Although keratoconus remains a mystery in many ways, we have many treatment options for successful management.



Although keratoconus remains a mystery in many ways, we have many treatment options for successful management.


Keratoconus is a peculiar and fascinating corneal disease in which the cornea weakens biomechanically, sheds tissue, and scars. We have known about this pathology for well over 150 years; it was first described in the early literature by Nottingham in 1854. Still, we do not know the etiology, nor do we know where keratoconus starts. However, recently a procedure has been developed that can arrest progression of the disease. Furthermore, we have many more options today in the refractive management of our keratoconus patients.

The intent of this article is to review what has happened in the field of keratoconus in recent years, including the work we have conducted at the Texas Eye Research & Technology Center (TERTC) at the University of Houston College of Optometry.


In keratoconus, the visual compromise results from ectasia of the cornea, a forward protrusion of the tissue. The structural pathological change leading to this phenomenon is still under debate. We know that the thinning of the cornea, concomitant with the ectasia, cannot on its own be the explanation for the anterior corneal collapse. In laser-assisted in situ keratomileusis (LASIK) surgery, it is well known that the healthy corneal stroma may be thinned down to 250µm without ectasia occurring, but many keratoconus patients suffer ectasia with a stromal thickness greater than 400µm. Therefore, there must be another factor that weakens the cornea and biomechanically provokes the ectasia, which is in an anterior direction most likely due to the intraocular pressure.

To understand the pathophysiology of keratoconus-associated ectasia, we at TERTC realized that our understanding of the stromal lamellar architecture was too incomplete. This anatomical research is still ongoing and is indeed fascinating. One question relating to corneal surgery and keratoconus that we are studying is: What makes the cornea thicker in the periphery? We are now monitoring keratoconus patients for progression, and we see the thickness difference between the central and the peripheral cornea. How is the cornea accomplishing this thickness variation? Is it affected by keratoconus, and does it contain secrets about which corneal surgeons need to know?

Recognizing that the cornea becomes progressively thinner in keratoconic corneas, we were interested in learning what is lost, because this tissue loss had yet to be detailed in the literature. The basic building block in the stroma is the collagenous lamellae, which consist of bundles of parallel collagen fibers—Type 1 collagen—and these bundles, or lamellae, are parallel to the ocular surface (Bergmanson, 2016).

The first question to answer was: How many lamellae should a normal cornea have? We have now learned that the human cornea contains 234 to 251 lamellae centrally (Bergmanson et al, 2005; Mathew et al, 2015). This new data shows that there is little individual variation in the number of lamellae that forms the central cornea.

In studying keratoconus, we were surprised to find that these pathological corneas had as much as 56% more lamellae anteriorly. How was this possible in a thinner cornea? Precise measurements and plain observations revealed that the lamellae in keratoconic corneas break up into smaller units. We have termed this pathological phenomenon lamellar splitting (Mathew et al, 2015). Similar tissue degradation has occurred in post-refractive corneal surgery in which the cornea becomes ectatic, and this phenomenon, lamellar splitting, may be seen (Dawson et al, 2008). Post-refractive surgery ectasia can occur following both photorefractive keratectomy (PRK) and LASIK and may occur within days, weeks, or several years following the surgery. When ectasia occurs soon following the surgery, lamellar splitting has already taken place in that cornea. However, when development of ectasia occurs several years after surgery, the lamellar splitting most likely had not occurred at the time of the surgery. We believe that ectasia develops when both lamellar splitting and corneal thinning have occurred.

Another hypothesis for post-refractive surgery- and keratoconus-induced ectasia was presented by Dawson et al (2008), who proposed lamellar slippage, in which corneal lamellae lose their inter-lamellar anchorage and start sliding over each other, creating the thinned cone area. However, this does not explain the overall thinning of the cornea that occurs in keratoconus (see below). If slippage occurs without tissue loss, the cornea can indeed become thinner in one area; but, by definition, it must become thicker in another area that, in this case, would be the peripheral cornea.


Our histopathological research documented that, in addition to lamellar splitting, we had indeed lost predominantly anterior lamellae. In addition, over large areas, the anterior limiting lamina was either thinned or completely absent.

