Emerging strategies yield improvements in keratoconus management
Recent research that looks at keratoconus from different and sometimes unusual angles might help penetrate the “mystery” of the condition and pave the way toward new management strategies.
“New discoveries are reversing a lot of the old interpretations,” David Touboul, MD, PhD, said. “We are nearer to understanding the pathophysiology of the disease, developing new screening approaches, new interpretations of the cross-linking effects, and there is a lot of discussion on combined procedures. Things are moving on.”
Keratoconus is a condition whose origin has long been unknown and hotly debated. Genetic and environmental factors have been widely investigated, as well as, more recently, the inflammatory component.
Some studies have suggested that keratoconus has great clinical variability and may be linked to multiple chromosomal regions, consistent with polygenic mode of inheritance. Candidate genes, such as VSX1, DOCK9 and TGFB1, were identified as potentially involved in the pathogenesis of the disease, but no validation was achieved in larger numbers.
“This suggests that there is a large genetic heterogeneity in keratoconus,” Damien Gatinel, MD, PhD, said. “I personally do not believe that one specific gene mutation will ever be discovered. Familial forms of the disease only occur in less than 20% of cases, and common environmental factors may be better incriminated to account for familial penetrance. In addition, some studies have also reported discrepancies in twins.”
Fresh look at causative factors
Keratoconus may recur in families, but shared environmental factors and heritable traits, such as orbital, eyeball and corneal morphology, are likely to play a greater role than gene mutation, according to Touboul.
“Corneal topography shows that there are genetically inherited patterns. Corneal shape, orbital volume, eyelid positions and muscle insertion, which are part of face morphology, are passed on across generations. All these components affect the biomechanics of the eye, which is correlated with the stress and strain around the cornea. Muscle tension, the position of the eyeball in the orbit and the stress of eye movements may play a role in what we may call the genetics of keratoconus,” he said.
In addition, allergies, atopy and myopia, which have been found in several studies to be correlated with keratoconus, have both a genetic and an environmental component.
“We should not forget that families share the same environment,” Touboul said. “Thus, when screening inherited risk of keratoconus in families of keratoconus patients, based on corneal topography analysis, we usually find some similarities in terms of corneal thickness, keratometry and shape, which are not keratoconus cases but family-inherited biometrical patterns. I think we should define eye rubbing risk score including an eye rubbing questionnaire and corneal surface examination that will define the eye rubbing risk for each patient. Since I think that keratoconus is probably a behavioral disorder of young patients, we are currently working on keratoconus families (AFRIK study) and on an eye rubbing questionnaire in Bordeaux.”
A spurious variable explains the correlation of allergy and keratoconus, according to Gatinel, and this variable is excessive eye rubbing.
“Allergic conjunctivitis patients begin to rub their eyes early in life and do it quite frequently and vigorously. Over time, the detrimental impact of eye rubbing on the cornea is expressed in two ways. First, the mechanical trauma conveyed directly onto the corneal dome [through increased eyelid pressure] results in the progressive disorganization of the stromal harmonious structure and causes progressive focal stromal weakening. Second, eye rubbing triggers local inflammation and activation and/or accentuation of local inflammatory biomechanical cascades, which can also cause tissue impairment. Studies have shown that 30 seconds of consecutive eye rubbing in normal, healthy patients results in an increase in proteolytic enzymes in the tear film,” he said.
Not allergy in and of itself, but instead a behavior that is classically associated with allergy, may therefore be the cause of corneal ectasia.
“The role of eye rubbing in the pathogenesis of the disease has been underestimated. I believe that it is in fact the most important and indispensable ingredient in the keratoconus recipe, the root cause of keratoconus,” Gatinel said.
Gatinel’s personal experience and recent prospective work conducted within the Rothschild Foundation led to the discovery that, in keratoconus, collagen is not actually lost but simply redistributed within the cornea by slippage between the lamellae and possibly via external forces of rubbing.
“When interviewed thoroughly and assisted by clear explanations about the possibility of unconscious rubbing, all of our patients diagnosed with and followed for [keratoconus] were found to have, in fact, rubbed one or both of their eyes,” he said.
He also found that the eye in which keratoconus is more advanced is always the one that is rubbed more often and that a striking correlation exists with the side on which patients preferentially sleep. Typically, patients with keratoconus tend to sleep in the prone position or on one side, giving rise to prolonged nocturnal ocular compression. This may cause local inflammation via heating and contamination from dust mites in the bed linen, which triggers rubbing to calm the itch of that eye.
