 7mm. Next, 0.1% riboflavin in dextran solution was applied to the corneal surface five minutes prior to irradiation and every five minutes throughout irradiation. The saturation of riboflavin into the corneal stroma took place over approximately 30 minutes.

Prior to application of the UV light, the cornea is checked at the slit lamp to observe the appearance of “flashing” or “flare,” which indicates the presence of riboflavin in the anterior chamber. The clinical appearance of this flashing is not dissimilar to observing the presence of sodium fluorescein in the anterior chamber following tonometry. Once flashing is noted, the cornea is irradiated with UV light at an energy level of 3.0mW/cm2 for a period of 30 minutes.

Patients who underwent CXL noted a halting of progressive keratoectasia in both keratoconic and post-excimer ectasias. This was accompanied by reduced maximum keratometric readings, reduced mean cylinder, and, in some cases, improved vision within months of the procedure. Very few complications were noted in the early iterations of CXL; rarely, secondary treatments needed to be performed, and occasional patients would develop corneal haze.10

Needless to say, despite the success of the treatment and the relatively simple procedure, the hour-long process of corneal saturation and irradiation was cumbersome and lengthy, and patient recovery was slower than desired in a post-laser-assisted in situ keratomileusis (LASIK) world. Out of these needs came attempts to modify the procedure in several ways.


Since the inception of riboflavin-UV-A CXL, the corneal epithelium has presented a significant barrier to the absorption of riboflavin molecules by the stroma. The epithelial surface is hydrophobic, and the water-soluble riboflavin molecules cannot readily pass through an intact, non-manipulated epithelial surface. Thus, the first iterations of CXL removed the epithelium prior to instillation of riboflavin to improve the speed and homogeneity of riboflavin saturation. This left patients with a sizable epithelial defect, which was associated with some discomfort, and a small percentage of patients developed corneal haze.10,11 It was noted that in the first three to six months following CXL using the epi-off method that improvements in keratometric readings were delayed by epithelial remodeling.

Innovations to bypass removal of the corneal epithelium while allowing stromal saturation of riboflavin were developed, and experiments were performed to assess efficacy and impact on tissue physiology. Among the modifications to the original technique included alterations to the riboflavin solution. These ranged from examination of excipients and preservatives to assist in penetration12,13 to alterations of osmolar properties.14 Another method of maintaining an epi-on status is to “minimally manipulate” the epithelium to break up the tight junctions between adjacent epithelial cells and improve permeability.15

Outcomes of studies comparing epi-on versus epi-off vary, with the majority of studies showing good efficacy in the epi-on groups, but with a slightly higher risk of progression compared to epi-off. Several studies report equivalent outcomes between the two groups with improved patient comfort in those receiving epi-on treatment.16,17 Other studies show either a decrease in the flattening effect on maximum keratometric readings or increased risk for retreatment (i.e., decreased effect) in the epi-on group compared to epi-off.18-20

More recently, iontophoresis has been used to deliver riboflavin across the epithelium and into the corneal stroma.21-23 This method utilizes an electric current to pass the negatively charged riboflavin molecules across the epithelial barrier while still maintaining epithelial integrity. The iontophoresis method decreases the saturation time to between five and 10 minutes, which significantly decreases overall treatment time.

While most studies show that iontophoresis-assisted CXL does not have quite the same level of efficacy as standard Dresden CXL, it still seems to be a promising method to provide a more rapid and comfortable means of loading the cornea with riboflavin;21-23 more studies will examine whether this method will prove useful in decreasing the loading time for those undergoing CXL.


As discussed previously, although clinical trials began in the United States in April 2008, the FDA finally approved the CXL procedure in April 2016. Specifically, approval was granted for one company’s patented form of two riboflavin formulations in addition to its UV light-emitting unit that uses the standard 3mW/cm2 irradiance level. Of particular note, the FDA granted approval for keratoconus and post-LASIK ectasia using the Dresden protocol, which means that physicians who elect to perform this procedure in an on-label fashion will be utilizing an epithelial-off process with a 9mm denuding of epithelium, followed by a drop of riboflavin solution every two minutes for 30 minutes (Figure 1), and then irradiation and repeating the riboflavin drops every two minutes throughout irradiation (Figures 2 and 3).