Interestingly, the epithelium is involved in this pathological process, which was first reported more than 50 years ago (Teng, 1963). Many of the epithelial cells have abnormal cytoplasm and adopt an abnormal behavior, while also overproducing an incomplete basement membrane material (Mathew et al, 2011). Abnormal cells and a thickened, defective basement membrane laid down on a partially or completely removed anterior limiting lamina is a recipe for epithelial fragility, which often clinically manifests in the keratoconic eye in the form of epithelial staining.

We noticed histopathologically that all of these epithelial and stromal changes also occur outside of or peripheral to the cone, which is demarcated by the Fleischer ring, suggesting that this disease is not limited to the cone area. Clinical research conducted at TERTC confirmed the hypothesis that keratoconus is a pan-corneal disease. This clinical study demonstrated that keratoconus patients, when compared to a control group, have thinner corneas across their full width (Brautaset et al, 2013). The thickness difference was greater centrally, but diminished toward the corneal periphery.

In clinical practice, we should always keep in mind that the entire cornea is compromised in this disease; for this reason, we feel that a scleral contact lens vaulting over the cornea and limbus makes good clinical sense. In clearing the cornea and limbus, the scleral lens places less physical stress on the weakened epithelium, with a consequent reduced incidence of staining.

Now that we know that keratoconus involves the entire cornea, we may wish to take another look at keratoconus “recurrence” following transplant surgery. A decade or two after receiving a transplant, the recipient may develop keratoconus in the donor tissue. The old thinking used to be that the former keratoconus patient was simply unlucky and received someone else’s subclinical keratoconus via the transplant.

We have two patients who underwent bilateral penetrating keratoplasty (PKP)—left and right eyes were operated on three to six months apart—25 years ago (Figure 1). The time separation between left and right eyes was sufficiently great to rule out the possibility that the donor corneas were harvested from the same individual.

Figure 1. A 53-year-old patient with a history of PKP OD 25 years ago. Topography indicates marked corneal steepening inferiorly with accompanying thinning consistent with keratoconus (A). Recently, the patient required a new transplant due to bullous keratopathy from low endothelial cell count. Histopathological assessment of this patient’s donor cornea removed for re-transplantation, in full thickness corneal keratoplasty surgery. The pathological features below are consistent with keratoconus (B):
E: the epithelium shows reacting and distorted cells.
Asterisk: Anterior limiting lamina has been lost in this view.
Arrow: Keratocyte process less than 1µm from the thickened basement membrane of the epithelium.
S: Stroma shows thin stromal collagenous lamellae, as is also consistent with the lamellar splitting that leads to the biomechanical weakening of the cornea in ectasia.

So what is the likelihood that a patient is so unfortunate that he or she received corneas from two separate subclinical keratoconus donors? The incidence of keratoconus is believed to be 1 in 2,000 (Rabinowitz, 2005), and therefore, it could be argued that there is a probability of a patient obtaining someone else’s keratoconus transplanted to his or her eye. Now, the likelihood for this to occur to both eyes makes it mathematically one chance in 4 million, which makes this a very unlikely scenario. Given that eye banks do screen for keratoconus, we are, as a consequence, concerned with the rarer subclinical keratoconic cadaver eye. The chance of accidentally receiving two such corneas is far less than one in 4 million—perhaps it is less than one in 40 million.

Acknowledging that the entire cornea is involved in keratoconus and that, in transplant surgery, a button measuring approximately 7mm is removed, we still have host cornea left to which the donor cornea is attached using stitches. It follows that the surgery did not eliminate the disease—keratoconus—from the eye. Therefore, the most likely explanation for recurrence of keratoconus is that the disease migrates from the host into the donor tissue (Bergmanson et al, 2014). This is also how disease “recurrence” is believed to occur in corneal dystrophies. It may be argued that this truly is not a recurrence, but rather a re-emergence of the disease.

In counseling patients facing possible transplant surgery, we can discuss the possibility of the disease re-appearing in the donor tissue. In an analysis of cases diagnosed to have a re-emergence of keratoconus, it was reported that for PKP, the average latency period was 19 years; for lamellar (LKP) or deep anterior lamellar keratoplasty (DALK), the latency period was only three to four years. The faster re-emergence time in LKP is due to the fact that more of the keratoconic tissue is left untouched by the surgery (Bergmanson et al, 2014). However, this number is also based on fewer reported cases and, for this reason, is less reliable. Nevertheless, the latency period is likely to be shorter for LKP than for PKP.


