Ectasia—whether post-LASIK or keratoconus—has always been an unwelcome occurrence. Traditional options for treatment, most notably penetrating keratoplasty, can salvage vision but have significant downsides. Fortunately, new options have proliferated in recent years. These have included partial transplant techniques, and more recently using riboflavin and ultraviolet light to increase the biomechanical strength of corneal tissue. Here, we describe the latest work with some of these options, including crosslinking (with and without PRK), deep anterior lamellar keratoplasty and tectonic lamellar sclerokeratoplasty.
Crosslinking Corneal Collagen
One of the most promising new approaches for managing ectasia is collagen crosslinking, or C3-R, in which riboflavin is applied to the cornea (usually after epithelial removal) followed by about 30 minutes of UV light irradiation. This causes new bonds to form between adjacent collagen molecules in the cornea, increasing corneal rigidity as much as three or four times. In several recent studies this has not only stopped the progression of keratoconus, but has caused some degree of reversal and improved visual acuity in many eyes.1,2 The procedure is not yet approved by the U.S. Food and Drug Administration, but has received the CE mark in Europe.
"Studies in both the lab and clinic have shown that the C3-R procedure does strengthen the cornea and stop the progression of primary occurring keratoconus," says Brian S. Boxer Wachler, MD, director of the Boxer Wachler Vision Institute in
"There are two ways the procedure can be done," he explains. "You can leave the epithelium intact, which is referred to as 'epi-on,' or you can remove it—'epi-off.' The original studies in
John Kanellopoulos, MD, clinical associate professor of ophthalmology at
"I became involved in the crosslinking technique out of desperation, trying to manage post-LASIK ectasia patients," he explains. "I used Intacs to treat keratoconus for more than 10 years, and patients were thrilled with the change in their visual function. However, keratoconic corneas are much softer than normal corneas. More than half of the Intacs segments migrated through the cornea and surfaced.3 We had to remove them because of the risk of infection.
"Once I learned the crosslinking technique, I discussed it with a patient who was due for a cornea transplant because of severe ectasia and an inability to rehabilitate with contact lenses or Intacs," he continues. "Not only did the crosslinking freeze the ectasia for about a year, it appeared to reverse some of it."
Dr. Kanellopoulos says the treatment has produced excellent results, but that he's had to graft some patients even after crosslinking. "In a few cases the cornea was stabilized at six months, but we couldn't visually rehabilitate the patient enough to allow him to return to normal function," he explains. "For example, if I freeze a cornea with 56 D of keratometry, and the patient is contact lens intolerant, that patient can't function. So, we have to resort to a graft."
Nevertheless, he considers crosslinking an advantage even if corneal transplantation is still necessary. "At least the residual host cornea is now a little tougher for receiving the donor cornea," he says. "In many cases after a transplant the weakened foundation of the ectatic cornea leads to a lot of postop astigmatism and instability for the graft. So crosslinking may be an advantage even in the worst cases."
Ronald R. Krueger, MD, medical director of the Department of Refractive Surgery at the Cleveland Clinic's Cole Eye Institute, has not performed the C3-R procedure in his own practice, but has worked with other surgeons who have. "One of the biggest limitations of the riboflavin treatment," he says, "is that you need to have a cornea that's thick enough to allow the riboflavin to be strongly absorbed and prevent the UV light from penetrating to the endothelium where it would cause damage. Right now, the accepted cutoff for corneal thickness is about 400 µm. Ironically, ectasia and keratoconus are corneal thinning conditions, and some keratoconic eyes can be as thin as 400 µm." Dr. Krueger notes that it might be possible to adjust the UV light exposure or the concentration of the riboflavin so as to make this work safely on a thin cornea—or even increase safety by using dextran or other medications to cause temporary swelling of the tissue. But for now, he says, most researchers are sticking with the established protocol.
"The possibility of using this to treat post-LASIK ectasia raises interesting questions," he adds. "If the flap is part of the problem, perhaps because it's thicker than intended, and now the anterior tissue is not adherent to the posterior tissue, crosslinking could conceivably knit the flap together better or just make the stroma in the flap stronger. And what will happen if we lift the post-LASIK flap, put riboflavin on the stromal bed, lay the flap back down and then shine the UV light in so it focuses on the area underneath the flap? Researchers are exploring these kinds of possibilities. It might even be possible to use different strategies as the ectasia progresses—assuming it does progress."