Figure 1. The riboflavin solution is applied to the debrided corneal surface prior to irradiation.
Courtesy of Avedro

Figure 2. The riboflavin solution is applied every two minutes throughout irradiation.
© 2017 Shawn Rocco/Duke Health

Figure 3. Once flare is noted, the cornea is irradiated with UV light at an energy level of 3.0mW/cm2 for 30 minutes.
Courtesy of Avedro

While the approval in the United States is a definite boon for patients, increasing access to the much-needed technology, it still is an incomplete picture regarding the full scope of uses and options for the technology, including further investigation into epi-on procedures and utilization of CXL in areas outside of keratoectatic disease, such as infectious keratitis. While these uses are outside of the FDA-approved guidelines, it is important to remember that the European Union had universally accepted CXL as the standard of care a full two years before the FDA trials in the United States had begun; much of the research regarding safety and efficacy has been and continues to be generated outside of the United States.


The fact that the traditional Dresden protocol takes a total of one hour to complete is a serious drawback to widespread utilization of the technology, especially in the younger population, for whom it may be the most beneficial. Currently, the procedure is time-consuming for both patients and practitioners. It is not surprising then that several alterations to the original energy settings are being examined for safety and efficacy in patients who have refractive instability.

These accelerated protocols are based on the Bunsen-Roscoe law of photochemical reciprocity, which states that the same photochemical effect may be achieved by reducing the irradiation interval, provided that the total energy level is kept constant by increasing irradiation intensity.24 For example, 10mW/cm2 for nine minutes, 18mW/cm2 for five minutes, and 30mW/cm2 for three minutes, each at a constant dose of 5.4J/cm2, may all have the same photochemical impact as the standard dosage of 3mW/cm2 for 30 minutes.

Wernli et al showed that there may also be a limit to the Bunsen-Roscoe law for CXL.25 In an in-vitro experiment, they found that with intensities beyond 40mW/cm2 to 50mW/cm2 and illumination times shorter than two minutes, corneal tissue did not demonstrate the same strengthening effect seen in the original protocols. This may provide a ceiling for how “fast” the procedure can be performed using the current technology. The majority of the studies utilizing accelerated protocols now fall beneath that threshold.

The most common alterations to treatment time and irradiance intensity fall into three subcategories. Those being studied most intently are the reduction from 30 minutes to 10, five, or three minutes, with corresponding irradiance levels of 9mW/cm2, 18mW/cm2, and 30mW/cm2, respectively. Outcomes generally seem favorable in the 10min/9mW/cm2 group at 12 months, without any adverse events in the accelerated study group.26,27

Razmjoo et al also performed a six-month study examining the 18mW/cm2 irradiance at five minutes compared to the original irradiance level of the Dresden protocol.28 They found no statistically significant differences in visual acuity, refractive, or topographic results at six months postoperatively.

Shetty et al compared the conventional Dresden protocol with each of the durations/irradiance levels mentioned above. Only the 10min/9mW/cm2 showed comparable efficacy in corneal flattening; however, all of the accelerated groups had reduced endothelial cell density.29

There have also been studies showing that the traditional Dresden protocol has superior results compared to accelerated protocols with respect to efficacy.30

There is a definite desire from a clinical efficiency standpoint to decrease procedure time. However, with the questionable efficacy of current accelerated protocols in combination with potential safety issues and concerns regarding the corneal endothelium, further longitudinal studies are needed to determine the place of accelerated CXL in patient care.


Discussing CXL with patients in a clinical setting initially tended to revolve around halting disease progression; namely, their keratoconus or keratoectatic disease would no longer become worse with time. This is a tremendous benefit to patients from a clinical standpoint for avoiding more invasive surgery; however, patient expectations also often include the perception that their vision will improve following the treatment. Thus, the discussion usually included that while their disease would be halted, they likely would not notice any real improvements in their vision. But, this is changing.

