(PDF 165 KB)
Deep anterior lamellar keratoplasty (DALK) is becoming the surgical procedure of choice for corneal diseases that do not involve the endothelium. The idea of removing only the anterior layers of the cornea to treat conditions like keratoconus, lattice dystrophy and corneal scarring has been around since the 1950s. But DALK procedures languished for decades because of poor visual results, due primarily to interface haze.
The results have improved in recent years with advances like Anwar’s “big bubble” technique to separate the stroma from Descemet’s membrane. The big bubble approach relies on an air bubble to achieve a complete and atraumatic baring of Descemet’s membrane. Despite the technical difficulty of DALK, this innovative technique—as well as recent forays into automated procedures with the microkeratome and the femtosecond laser—have sparked renewed interest in DALK procedures, which have several advantages over full-thickness grafts.
Maintaining an intact endothelium obviates or reduces several of the major challenges and vulnerabilities associated with corneal transplantation. “Theoretically, DALK leaves the patient with a stronger wound because you have not violated Descemet’s membrane, which is the strongest layer of the cornea,” said Woodford S. Van Meter, MD, professor of ophthalmology and director of the cornea and external disease service at the University of Kentucky in Lexington.
William Barry Lee, MD, attending physician at Piedmont Hospital and Children’s Health Care of Atlanta, pointed to the near elimination of interface haze when DALK is performed properly. “For decades the big struggle with DALK techniques involved how to do the procedure without leaving a scar in the patient’s cornea,” he said. “But if you can remove all of the patient’s stroma you’ve alleviated the cause of the scarring, and the big bubble technique really affords us the ability to do that now.” (See “Big Bubbles, Small Bubbles.”)
Safer than whole-cornea procedures. Dr. Lee added that during full-thickness corneal transplants the contents of the eye are exposed, potentially leading to a hemorrhage and loss of the eye. “The risk of an expulsive hemorrhage is significantly reduced with DALK,” he said, “and if an expulsive hemorrhage did occur, it would be easy to control and very unlikely to cause vision loss.”
And better recovery. Another potential advantage of DALK transplants is earlier suture removal, which translates into a faster recovery time and a more rapid reduction in suture-related astigmatism. In addition, the reduced threat of rejection means less need for steroids and, in turn, fewer steroid-related side effects.
Trauma resistant. Anecdotal reports also suggest that corneal transplants performed with DALK may stand up to mild-to-moderate blunt force trauma better than full-thickness grafts, which are especially vulnerable because of the weakness of the scar. Dr. Lee noted that a 12-year-old patient of his was jabbed in the eye with an elbow three months after undergoing a DALK transplant for keratoconus. “It broke all the sutures but didn’t pop the corneal transplant off—it stayed essentially intact, and the inside of the eye was not exposed,” he said. “I was able to just put more sutures in and didn’t even have to enter the inside of the eye.”
According to Natalie A. Afshari, MD, associate professor of ophthalmology at Duke University in Durham, N.C., “The overall trend in corneal transplantation has shifted more and more toward lamellar surgery as we try to leave some of the healthy cornea behind. For endothelial disease, we are moving more toward Descemet’s stripping endothelial keratoplasty, or DSEK. For anterior disease, if we can do DALK, it is our preferred approach.”
Big Bubbles, Small Bubbles
The big bubble technique, developed by Saudi surgeon Mohammed M. Anwar, MD, has made performing DALK easier and more likely to be successful. Creating a big bubble involves injecting air into the deep stroma to form a circular air pocket that will separate Descemet’s membrane from the stromal tissue.
Achieving a big bubble can be difficult, even among experienced surgeons. In one study, 60 percent of attempted big bubble detachments were unsuccessful.1 A normally developing bubble is round and appears to have a white band outlining the border. Signs of trouble include a bubble that begins to move to one side more than the other or a bubble that has an eccentric shape or looks like a water bubble without a clearly defined circular edge.
Singaporean surgeon Donald T. Tan, MD, developed the “small bubble” technique to assist surgeons in confirming that a big bubble has been created.2 The technique involves injecting a small air bubble into the anterior chamber from the limbal periphery. If the small bubble can be seen at the periphery, the big bubble separation has been successful. This is because the convexity of the big bubble protrudes posteriorly, pushing the anterior chamber small bubble to the periphery. In contrast, if the small bubble disappears immediately after being injected into the anterior chamber, a big bubble has not been achieved.
1 Borderie, V. M. et al. Arch Ophthalmol 2008;126(1):31–37.
2 Parthasarathy, A. et al. Br J Ophthalmol 2008;92(3):422.
A Difficult Procedure
The main drawback to greater use of DALK is its technical difficulty. Dr. Van Meter noted that it is a surgically demanding dissection, both in the donor and in the recipient. In the recipient, most of the cornea needs to be hand-cut.
“The big challenge with DALK is knowing how to go deep enough without cutting through Descemet’s membrane and making it a penetrating procedure,” Dr. Van Meter said. “Leaving any stroma will cause scarring, so the trick is to get as close as you can without entering Descemet’s.”
Stop after the stroma. Injecting air into the stroma can help the surgeon visualize where the stroma ends and Descemet’s membrane begins, Dr. Van Meter said. “After you make your cut around the periphery, you inject intra- stromal air, which turns the stroma white. Then you feather it down—and you can do this in quadrants—trying to reach Descemet’s membrane without entering the anterior chamber.”
Dr. Lee noted that Descemet’s membrane is only 15 to 20 µm thick. “It’s difficult to dissect the very thin layers of the cornea without causing a perforation. A large perforation would potentially cause poor vision recovery and would typically require converting to a full-thickness procedure intraoperatively.” He added that if a large perforation occurs during a big bubble procedure, it is easy to convert to a standard corneal transplant. “You really don’t lose anything by trying.”
Adding Automation to DALK
Automated procedures reduce the risk of perforation in DALK, but with some caveats.
Microkeratome. A device like the microkeratome can add precision to partial-thickness cuts, but it cannot remove the entire stroma. “The microkeratome can’t dissect down that deep without risking perforation, so it will always leave a little posterior stroma on the host cornea,” Dr. Lee said. “As a result, your risk of scarring is always a little higher with this technique than with the big bubble. But the risk of perforation is a lot lower because it’s an automated procedure.”
Femtosecond laser. Anecdotal reports suggest that the femtosecond laser may prove to be a better automated approach for DALK. The laser can be programmed for exact depths, allowing the surgeon to go deeper into the stroma. “It probably gives you a little more accuracy in the depth of the cut, and there have been some really promising early reports,” Dr. Lee noted.
For now, femtosecond-assisted DALK is experimental, and cost is a limiting factor. “The IntraLase device is expensive to begin with and then it requires special corneal transplant software, which is also expensive,” Dr. Afshari said.
Dr. Van Meter described the femtosecond laser as a “useful but theoretical tool” for corneal transplantation. The logistics are challenging, he said, because the laser is not geared for being in a sterile operating room, requiring the patient to be moved during the procedure.
Dr. Lee noted that in addition to the costs of the laser and the add-on software, the company charges a “click fee” every time the laser is activated. “Right now insurance doesn’t reimburse for that click fee, so the financials of doing femtosecond DALK don’t work out well for patients or cornea surgeons.” He added that in the future it is conceivable that reimbursement might be handled the way multifocal implants are today, with insurers paying for a portion of the surgery and patients paying an add-on fee for the femtosecond laser.