Given the early success of the C3-R treatment, researchers in the
One of the most promising uses of the C3-R procedure is in combination with a modified version of PRK. In about 200 of the 300 cases in which Dr. Kanellopoulos has used the crosslinking technique he has also performed PRK, using the WaveLight topography-guided laser platform to normalize the shape of the cornea.
"This is not a straightforward PRK," he explains. "It combines a myopic PRK over the apex of the cone with a segment of hyperopic PRK at the opposite side of the cornea. The combination flattens the ectatic part of the cornea and steepens the part of the cornea that's very flat. [See illustration, above.] As a result, you're removing very little tissue but making the cornea smoother."
Dr. Kanellopoulos notes that this type of PRK is a specialized intervention because treating the refractive error isn't necessarily the goal. "The goal is to normalize the cornea as much as possible to increase BSCVA," he says. "Invariably BSCVA increases if you decrease the amount of irregular astigmatism. So the number one treatment target is cylinder—i.e., the irregular astigmatism—and the number two target is correcting some of the sphere. Here, success is turning a cornea that's -4 -4 at 70 degrees into a cornea that's -3.5 -0.5 at 70 degrees. That's a tremendous success story because that eye can see 20/20 with a soft -3.5 contact lens. More importantly, that eye does not need a cornea transplant.
"Some surgeons might argue that this could be done using wavefront-guided technology," he says. "In mild cases of ectasia that's true. However, a wavefront-guided system will try to ablate tissue until it's all equal, so it removes three times as much. Tissue reserve is a big issue in ectasia cases."
Dr. Kanellopoulos adds that one surprising effect of the crosslinking treatment is that corneal tissue reacts differently to laser ablation, which is an issue if the cornea is crosslinked before ablation. "In crosslinked corneas we consistently get more refractive result than we expect," he says. He hasn't yet found a formula to compensate for the difference. "We're talking about very irregular ablations," he notes. "However, we recommend undercorrecting by 25 percent to play it safe. If the patient is -2, -2 after crosslinking, I would treat -1.5, -1.5."
Perfecting Crosslinking with PRK
Dr. Kanellopoulos says that at the outset he crosslinked the cornea first and came back to normalize the cornea with PRK if it seemed necessary. But once it was clear that this was a safe and effective combination, he thought: Why not do both procedures at the same time?
At that point it seemed reasonable to try doing the PRK first. "The crosslinking occurs more vigorously in the surface of the cornea, dissipating as you go deeper," he says. "So the best cornea you're creating is right under the surface. Going back six months later and ablating some of that tissue by doing a limited PRK will undercut the biomechanical strength you've added. If you ablate first and then crosslink, the tissue that's left behind is crosslinked the best possible way, and you have a more stable biomechanical effect." Dr. Kanellopoulos says his clinic now treats 80 percent of its patients with this one-stop/two-part procedure. Only 20 percent need a second enhancement.
Dr. Kanellopoulos notes that ablating an ectatic cornea is tricky because of the conflicting goals. "On the one hand, we're trying to normalize the cornea, which requires tissue thinning," he says. "On the other hand, we want to keep the cornea as thick as possible to make sure it doesn't go into ectasia. Right now we don't know exactly how much ablation is safe; hopefully that will be answered in the future."
In the meantime, he and his colleagues have settled on an arbitrary limit that seems to provide a sufficient margin of safety. "The average keratoconic eye we treat has a thickness of about 450 µm," he says, "so we've set 50 µm as the arbitrary limit to the amount of tissue we'll remove centrally. Within that limit, we try to remove as much cylinder as possible and some of the sphere.
"For example," he continues, "if a patient is -12 -4, with a corneal thickness of 440 µm, we'll treat 3 D of astigmatism and 1 or 2 D of myopia. With this much myopia, uncorrected vision won't improve much because he's still going to have about 10 D of myopia and 1 D of astigmatism. However, it's a completely different ballpark in terms of visual function. Before surgery this patient almost certainly needed an RGP contact lens. Postop, he can wear glasses or a soft contact lens to get vision in the range of 20/25 or 20/30. And in some cases we may be able to treat the entire refractive error and then crosslink the cornea—say, if the eye is -2, -3, with a thickness of 450 µm. The result is that we can get this cornea from a BSCVA of about 20/40 to a UCVA of 20/20.