“Strictly unilateral keratoconus patients also do exist, and they are found to just rub one eye excessively. What occurred to me over time is that it takes about 2 to 3 years of repeated rubbing to develop the first manifestations of keratoconus. Patients with newly discovered keratoconus often disclose that they started to rub their eyes greatly when they developed sudden ocular allergy or were exposed to dust or other irritants,” he said.
Other triggers for increased eye rubbing relate to chronic general fatigue. Working the night shift or investing heavily in studies with extended computer time is frequently disclosed and seems to be concomitant to the practice of repeated eye rubbing.
“It is very important to spread awareness about eye rubbing as a risk factor. Today, with the increasing use of computers, tablets and smartphones, dry eye disease is dramatically increasing, and therefore eye rubbing is increasing. As well as more and more myopia, we are beginning to see more and more forme fruste keratoconus, and our impression is that we should already abandon the classification of keratoconus as a rare disease,” Touboul said.
Both specialists agree that educating patients not to rub their eyes might prevent keratoconus in some cases and, at least in the early and moderate forms, arrest disease progression.
“Over an average 2-year follow-up, adult patients who completely stop rubbing their eyes have also stopped progressing. Complete cessation of eye rubbing has resulted in stabilization of the corneal deformation. This is a very exciting finding. Interestingly, progression is only seen in patients who admit to be unable to abandon their eye rubbing habit. This is strong indirect evidence that eye rubbing is the root cause of this disease,” Gatinel said.
Touboul also hypothesized that cessation of eye rubbing might explain the efficacy of corneal cross-linking.
“Photopolymerization not only induces corneal stiffening, it also inhibits the sensitivity of corneal nerves for months and perhaps for years, and this means that patients stop rubbing their eyes. In addition, after CXL, the epithelium is renewed, the surface is smoother, and you treat the inflammation with steroids and lubricants. This also makes patients feel better and stop rubbing their eyes,” Touboul said.
CXL, in other words, might work because of the synergistic effects of multiple factors, some of which converge toward eye rubbing inhibition, “which is probably the main mechanism to explain the efficacy of the procedure. Thus, innovations in the field of CXL should focus on eye rubbing inhibition than corneal stiffening, in my point of view,” he said.
“After decades of teaching ophthalmic students that keratoconus is a dystrophy of unknown origin, it is difficult to accept the simple truth that it all comes down to eye rubbing. And it is not possible, for ethical reasons, to conduct a prospective study where patients are asked to rub their eyes vigorously and repeatedly to determine in which conditions keratoconus could be induced. However, when there are large number of patients, this is surprisingly easy to prove,” Gatinel said.
Impact on quality of life
Studies have shown that keratoconus has a heavy impact on quality of life. The Collaborative Longitudinal Evaluation of Keratoconus (CLEK) study group published the finding that the vision-related quality of life of keratoconus patients, with a mean age of 29 years, was similar to that of much older patients with stage 3 and 4 age-related macular degeneration.
“It was a big surprise for us at the time to see the profound effect of this disease,” Karla Zadnik, OD, PhD, said. “Many of the patients in the CLEK study were correctable with spectacles or contact lenses to 20/20 or 20/30, so we were shocked to see that they felt as bad about their vision as much older patients with a potentially blinding disease.”
Data were collected at 16 centers across the U.S., totaling 1,209 patients who were followed annually for 8 years.
“We thought we had made some big mistake at first, but this was reality. The data were collected from a large sample,” Zadnik said. “These data have been on my mind since then. These data should make doctors more empathetic when patients with keratoconus say they have been missing days at work or that, at the end of the day, they can do nothing but go to bed. We should all listen more carefully to our patients with keratoconus and stop thinking that they can just put on their contact lenses, feel fine and should not complain.”
A similar, more recent study, the Impact study, found that the quality-of-life scores for keratoconus patients in France were even below CLEK outcomes, despite the use of new procedures such as CXL.
“We enrolled keratoconus patients with different stages of the disease, including those who had had procedures like CXL, intracorneal ring segments (ICRS) or corneal transplantation. A total of 550 keratoconus patients were analyzed in 57 participating centers across France,” Valentine Saunier, MD, MSc, said.
Approximately 80% of patients had visual acuity more than 20/40, but there was a significant gap between visual acuity and subjectively reported quality of life, and this included patients after CXL, ICRS implantation and keratoplasty.
“Visual acuity is just a number, which does not reflect the way patients see. Other factors must be taken into account, like spherical aberration, coma and trefoil, which are significant in keratoconus and have a heavy impact on quality of vision,” Saunier said.
In her opinion, these findings point to the need for a new classification of keratoconus that includes parameters such as optical aberration and the psychological state of patients.
