September 2017


Update on Crosslinking
Refractive opportunities for crosslinking

by Liz Hillman EyeWorld Staff Writer

Dr. Elling and his group perform epi-on refractive crosslinking (PiXL) with an oxygen supply to boost its effect. The patient wears special oxygen goggles, and in the epithelial prep stage, a riboflavin soaked sponge is applied to the treatment area before starting riboflavin drops. This procedure uses a pulsed ultraviolet light.
Source: Matthias Elling, MD

A look at applications for crosslinking beyond keratoconus

Collagen crosslinking has already established its value in its ability to halt progressive keratoconus and strengthen corneas weakened by post-refractive surgery ectasia. The quality of vision is not directly addressed though with the standard Dresden protocol, whose main purpose is simply biomechanical strengthening.
However, research continues to progress on various fronts as to the potential refractive capabilities of crosslinking, including customized crosslinking for keratoconus to promote more visual improvement, LASIK combined with crosslinking to strengthen the cornea, and crosslinking for the correction of refractive error.

Customized crosslinking

Topography-guided crosslinking, zonal crosslinking, customized crosslinking: It goes by many names, but in short it describes the idea of targeting crosslinking treatment to the weakest area of the cornea.
William Dupps, MD, PhD, Cleveland Clinic, Cleveland, led studies that modeled this idea, building on his group’s research exploring the biomechanical origins of keratoconus.1
“The premise is there is a weakening that occurs in the cornea that progresses over time,” Dr. Dupps explained. Using clinical tomography of a patient’s eyes—one with keratoconus, the other without—Dr. Dupps said they created a virtual eye based on the real patient’s corneal shape. In the unaffected eye model, they weakened a point on the cornea and saw curvature increase.
“It developed a cone that was very similar to the one that had already manifested in the patient’s other eye,” he said.
From there, the next step was to study how to approach therapy. Dr. Dupps said they researched how different treatments changed stiffness and thus corneal shape. In the model of an eye with a more central cone and another with an off-center cone, they looked at the differences in response using the standard 9 mm global treatment vs. a more focal, or zonal, treatment centered on the steepest point.
“We found that in the few cases we simulated, it predicted we would produce more than two times the amount of flattening of the steepest portion of the cone compared to the standard treatment. Because this model incorporates the full 3-D shape of the cornea, we can pull out metrics that relate to the optical quality of vision. We computed aberrations like coma and spherical aberration and saw much larger decreases in coma with a zonal approach, which is the main aberration in keratoconus that we think degrades vision.
“The take-home was that we should continue to explore the potential of more customized treatments to not only stabilize the disease, but in some cases to achieve more optical rehabilitation than we could with a standard treatment,” Dr. Dupps said.
This research has moved from the theoretical/model domain into some clinical cases. Matthias Elling, MD, senior physician, Ruhr University Bochum, Bochum, Germany, pointed to research published in 2016 by Seiler et al. that found that 37% of eyes in the customized group (compared to 11% in the standard group) had 2 D or more of flattening at 1 year postop, as well as a better epithelial healing time, change in Kmax, and regularization.2 Another paper published earlier this year showed similar safety of standard crosslinking compared to topography-guided but with a stronger flattening effect and better improvement in corrected distance visual acuity in the latter group.3
When it comes to customized crosslinking, Dr. Dupps pointed to a couple of subtleties that he said could be important for this type of treatment. There are two different ways to calculate curvature of the cornea—axial curvature and tangential curvature—which he said with keratoconus could give two very different locations for the steepest part of the cone.
“We at least need to carefully report exactly what curvature metric is being used to guide [customized treatment] and then assess the outcomes in light of that,” Dr. Dupps said.
Dr. Dupps said there isn’t a tool yet that identifies the weakest point of the cornea—steepest curvature point acts as a surrogate for now—but he said emerging tools like Brillouin microscopy and OCT elastography are in translational trials and could provide this information in the future. Dr. Elling pointed to a 2014 study that used Brillouin microscopy to show that stromal weakening in keratoconus is concentrated in the region of the cone.4
The other factor brought up by Dr. Dupps is how long the customized treatment effect lasts. “Keratoconus is a progressive condition. If you initially have a small zone of weakening, but then 20 years later you have progressive weakening of untreated peripheral corneal regions, the shape may continue to change. This might explain some cases of progressive central corneal flattening after CXL, so there is a question of whether or not we should still be delivering some crosslinking treatment to the whole cornea while concentrating more treatment in the weakest area. That’s another important open question that needs further investigation,” he said.
“In terms of cons, a disadvantage [of customized crosslinking] that I can think of is that more physician input is required into determining the appropriate treatment pattern for each patient,” Dr. Elling said. “However, this may become easier over time as the ophthalmic community gains experience with the procedure.
“We have treated a few cases of irregular corneal astigmatism using this approach and have been very satisfied with the outcomes, particularly in terms of visual improvement,” Dr. Elling said.