We have a very central role in the counseling and caring of our patients who have keratoconus. As the disease progresses, our keratoconus patients will have developed a strong faith in our abilities and competence. As a consequence, they will listen very carefully to what we have to say regarding continued and future management. In other words, as eyecare practitioners, we have a great opportunity to guide them in an appropriate direction that meets their individual, social, and financial needs. In the past, options for managing keratoconus were few and simplistic; however, today keratoconus patients have an array of possible clinical management solutions.

Corneal Stromal Cross-Linking Eyecare professionals can and should be monitoring keratoconus for progression. Whether this is best performed with keratometry/topography or with pachymetry or another methodology will most likely be determined through future peer-reviewed literature. Such monitoring is necessary to determine whether cross-linking is indicated to stop further advancement of the disease. When applied properly, this is a wonderful option for patients.

The corneal stromal cross-linking (CXL) procedure originated in Dresden, Germany and has been applied across Europe for more than a decade, with promising outcomes when the treatment protocol is strictly followed (Wollensak et al, 2003). This year, the U.S. Food and Drug Administration (FDA) approved this procedure, which involves debriding the cornea and then exposing the full width of the cornea to a 365nm wavelength of ultraviolet-A (UVA) radiation while the corneal tissue is soaked with a riboflavin/dextran solution, to achieve a collagen fiber cross-linking and an ensuing increase in biomechanical rigidity. The FDA-approved method is the Avedro KXL System, using Photrexa Viscous as the riboflavin/dextran solution (Photrexa does not contain dextran) and delivering a dose of 3mW/cm2 for 30 minutes. Anything else must be considered off-label. For instance, in clinics in which an “accelerated” procedure that reduces exposure time to five minutes is applied, the energy of the beam must be increased a massive six times. The safety of such irradiation has not yet been demonstrated. In addition, not debriding the epithelium prior to exposure is an off-label application.

The fact that riboflavin is not opaque at the 365nm wavelength should be acknowledged—approximately 25% of the UVA radiation is transmitted. This transmitted wavelength is absorbed by the crystalline lens (Walsh and Bergmanson, 2016) and may also harm the corneal endothelium, especially if the cornea is thinner than 400µm. Because riboflavin is not opaque at that wavelength, another compound would be more ideal.

The FDA-approved protocol permits corneas thinner than 400µm to be exposed if the topical application of the riboflavin solution swells the cornea to 400µm and beyond. However, is the safety of this artificial swelling of the cornea based on the assumption that this solution has the same absorption coefficient as the corneal stromal tissue, which has not been demonstrated and is not likely to be the case? It should be noted that the FDA approval of this procedure is indicated for not only keratoconus, but also for post-refractive surgery ectasia. This FDA approval can be applied only to individuals between the ages of 14 and 65, because the FDA trials did not have patients outside of this range.

A curious aspect about cross-linking is the question of from where this name is actually derived and its validity as a term for this procedure. There are no images in the literature illustrating this cross-linking that is assumed to have occurred. Indeed, the question has been raised as to whether this is an appropriate term to describe the events provoked by UVA exposure of the riboflavin-saturated cornea (Meek and Hayes, 2013). What occurs in the stroma is not fully understood and is more complex than just simply fibrillar matter linking separate collagen fibers to produce enhanced biomechanical characteristics. Meek and Hayes (2013) reported that cross-linking may also occur between fibrillar surface molecules and proteoglycan proteins in the matrix.

Studies and reviews universally report favorable outcomes of the CXL procedure (Meek and Hayes, 2013; Raiskup and Spoerl, 2013; O’Brart et al, 2013). This consensus tells us that the UVA/riboflavin Dresden protocol is effective at stopping, or at least slowing down, progression of corneal ectasia. It is important to recognize that this treatment protocol is not a cure and is only indicated if disease progression is established. How this progression is best established will be the next step in managing ectasia patients. We look forward to benefiting from evidence-based information on how to best determine the advancement of the pathology.