Customized cross-linking to treat keratoconus holds promise and might soon become the gold standard, according to Theo G. Seiler, MD.
This new approach is based on the findings of several research groups working on theoretical corneal biomechanics and the concept that focal weakening, rather than uniform weakening of the cornea, occurs in keratoconus. Customized CXL creates an individual profile wherein maximum fluence is focused on the weak area and is progressively tapered toward the peripheral, nonectatic region.
“There are several aspects we still have to define,” Seiler said. “Right now we create the customized profile according to geometric data obtained from Placido-based topography, rotating Scheimpflug imaging or OCT. In the future, Brillouin spectroscopy should allow us also to detect localized biomechanical properties of the cornea and plan the treatment accordingly with micrometric precision.”
In the first preliminary studies, Seiler and his group found that the weakest point of the cornea is highly correlated to the point of maximum posterior float; based on this point, they have designed an irradiation pattern algorithm. Other groups, such as that of Anders Behndig, MD, in Sweden, centered on the maximum keratometry (Kmax).
“This may also be imprecise because the epithelium is thinner above the cone and thicker in the midperiphery. However, centration is similar in many cases with the two methods, and the idea of tapering toward the periphery is the same. What is still to be investigated is the exposure limit and total energy load that is safe for the eye. In Zurich, we use a maximum energy of 10 J/cm2 in the center, whereas other groups use 15 J/cm2. Safety standards will have to be defined,” Seiler said.
Irradiance and imbibition time are also under investigation. Based on experimental data, the best irradiance might be between 9 mW/cm2 and 18 mW/cm2, according to Seiler, and imbibition of 0.1% riboflavin with hydroxypropyl methylcellulose for 10 minutes is currently used.
“Since we have evaluated the outcome of the first prospective study on customized CXL in Zurich in 2015, every patient is routinely treated with customized CXL (Mosaic system, Avedro),” Seiler said.
An important advantage of the customized procedure is that a smaller portion of the epithelium is removed, thus reducing healing time and the risk for infection. In the studies published by Seiler and colleagues, epithelial healing time was 2.6 days in the customized CXL group compared with 3.2 days in the standard CXL group, with a debridement area of 9 mm.
“This is a difference of 0.6 days, which is quite a significant gain if you have in mind that the vulnerable phase is exactly that phase where the epithelium is open because there you don’t have a barrier against infections,” Seiler said.
In terms of visual gain, flattening of more than 1 D was achieved after 1 year in 40% of eyes in the standard group and 60% of eyes in the customized group, further increasing up to 3 years after treatment. With a higher Kmax decrease, the advantages of the customized procedure were even more apparent. A flattening of 3 D was achieved in double the number of eyes in the customized group compared with the standard group.
Subgroup analysis showed that the real benefit of the customized procedure in terms of visual acuity is in corneas with a Kmax up to 50 D before irradiation.
“Patients with a Kmax of 60 D or 65 D would not see a difference in [uncorrected distance visual acuity], and even with a flattening of 6 D, 7 D or 8 D, they would have to use rigid contact lenses. Using the customized rather than the standard procedure would not make a real difference. But patients with preoperative Kmax of 50 D and less who are flattened by 3 D or 4 D may recognize a significant increase in UDVA. This is something I want to emphasize, that some patients benefit a lot from the customized procedure,” Seiler said.
Corvis Biomechanical Index
Topography, tomography and, more recently, OCT have been the gold standard for detecting the morphological alterations that lead to the diagnosis of keratoconus.
“These are good parameters because keratoconus does change the corneal shape. However, what lies below the surface is a change in corneal biomechanics, and this is the main focus of research now,” Riccardo Vinciguerra, MD, said.
The Ocular Response Analyzer (ORA, Reichert) was the first instrument to detect indexes that are correlated with corneal biomechanics.
“The limitation of ORA is that it relies on a signal deriving only indirectly from the deformation of the cornea but is not able to visualize the deformation itself,” Vinciguerra said.
The Corvis ST (Oculus), on the other hand, was for a long time only able to show the deformation on a video that nobody yet knew how to use in practice.
“However, there were a lot of promising parameters there. We started working on them with a joint effort that involved researchers from Europe, the U.S., South America and, now, also Korea. Through successive steps, we came to develop the Corvis Biomechanical Index (CBI) that won us the 2017 Troutman Prize at the AAO,” Vinciguerra said.
The CBI includes several dynamic corneal response parameters and has shown high sensitivity and specificity in discriminating keratoconus from healthy corneas. It is also useful to detect subclinical keratoconus and very asymmetric ectasia. Another interesting field of application of corneal biomechanics is CXL.