"Some of these eyes will have some aberrations and night-vision problems," he admits. "But I don't compare these eyes with my LASIK population; I compare them with my cornea transplant population."
Dr. Kanellopoulos believes crosslinking has great potential in areas besides treating ectasia. "In the future, we may crosslink some of the high myopes we do laser surgery on," he says. "Crosslinking seems to have a remarkable effect on post-LASIK regression. One patient, a very high myope who was successfully treated with LASIK, was
Dr. Kanellopoulos admits that many questions remain. "Right now, almost all investigators are working with a certain concentration of riboflavin, a certain amount of energy and exposure to the cornea," he observes. "All of these variables need to be played with in order to find the gold-standard version of this treatment. It may be possible to achieve the same effect in less time or with greater safety."
He notes that it does take years for a procedure to establish its credibility, but adds that his own longest patient follow-up is close to five years, and researchers in Dresden have crosslinked patients they've been following for almost 10 years. "It's a shame that this technology appears to be nowhere near FDA approval in the
Dr. Krueger points to several factors standing in the way of FDA approval of the crosslinking treatment any time soon. "For one thing, the FDA isn't quite sure how to define this," he says. "Is it a drug or a device? That's two different approval processes at the FDA. Also, no company has fully gotten behind this in terms of funding—maybe because nobody is sure whether this will be a moneymaker for them. Nevertheless, I think a lot of physicians in the
The DALK Alternative
Another option for treating keratoconus and ectasia after LASIK that has been slowly gaining popularity is deep anterior lamellar keratoplasty, in which the diseased cornea is removed down to Descemet's membrane, leaving the host endothelium intact. Preserving the endothelium has the effect of minimizing the risk of donor tissue rejection.
Kenneth Goins, MD, professor of ophthalmology at the
"Whenever possible, I try to preserve a healthy and intact corneal endothelium, especially in diseases such as keratoconus," says Dr. Goins, noting that he doesn't see many post-LASIK ectasia cases. "The best results are obtained when you're able to fully dissect down to Descemet's membrane; this often results in 20/25 or better vision. If there is a microperforation of Descemet's membrane during DALK, a little stroma can be left to cover the defect, usually resulting in a good visual outcome. The patient may have less than 20/20 vision in that situation, but depending on what created the need for surgery, an excellent outcome may still ensue." Dr. Goins notes that some surgeons simply convert to penetrating keratoplasty if they fail to get a perfect dissection, which he says is a legitimate option.
"The main advantage of DALK over PK is preservation of the corneal endothelium," he explains. "The rate of endothelial decay is less after DALK than after PK. Also, the incidence of allograft rejection may be reduced, and it's basically an extraocular procedure, so it may leave the globe structurally stronger than a standard keratoplasty." (Other advantages often noted by surgeons include the absence of wound leakage because of the closed system, the ability to use less healthy donor tissue, earlier removal of stitches and more rapid cessation of topical corticosteroids.) "The long-term vision results are good," he adds. "A high percentage of patients achieve 20/40 or better."
Dr. Goins says the improved structural stability can be very beneficial if you're doing keratoplasty surgery on a patient who may be more likely to rub the eye. "I recently performed DALK on two patients with Down Syndrome," he says. "They both rubbed their eyes and broke the sutures, but because it was partial thickness surgery I was able to get them to surgery and close the eye. I didn't have to worry about losing the eye from a suprachoroidal hemorrhage."
Dealing with the Downsides
Despite his enthusiasm for DALK, Dr. Goins admits that it's a much more difficult procedure to master. "I've done 15 or 20 cases and I'm still learning," he says. "It requires skill and experience most surgeons don't have."
In addition to the learning curve, Dr. Goins notes that the procedure has some unique complications that can be difficult to avoid. "One is microperforation of Descemet's membrane, which sometimes happens for no apparent reason," he says. "The patient may also have a detachment of the membrane following surgery, which is very problematic; it may be necessary to inject air into the eye in order to reattach."