Kanellopoulos and Binder described the seminal case in 2007 for the successful combination of CXL and photorefractive keratectomy (PRK),31 which noted improved visual acuity in addition to halting of the keratoectatic process. Since then, CXL has been studied in combination with LASIK, intracorneal ring segments, and conductive keratoplasty for better refractive results. There has been growing acceptance that these technologies can be utilized either in combination or in close sequential order, with CXL typically being performed first. One study showed comparable stability over a three-year period between eyes that underwent CXL alone versus CXL and topography-guided PRK.32 Considering that CXL improves the strength of the cornea and reduces the susceptibility to ectasia, it is not surprising that the majority of the results regarding its combined use with refractive surgery have been positive.


The advantage of utilizing CXL for refractive procedures is that it is a non-ablative procedure; no tissue is removed to accomplish the alterations in corneal curvature that may lead to improvement in vision. The technique by which CXL is patterned to obtain refractive improvement is known as photorefractive intrastromal corneal collagen cross-linking (colloquially as PiXL). It is currently being studied in Europe and Canada and is used to treat low levels of myopia under 3.00DS, without astigmatism. The application of the patterned UV-A light causes a tightening of the cornea in the region of the light, and this causes a flattening of the corneal curvature, similar to what is noted with excimer laser surgery for myopia.

Nordström et al published one-year results of topography-guided CXL for keratoconus, showing results of PiXL versus traditional CXL on patients who had progressive keratoconus.33 Significant improvements were noted in spherical refractive error and in visual acuity at three, six, and 12 months following treatment in the PiXL group, but not in the traditional CXL group. No adverse effects were noted in the PiXL group.


It has long been accepted that UV light itself has a disinfectant/biocidal property, as the short-wavelength light is disruptive to the DNA of living organisms. Combined with the collagen strengthening properties inherent to the CXL procedure, it was quickly understood that there may be a benefit to using CXL in more severe forms of infectious keratitis, which may lead to better outcomes. PACK-CXL is the acronym for photoactivated chromophore for infectious keratitis—corneal cross-linking, which is the current term for describing the utilization of riboflavin-UV-A CXL for infectious keratitis.

Iseli et al examined a cohort of patients who had corneal melting associated with infectious keratitis.34 In this small study, patients presenting with organisms resistant to topical and systemic therapy were treated with the standard Dresden protocol. Corneal melting was halted in all but one patient.

Makdoumi et al treated 16 patients who had bacterial keratitis using CXL with the standard Dresden protocol as first-line therapy.35 In this study, 15 patients showed complete closure of the epithelial defect, and all showed improved inflammatory reaction. This step toward utilizing CXL as a first-line therapy in infectious keratitis was a significant one.

A recent meta-analysis of CXL for infectious keratitis showed promise across nearly all forms of infectious keratitis with the exception of viral keratitis.36 With concerns of bacterial resistance to antibiotics and the cost of developing new antibiotic therapies, the future looks bright for utilizing CXL as a potential first-line therapy in cases of bacterial and other forms of infectious keratitis.


CXL has been a transformational procedure and has demonstrated exceptional safety and efficacy since its inception. Widespread acceptance and utilization of the original technique has led to further investigation into several modifications of the protocol to improve efficiency without sacrificing its original efficacy.

In addition, an understanding of the flexibility of the technology has led to exploration and discovery of additional potential for CXL technology. While the “Oldsmobile” version of CXL has been approved in the United States for more than a year, the modifications and advances that will make it a more feasible, flexible, and faster procedure are imminent on the horizon. It is CXL—full speed ahead! CLS