Although CXL appears to consistently achieve an arrest or slowing down of progression, there are various complications associated with this protocol. The most common complications listed under the FDA approval include corneal opacity (haze), punctate keratitis, striae, pain, and visual acuity loss. In our clinical experience, haze is the most common and can be discerned in most corneas exposed to this procedure (Figure 2). Therefore, it may be argued that the UVA/riboflavin treatment protocol should be applied in a judicious manner and only on corneas in which ectasia has been demonstrated to be in a progressive phase.

Figure 2. Keratoconus patient five months after corneal cross-linking treatment. The arrow shows Vogt’s striae, the asterisk shows anterior stromal haze consistent with the cross-linking procedure.

Contact Lens Options On the contact lens side, we have a large number of options that include soft toric contact lenses, custom soft lenses, corneal GP lenses, piggyback lens systems, hybrid contact lenses, and scleral GP lenses. For the mild, almost subclinical keratoconus cases, a soft toric lens would make sense. Such a lens is cost effective and easily replaced if lost. Soft lenses can now be ordered to correct greater astigmatic errors, and these lenses produce rather consistent results.

We also have specialty lenses manufactured in hydrogel or silicone hydrogel materials in which their optical performance is in part achieved through increased thickness, and their higher modulus provides a masking effect similar to a GP lens. However, ultimately GP lenses tend to be better contact lens options for keratoconus patients.

In GP materials, we may select a corneal, hybrid, or scleral lens design. It is beyond the scope of this article to discuss details about all of the factors that play into the lens choice. They all have their pros and cons.

However, in our referral clinic setting, once a soft, inexpensive toric lens can no longer provide adequate correction, the trend is to recommend a scleral GP lens for keratoconus patients. This type of lens offers amazing comfort and is preferred by most such patients (Bergmanson et al, 2015). The visual acuity is usually better compared to that with any other optical device, and it may be obtained using, to good effect, wavefront-correcting optics.

Another advantage with a scleral lens is that it vaults over the cornea and limbus and does not present an additional challenge to the already compromised keratoconic cornea and its a fragile epithelium. With these traits, scleral lenses have the potential to further delay or even eliminate the need for surgery.

Other Surgical Options Outside of CXL, on the surgical side, we have an equally wide range of possibilities involving PKP, LKP, and intracorneal ring segments. Apart from late stages of keratoconus and when hydrops have occurred, all pathological changes appear to occur in the more anterior aspects of the cornea (Mathew et al, 2012). Therefore, LKP is a great way to approach a cornea in need of a transplant. This partial-thickness procedure has a shorter healing time, reduced risk of rejection, and better maintains global strength. With good surgeons, there is no difference in visual acuity outcomes. We ideally interact with surgeons skilled in the full range of corneal surgeries so that our patients get the surgery that is best for them and not simply the one procedure that a surgeon knows how to perform.

Intracorneal ring segments have not gained wide usage, and this may be due, at least in part, to the limited treatment effect that they appear to have. Intracorneal ring segments may be combined with contact lens wear, and this approach may extend the use of a soft lens correction.

Even though there is much that we do not know about the pathology of keratoconus, there is still a great deal that we can do for our patients who have this disease. Indeed, we have a smörgåsbord of options! CLS

For references, please visit and click on document #253.

Dr. Bergmanson is the Brien A. Holden Professor of Optometry at University of Houston College of Optometry and founding director of its Texas Eye Research and Technology Center (TERTC). He is licensed in Texas as a therapeutic optometrist and an optometric glaucoma specialist. Dr. Bergmanson is a Life Foundation Fellow of the College of Optometry in the United Kingdom and Fellow of the American Academy of Optometry, where he is a Diplomate in the Cornea and Contact Lens Section. He is a lifetime honorary member of the Swedish and Dutch Optometric Associations. He has written 150 publications, eight chapters, and one book.
Dr. Martinez is the current Cornea and Contact Lens Fellow at the TERTC. His work primarily concerns prescribing and managing patients who have irregular corneas including keratoconus, ectasia, and post-keratoplasty utilizing a range of scleral GP contact lenses.