“Currently, all we have in our hands to evaluate CXL effects is either the demarcation line or the change in shape, which are not direct ways of evaluating the real effect of CXL, which is the stiffening effect,” Vinciguerra said.
Vinciguerra now uses Corvis in addition to Pentacam (Oculus) as a fundamental part of his keratoconus screening because the CBI is increasing his chances to detect the disease early. He also uses the Corvis after CXL but is currently working with his research group and Oculus to improve this application.
“We are trying to create two new displays, one to evaluate stiffness changes after cross-linking and the other to detect and monitor keratoconus progression. Here in Liverpool, with the team of Prof. Elsheikh, we are working on a new index called BETA, which will be directly evaluating biomechanics with no correlation to thickness or IOP,” Vinciguerra said.
The next step might be biomechanical modeling of different surgeries to predict the effects on individual corneas and select patients for specific procedures.
“Biomechanical modeling of refractive surgery will allow us, for instance, to decide whether to do surface ablation, LASIK or SMILE. In patients with keratoconus, modeling will help us decide where to implant intracorneal rings, how to cross-link the patient and how to design the customized cross-linking pattern based on corneal biomechanics rather than morphological indexes,” Vinciguerra said.
Brillouin spectroscopy is another promising method to measure biomechanics in vivo, with the benefits of a noncontact mode and a high spatial resolution of 5 µm.
“With air puff-based systems you obtain one number for the entire cornea, while Brillouin spectroscopy maps the tissue response with thousands of numbers, leading to a high-resolution map of the bulk elastic modulus of the cornea,” Seiler said.
The third benefit is that no IOP changes and no other ocular structures are involved. With the Corvis, the air puff produces deformation of the cornea by applying pressure, but the response is influenced by the biomechanical properties of the sclera and by IOP.
“In the air puff deformation response we have to consider the interaction of many factors, whereas with Brillouin spectroscopy the bulk elastic modulus is directly determined with a high spatial resolution. Compared with the Corvis, Brillouin spectroscopy may detect keratoconus at earlier stages and will overcome the limitations of geography-based tomographic methods, which so often lead us to wonder, ‘Is this a keratoconus cornea or not?’” Seiler said.
A paradigm shift in ICRS use
ICRS use is also undergoing a paradigm shift, according to Touboul. In the past, segments were used in patients with progressive keratoconus to avoid or postpone corneal transplantation but without targeting an accurate refractive correction.
“Now patients are referred earlier because we can offer CXL, and ICRSs have become a minimally invasive, tissue-sparing option to flatten and regularize the surface and improve VA and refraction by reducing coma aberration. Classically, we wait 2 to 3 months after the procedure, see if there are the conditions for contact lens fitting or enhance the results with transepithelial topo-guided PTK,” Touboul said.
“Although in terms of progression we are currently relying to a large extent on CXL as the answer, CXL does not specifically address vision. ICRSs stabilize the cornea and change the shape, making it easier to wear glasses and fit a contact lens, even a soft lens, which is generally easier to wear and more comfortable. It is the least invasive procedure, particularly now that we can make the tunnel with the femtosecond laser. I also like it because it is easily reversible,” Penny A. Asbell, MD, FACS, MBA, said.
The question today is whether, how and in what sequence ICRSs should be combined with CXL. There are several specialists who perform simultaneous CXL and ICRS implantation and some who perform them sequentially in various orders and time intervals.
“We don’t have comparative data so we really don’t know what the best option is. If there is a risk of progression, I would personally do CXL first and then take at least a couple of months, preferably 6 months, to evaluate whether there are the conditions and the indications to implant the rings. On the other hand, in cases where you cannot document progression or if there is significant keratoconus and you need to do something a little faster, I think ICRSs are the way to go as the first step,” Asbell said.
A combination of CXL with an excimer laser procedure such as PRK is not approved in the United States and will remain controversial until larger studies can definitively assess risks and benefits, she said.
“Excimer laser procedures take away tissue and therefore make the cornea even less stable. It’s a double-edged sword because on the other hand you might be able to correct a good part of the refractive error and make life easier and better for the patient,” she said.
In her opinion, these options are worth further careful investigation. “First because refractive surgery is difficult and more risky on the very irregular cornea of keratoconus eyes, and second because we cannot be sure of the long-term effects of the treatment on corneas that may still undergo progressive thinning,” Asbell said.
“We need to outline a strategy for combining procedures, where each subsequent step is decided on the basis of regular outcome monitoring and the vision improvement we have planned might take time to be achieved,” Touboul said.