  1. Spörl E, Huhle M, Kasper M, Seiler T. [Increased rigidity of the cornea caused by intrastromal cross-linking]. Ophthalmologe. 1997 Dec;94:902-906.
  2. Wollensak G, Spoerl E, Seiler T. Riboflavin/ultraviolet-a-induced collagen crosslinking for the treatment of keratoconus. Am J Ophthalmol. 2003 May;135:620-627.
  3. Krueger RR, Ramos-Esteban JC, Kanellopoulos AJ. Staged delivery of riboflavin with UVA cross-linking in advanced bullous keratopathy: laboratory investigation and first clinical case. J Refract Surg. 2008 Sep;24:S730-S736.
  4. Michael R. Development and repair of cataract induced by ultraviolet radiation. Ophthalmic Res. 2000;32 Suppl 1:ii-iii, 1-44.
  5. Spoerl E, Seiler T. Techniques for stiffening the cornea. J Refract Surg. 1999 Nov-Dec;15:711-713.
  6. Wollensak G, Spoerl E, Seiler T. Stress-strain measurements of human and porcine corneas after riboflavin-ultraviolet-A induced cross-linking. J Cataract Refract Surg. 2003 Sep;29:1780-1785
  7. Spoerl E, Wollensak G, Seiler T. Increased resistance of crosslinked cornea against enzymatic digestion. Curr Eye Res. 2004;29:35-40.
  8. Wollensak G, Spoerl E, Wilsch M, Seiler T. Endothelial cell damage after riboflavin-ultraviolet-A treatment in the rabbit. J Cataract Refract Surg. 2003 Sep;291786-1790.
  9. Wollensak G, Spörl E, Seiler T. [Treatment of keratoconus by collagen cross linking]. Ophthalmologe. 2003 Jan;100:44-49.
  10. Mazzota C, Balestrazzi A, Baiocchi S, Traversi C, Caporossi A. Stromal haze after combined riboflavin-UVA corneal collagen cross-linking in keratoconus: in vivo confocal microscopic evaluation. Clin Exp Ophthalmol. 2007 Aug;35:580-582.
  11. Raiskup F, Hoyer A, Spoerl E. Permanent corneal haze after riboflavin-UVA-induced cross-linking in keratoconus. J Refract Surg. 2009 Sep;25:S824-S828.
  12. Torricelli AA, Ford MR, Singh V, Santhiago MR, Dupps WJ Jr, Wilson SE. BAC-EDTA transepithelial riboflavin-UVA crosslinking has greater biomechanical stiffening effect than standard epithelium-off in rabbit corneas. Exp Eye Res. 2014 Aug;125:114-117.
  13. Koppen C, Wouters K, Mathysen D, Rozema J, Tassignon MJ. Refractive and topographic results of benzalkonium chloride assisted transepithelial crosslinking. J Cataract Refract Surg. 2012 Jun;38:1000-1005.
  14. Raiskup F, Pinelli R, Spoerl E. Riboflavin osmolar modification for transepithelial corneal cross-linking. Curr Eye Res. 2012 Mar;37:234-8.
  15. Rechichi M, Daya S, Scorcia V, Meduri A, Scorcia G. Epithelial-disruption collagen crosslinking for keratoconus: one-year results. J Cataract Refract Surg. 2013 Aug;39:1171-1178.
  16. Magli A, Forte R, Tortori A, Capasso L, Marsico G, Piozzi E. Epithelium-off corneal collagen cross-linking versus transepithelial cross-linking for pediatric keratoconus. Cornea. 2013 May;32:597-601.
  17. Hirji N, Sykakis E, Lam FC, et al. Corneal collagen crosslinking for keratoconus or corneal ectasia without epithelial debridement. Eye (Lond). 2015 Jun;29:764-768.
  18. Kocak I, Aydin A, Kaya F, Koc H. Comparison of transepithelial corneal collagen crosslinking with epithelium-off crosslinking in progressive keratoconus. J Fr Ophthalmol. 2014 May;37:371-376.
  19. Rush SW, Rush RB. Epithelium-off versus transepithelial corneal collagen crosslinking for progressive corneal ectasia: a randomised and controlled trial. Br J Ophthalmol. 2017 Apr;101:503-508.
  20. Çerman E, Toker E, Ozarslan Ozcan D. Transepithelial versus epithelium-off crosslinking in adults with progressive keratoconus. J Cataract Refract Surg. 2015 Jul;41:1416-25.