Postoperative wrinkles in Descemet's membrane are another possible complication. "These may be related to suture tension," he says. "If they occur in the visual axis, they can be devastating. Fortunately, most wrinkles go away when the sutures are removed. Last, but not least, if the dissection is not carried all the way down to Descemet's membrane, you may encounter interface haze that can interfere with vision."
Dr. Goins says that it's still difficult for him to predict which patients will have the best outcomes, because he doesn't know when he might have to stop the dissection short of Descemet's membrane. For that reason, he says it's important to choose DALK patients carefully. "If a patient needs to be 20/20 BCVA after the surgery, I'd probably do a PK or IntraLase-enabled keratoplasty instead of DALK," he says. "But most of these patients have really horrible vision in the affected eye when initially evaluated, so any improvement is appreciated. Many are very happy to keep their own endothelium and minimize the chance of rejection."
Dr. Goins believes DALK use in the
Dr. Goins says the bottom line is not that DALK is necessarily better for treating ectasia than PK, but is a valuable alternative option. "DALK can provide visual results similar to PK with decreased chance of allograft rejection and longer graft survival," he says. "Overall, the complication rate is higher than with PK, but the advantages are worth the risk."
Replacing the Entire Cornea
In most cases, anterior segment ectasia only affects the central or paracentral cornea, allowing a graft to be sutured to the healthy corneal tissue that remains at the periphery. In some cases of advanced keratoconus, however, the thinning can involve the entire cornea.
In this situation, the procedure of choice may be tectonic lamellar sclerokeratoplasty, in which a much larger graft including scleral tissue is used. Salomon Esquenazi, MD, assistant professor of ophthalmology and neurosciences research at
"This is a last resort for a patient with end-stage ectasia," he says. "Of course, these eyes have very poor, compromised vision, and a very poor prognosis for recovery of vision. But all of our patients are doing well—the tectonic grafts have been successful.
"The case we published last year involved a 41-year-old patient we saw at LSU who had complete ectasia of the anterior segment," he continues. "Because the whole cornea was involved, we didn't have a bed onto which we could suture a corneal graft. So, we used a huge 14-mm graft that included cornea and a piece of sclera and sutured it onto the patient's sclera. We published this case because this was a very young patient and we were able to rehabilitate his eye. He would have had no chance for success with any other procedure."
Dr. Esquenazi notes that in some cases following a tectonic graft, the center of the cornea may become opaque, although this is not the result of secondary ectasia. "With standard PK where you use a small graft, the chances of rejection are much less," he explains. "When you use bigger grafts, rejection is more likely, although in this situation you don't have any other option. If the central cornea does become clouded, you can do a penetrating keratoplasty involving only the center of the cornea on top of the tectonic graft. You'll have a healthy bed onto which you can suture the secondary graft."
Because the graft is larger and involves scleral tissue, Dr. Esquenazi says tectonic lamellar sclerokeratoplasty is much more difficult than a standard PK. "The technique is similar to penetrating keratoplasty," he notes. "It takes a little more time, but you employ the same instruments. However, in a normal penetrating keratoplasty you transplant a corneal graft that measures between 7 and 8 mm in diameter. Here you're transplanting a 14-mm graft. You have to open the conjunctiva and do a peritomy. So it's much more involved."
1. Wollensak G. Crosslinking treatment of progressive keratoconus: new hope. Curr Opin Ophthalmol. 2006;17:4:356-60.
2. Caporossi A, Baiocchi S, Mazzotta C, Traversi C, Caporossi T. Parasurgical therapy for keratoconus by riboflavin-ultraviolet type A rays induced cross-linking of corneal collagen: Preliminary refractive results in an Italian study. J Cataract Refract Surg 2006;32:5:837-45.
3. Kanellopoulos AJ, Pe LH, Perry HD, Donnenfeld ED. Modified intracorneal ring segment implantations (INTACS) for the management of moderate to advanced keratoconus: efficacy and complications. Cornea 2006 Jan;25:1:29-33.
4. Chan CCK, Sharma M, Boxer Wachler BS. The effect of inferior segment Intacs with and without corneal collagen crosslinking with riboflavin (C3-R) on keratoconus. J Cataract Refract Surg 2007;33:75-80.
5. Esquenazi S,