After PTK on ICRS implantation, steroids should be given for at least 3 months to prevent haze. When it occurs, haze might spread from the location around the ring toward the center, leading to blurred vision despite having nearly regular topography. Optimized visual recovery might take 6 to 12 months in these cases, delaying the result of the second procedure.
“Even in the best case, I tell patients that from the day we implant the ring, wait, do PTK and then wait for PTK to work at its best; about 1 year will pass. The time is eventually approximately the same that you need for full recovery after corneal transplantation, but the difference is that in one case you have a graft, in the other you have the native cornea of the patient,” Touboul said.
On a final note, he said that the latest advances have reduced the number of transplantation procedures in patients with keratoconus.
“The toll we pay is that now our DALK procedures are really difficult, challenging cases with scars, thinning and huge ectasias because CXL, rings, PTK and improved contact lens fitting have driven away all the other cases from the operating table, and it is good news,” Touboul said. – by Michela Cimberle
- Ambrósio R Jr., et al. J Refract Surg. 2017;doi:10.3928/1081597X-20170426-02.
- Ambrósio R Jr., et al. Open Ophthalmol J. 2017;doi:10.2174/1874364101711010176.
- Asbell PA. Eye Contact Lens. 2014;doi:10.1097/ICL.0000000000000097.
- Bral N, et al. Case Rep Ophthalmol. 2017;doi:10.1159/000484712.
- Galvis V, et al. Cornea. 2017;doi:10.1097/ICO.0000000000001052.
- Galvis V, et al. Eye (Lond). 2015;doi:10.1038/eye.2015.63.
- Kymes SM, et al. Am J Ophthalmol. 2008;doi:10.1016/j.ajo.2007.11.017.
- Mercer RN, et al. J Refract Surg. 2017;doi:10.3928/1081597X-20170621-03.
- Saad A, et al. J Refract Surg. 2016;doi:10.3928/1081597X-20160523-01.
- Saunier V, et al. J Cataract Refract Surg. 2017;doi:10.1016/j.jcrs.2017.08.024.
- Scarcelli G, et al. Invest Ophthalmol Vis Sci. 2013;doi:10.1167/iovs.12-11387.
- Scarcelli G, et al. Invest Ophthalmol Vis Sci. 2014;doi:10.1167/iovs.14-14450.
- Seiler TG, et al. Am J Ophthalmol. 2016;doi:10.1016/j.ajo.2016.02.029.
- Seiler TG, et al. J Refract Surg. 2017;doi:10.3928/1081597X-20170705-01.
- Shetty R, et al. Indian J Ophthalmol. 2017;doi:10.4103/ijo.IJO_217_17.
- Smadja D, et al. J Refract Surg. 2013;doi:10.3928/1081597X-20130318-07.
- Tellouck J, et al. Cornea. 2016;doi:10.1097/ICO.0000000000000833.
- Vinciguerra R, et al. J Refract Surg. 2017;doi:10.3928/1081597X-20170213-01.
- Vinciguerra R, et al. J Refract Surg. 2018;doi:10.3928/1081597X-20180103-01.
- For more information:
- Penny A. Asbell, MD, FACS, MBA, can be reached at Mount Sinai Hospital, Box 1183, New York, NY 10029; email: firstname.lastname@example.org.
- Damien Gatinel, MD, PhD, can be reached at Fondation Rothschild, 25 Rue Manin, 75019 Paris, France; email: email@example.com.
- Valentine Saunier, MD, MSc, can be reached at Group Hospitalier Pellegrin, Place Amélie Raba Léon, 33000 Bordeaux, France; email: firstname.lastname@example.org.
- Theo G. Seiler, MD, can be reached at Institut für Refraktive und Ophthalmo-Chirurgie (IROC), Stockerstrasse 37, 8002 Zürich, Switzerland; email: email@example.com.
- David Touboul, MD, PhD, can be reached at Group Hospitalier Pellegrin, Place Amélie Raba Léon, 33000 Bordeaux, France; email: firstname.lastname@example.org.
- Riccardo Vinciguerra, MD, can be reached at Royal Liverpool and Broadgreen University Hospital, Prescot St., Liverpool, England L7 8XP; email: email@example.com.
- Karla Zadnik, OD, PhD, can be reached at The Ohio State University College of Optometry, A400 Starling Loving Hall, 338 W. 10th Ave., Columbus, OH 43210; email: firstname.lastname@example.org.
Disclosures: Asbell, Gatinel, Saunier, Seiler, Touboul and Zadnik report no relevant financial disclosures. Vinciguerra reports he is consultant to Oculus.
Click here to read the , "Which combination of procedures would be your best choice to provide vision improvement in patients with progressive keratoconus?"