It’s well established that LASIK can weaken the cornea and in some rare cases result in refractive drift or corneal ectasia. Crosslinking has been established to strengthen the corneal collagen bonds, so why not combine the two procedures? That’s the concept behind LASIK Xtra (Avedro, Waltham, Massachusetts).
“Published data on corneal crosslinking demonstrates that the procedure adds strength to the cornea, therefore the concept behind the combination procedure is an interesting one,” Dr. Elling said, adding that he does not have personal experience with it, but noted that “the published literature supports the concept that CXL can add safety to the LASIK procedure, and that there is likely an improvement in refractive stability in highly myopic or hyperopic patients.” He also mentioned other work that suggests better flap adhesion in these cases.5–7
Anders Behndig, MD, professor and chief physician, Department of Clinical Sciences, head of the Department of Ophthalmology, Umeå University, Umeå, Sweden, offered his perspective on LASIK Xtra, calling it a “promising method for preventing iatrogenic ectasia after LASIK.”
“It has a theoretical rationale, and data are coming in supporting its clinical value. Furthermore, it appears the method may have optical benefits, [such as] the stability of the postoperative refraction, but it’s a bit early to say much about that,” Dr. Behndig said.
Though he doesn’t have clinical experience with LASIK Xtra, Dr. Dupps said they’ve simulated it in computational models. The questions they hoped to answer were whether additional flattening of the cornea with crosslinking would cause over or under corrections that would require physicians to change their nomogram for LASIK and if it would have a protective effect against ectasia.
Dr. Dupps said there was less than a quarter diopter difference in the simulation between the LASIK and the LASIK Xtra model, suggesting that mild crosslinking effects wouldn’t have an impact on nomograms. Protection against ectasia was a harder question to answer because it’s a rare condition that would require long-term follow-up to assess the risk.
“In the modeling exercise, we saw that if you stiffen that bed it does give the cornea more resistance to shape change,” he said, adding later that the most compelling argument for LASIK plus crosslinking he’s seen is by Kanellopoulos et al. in hyperopic eyes.6 This study showed the crosslinking LASIK group had less hyperopic regression over the year follow-up and better refractive outcomes.

Correcting refractive error

Another avenue for crosslinking is correcting refractive error. Photorefractive intrastromal crosslinking (PiXL) is in clinical trials with Avedro’s Mosaic system to provide topography-guided accelerated crosslinking to correct low levels of refractive error.
“Our experience with PiXL for low myopia has been very positive,” Dr. Elling said. “We have been performing epithelium-off PiXL since 2014, and since last year, we began offering epi-on treatments with supplemental oxygen. Our patients are very satisfied with the vision improvement, and the idea of a minimally invasive procedure is very appealing. In our clinical studies, we had the possibility to show that we are able to reach a refractive change with PiXL.” Dr. Elling added that this technique hasn’t shown itself to be as precise as LASIK or PRK.8
Referencing a company-sponsored trial with 1-year follow-up, Dr. Elling noted statistically significant improvements in mean manifest sphere at all follow-up visits compared to baseline with a mean change of about 1 D. This reduction remained stable.
“Given its noninvasive nature, PiXL presents a potential treatment option for patients with low myopia who require a less invasive refractive correction procedure than conventional laser vision correction. In comparison to LASIK or PRK we have no tissue ablation. In the future, PiXL may give us the possibility … for refractive adjustments after cataract surgery, to treat irregular astigmatism, or even to do a touch-up after LASIK with thin residual stromal bed,” Dr. Elling said, adding that they are currently evaluating techniques to balance adequate epithelial disruption and postop comfort.
Dr. Behndig said results with PiXL have shown stable and “sufficiently consistent” results.
Going forward, both Drs. Behndig and Elling mentioned the possibility of PiXL for presbyopic correction. Dr. Dupps agreed that early trials of refractive crosslinking are encouraging, and he thinks that enhancing the predictability of refractive crosslinking is the next critical step in moving the concept forward.
“By building on advances in crosslinking science to enhance and more tightly control the stiffening effect, other personalization tools such as corneal tomography, emerging biomechanical measurement technologies, and simulation-based planning tools can help leverage our predictive capabilities toward more personalized treatments,” he said.


1. Roy AS, et al. Patient-specific computational modeling of keratoconus progression and differential responses to collagen cross-linking. Invest Ophthalmol Vis Sci. 2011;52:9174–87.
2. Seiler TG, et al. Customized corneal cross-linking: One-year results. Am J Ophthalmol. 2016;166:14–21.
3. Cassagne M, et al. Customized topography-guided corneal collagen cross-linking for keratoconus. J Refract Surg. 2017;33:290–97.
4. Scarcelli G, et al. Biomechanical characterization of keratoconus corneas ex vivo with Brillouin microscopy. Invest Ophthalmol Vis Sci. 2014;55:4490–5.
5. Tomita M. Combined laser in-situ keratomileusis and accelerated corneal crosslinking: an update. Curr Opin Ophthalmol. 2016;27:304–10.
6. Kanellopoulos AJ, et al. Topography-guided hyperopic LASIK with and without high irradiance collagen cross-linking: initial comparative clinical findings in a contralateral eye study of 34 consecutive patients. J Refract Surg. 2012;28:S837–40.
7. Seiler TG, et al. Superficial corneal crosslinking during laser in situ keratomileusis. J Cataract Refract Surg. 2015;41:2165–70.
8. Elling M, et al. Photorefractive intrastromal corneal crosslinking for the treatment of myopic refractive errors: Six-month interim findings. J Cataract Refract Surg. 2017;43:789–795.

Editors’ note: Dr. Dupps has financial interests with Avedro, OptoQuest (Cleveland), and Carl Zeiss Meditec (Jena, Germany). Dr. Elling has financial interests with Avedro. Avedro supports Dr. Behndig’s ongoing study on PiXL with an unrestricted grant.

Contact information


Refractive opportunities for crosslinking Refractive opportunities for crosslinking
Ophthalmology News - EyeWorld Magazine
283 110
220 155
True, 9