  21. Bikbova G, Bikbov M. Standard corneal collagen crosslinking versus transepithelial iontophoresis-assisted corneal crosslinking, 24 months follow-up: randomized controlled trial. Acta Ophthalmol. 2016 Nov;94:e600-e606.
  22. Vinciguerra P, Romano V, Rosetta P, et al. Transepithelial iontophoresis versus standard corneal collagen cross-linking: 1-year results of a prospective clinical study. J Refract Surg. 2016 Oct 1;32:672-678.
  23. Lombardo M, Giannini D, Lombardo G, Serrao S. Randomized Controlled Trial Comparing transepithelial corneal cross-linking using iontophoresis with the Dresden protocol in progressive keratoconus. Ophthalmology. 2017 Jun;124:804-812.
  24. Kymionis GD, Kontadakis GA, Hashemi KK. Accelerated versus conventional corneal crosslinking for refractive instability: an update. Curr Opin Ophthal. 2017 Jul;28:343-347.
  25. Wernli J, Schumacher S, Spoerl E, Mrochen M. The efficacy of corneal cross-linking shows a sudden decrease with very high intensity UV light and short treatment time. Invest Ophthalmol Vis Sci. 2013 Feb 1;54:1176-1180.
  26. Cummings AB, McQuaid R, Naughton S, Brennan E, Mrochen M. Optimizing corneal cross-linking in the treatment of keratoconus: a comparison of outcomes after standard- and high-intensity protocols. Cornea. 2016;35:814-822.
  27. Sadoughi MM, Einollahi B, Baradaran-Rafii A, Roshandel D, Hasani H, Nazeri M. Accelerated versus conventional corneal collagen cross-linking in patients with keratoconus: an intrapatient comparative study. Int Ophthalmol. 2016 Dec 29. [Epub ahead of print]
  28. Razmjoo H, Peyman A, Rahimi A, Modrek HJ. Cornea collagen cross-linking for keratoconus: a comparison between accelerated and conventional methods. Adv Biomed Res. 2017 Feb 22;6:10.
  29. Shetty R, Pahuja NK, Nuijts RM, et al. Current protocols of corneal collagen cross-linking: visual, refractive, and tomographic outcomes. Am J Ophthalmol. 2015 Aug;160:243-249.
  30. Hashemi H, Miraftab M, Seyedian MA, et al. Long-term results of an accelerated corneal cross-linking protocol (18mW/cm2) for the treatment of progressive keratoconus. Am J Ophthalmol. 2015 Dec;160:1164-1170.
  31. Kanellopoulos AJ, Binder PS. Collagen cross-linking with sequential topography-guided PRK: a temporizing alternative for keratoconus to penetrating keratoplasty. Cornea. 2007 Aug;26:891-895.
  32. Kontadakis GA, Kankariya VP, Tsoulnaras K, Pallikaris AI, Plaka A, Kymionis GD. Long-term comparison of simultaneous topography-guided photorefractive keratectomy followed by corneal cross-linking versus corneal cross-linking alone. Ophthalmology. 2016 May;123:974-983.
  33. Nordström M, Schiller M, Fredriksson A, Behndig A. Refractive improvements and safety with topography-guided corneal crosslinking for keratoconus: 1-year results. Br J Ophthlamol. 2017 Jul;101:920-925.
  34. Iseli HP, Thiel MA, Hafezi F, Kampmeier J, Seiler T. Ultraviolet A/riboflavin corneal cross-linking for infectious keratitis associated with corneal melts. Cornea. 2008 Jun;27:590-594.
  35. Makdoumi K, Mortensen J, Sorkhabi O, Malmvall BE, Crafoord S. UVA-riboflavin photochemical therapy of bacterial keratitis: a pilot study. Graefes Arch Clin Exp Ophthalmol. 2012 Jan;250:95-102.
  36. Papaioannou L, Miligkos M, Papathanassiou M. Corneal collagen cross-linking for infectious keratitis: a systematic review and meta-analysis. Cornea. 2016 Jan;35:62-71.