DMEK combines ample benefits and daunting challenges

Full-thickness corneal transplantation, or penetrating keratoplasty, was the gold standard treatment for corneal disease until the advent of endothelial keratoplasty in the late 1990s.

Penetrating keratoplasty (PK) is indicated for combined stromal and endothelial pathologies. Endothelial keratoplasty (EK) is indicated for endothelial dysfunction. EK is performed through a small incision in a closed-globe setting, allowing faster visual recovery, minimizing refractive change and maintaining structural integrity to a greater extent than PK.

Currently, the leading EK technique is Descemet’s stripping endothelial keratoplasty (DSEK), also known as Descemet’s stripping automated endothelial keratoplasty (DSAEK). Donor tissue is dissected with a microkeratome. The endothelium and Descemet’s membrane are removed from the recipient cornea and replaced with a donor disc comprising endothelium and posterior stroma.

A newer variant, Descemet’s membrane endothelial keratoplasty (DMEK), involves replacement of tissue with donor endothelium and Descemet’s membrane.

Indications and contraindications for DSEK and DMEK are similar, leaving the choice of procedure largely up to surgeon preference, David R. Hardten, MD, OSN Cornea/External Disease Section Editor, said.

“The indications would be endothelial disease without significant stromal disease for both,” he said. “The contraindications for both would be patients who have significant stromal disease that would be better modified by a full-thickness or an anterior lamellar keratoplasty.”

Indications and contraindications for DSEK and DMEK are similar, according to David R. Hardten, MD. Image: Shari Fleming Photography

However, recent studies have shown that DMEK may offer superior visual outcomes and markedly lower graft rejection rates than DSEK. Additionally, the increasing availability of pre-stripped corneal tissue in the United States makes DMEK an attractive treatment option, Mark A. Terry, MD, said.

“Eye banks now are starting to prepare tissue for the surgeon, so the surgeon is no longer going to have to strip their own tissue,” he said. “That’s a huge advantage, just like it was with DSEK.”

Francis W. Price Jr., MD, said DMEK outcomes have improved steadily in the few years since he started doing the procedure.

“We found out fairly early that we get better visual results, definitely one or two lines better on the eye chart. Even more than that, the vision comes back much more quickly,” Dr. Price said.

Thomas John, MD, characterized DMEK as a breakthrough in corneal transplantation, akin to phacoemulsification as a groundbreaking innovation in cataract surgery.

“Once I switched from DSEK to DMEK, I have not gone back to DSEK,” Dr. John said. “DMEK is really a huge step forward. Once the surgeon gets comfortable with the technique of preparing the donor tissue and attaching it to the recipient eye, it’s a win-win situation both for the surgeon and patient because the quality of vision is significantly better and recovery is faster.”

Near normal appearing cornea following DMEK with restoration of normal anatomical layers of the cornea. Image: John T

Visual improvement, low rejection rate

“There’s a lot of new information on DMEK that’s coming out, a lot of hard work being done by the pioneers in DMEK that is really advancing the technology for utilizing DMEK,” Dr. Hardten said.

Dr. Price, who is considered one of those pioneers, said that DMEK offers faster, more complete visual recovery than DSEK or PK.

“DSEK had been as good as or better than PK with vision recovery. Not only was it quicker, but the overall vision was better. We just didn’t get that many people that were 20/20,” Dr. Price said. “Now, with DMEK, we’ve been getting 20/15 vision, 20/20 vision. We finally had gotten a procedure that was more or less like cataract surgery, where you would start getting the visual recovery in people you would expect.”

In Ophthalmology, Guerra and colleagues reported that DMEK offered better visual outcomes and yielded fewer refractive changes in the first postoperative year than DSEK. DMEK and DSEK had similar endothelial cell counts.

Donald T.H. Tan, MBBS, FRCSG, FRCSE, FRCOphth, FAMS, attributed strong visual gains to the absence of posterior stromal tissue in the DMEK graft.

Donald T.H. Tan

“DMEK is a more anatomically correct procedure than DSEK, as there is only an exchange of Descemet’s membrane and endothelium, without an additional layer of donor posterior stroma, as occurs with the DSEK procedure,” Dr. Tan said. “The lack of a stroma-to-stroma interface also results in the potential for better vision, and studies clearly show that more DMEK patients attain 20/20 vision as opposed to DSEK patients. In view of this, DMEK is therefore ideal for patients with relatively milder forms of endothelial decompensation.”

In a study published online ahead of print in Ophthalmology, Anshu and colleagues, including Dr. Price, reported that the likelihood of graft rejection was significantly lower with DMEK than with DSEK and PK (P = .004).

“Most people have noticed that their rejection rates are lower with DSEK, but they haven’t really quantified it like we have. We found that over the course of 2 years, we had a rejection rate of 12% for DSEK,” Dr. Price said.

DMEK had a graft rejection rate of 1% at 1 year and 2 years. DSEK had a rejection rate of 8% at 1 year and 12% at 2 years. PK had a rejection rate of 14% at 1 year and 18% at 2 years.

DMEK had a 15-fold lower risk of rejection than DSEK (P = .008) and a 20-fold lower risk than PK (P = .006), the researchers said.

“I think the last finding that we had about rejection is a game-changer,” Dr. Price said.

Francis W. Price

Suturing and topical steroids play a key role in rejection rates, he said. With PK, steroids often are stopped or decreased as soon as possible so the wound heals and the sutures can be removed because the corneal sutures distort the corneal surface and vision. However with PK, decreasing the steroids can lead to graft rejections.

“With DMEK, we typically use a 2.8 mm incision and a 5 mm one in DSEK, so sutures are not usually needed for the wounds and there is no need to taper the steroids to promote wound healing needed for visual recovery,” Dr. Price said. “With both DMEK and DSEK, we can keep patients on topical steroids unless they develop side effects like steroid-induced glaucoma.”

“Doctors across the board have different steroid dosing regimens,” Dr. Price said. He and German surgeon Frederick Kruse, MD, are embarking on a prospective study to evaluate different topical steroid dosing regimens for DMEK.

“Even though corneal transplant is the most successful type of tissue transplant done, and it’s been done routinely in the U.S. since the 1960s, there’s never been a prospective study during the first year after surgery evaluating different steroid dosing regimens,” Dr. Price said. “We are evaluating two different dosing regimens to see if we can cut the steroid doses, minimizing secondary glaucoma without getting rejections.”

Deep learning curve

Despite the reported advantages, DMEK presents an array of intraoperative and postoperative challenges, starting with graft preparation, Dr. Tan said.

“The major downside of DMEK today lies in the surgical challenges,” he said. “Donor preparation for DMEK is more difficult, as it remains challenging to consistently separate Descemet’s membrane without tearing, which may result in loss of that donor. Older donors are preferred, as their Descemet’s membrane is less adherent than in younger corneas and therefore easier to separate.”

Dr. Terry said that DMEK presents a steep learning curve, even for experienced surgeons.

Mark A. Terry

“It comes down to surgeons’ willingness to take on a brand-new learning curve,” Dr. Terry said. “DMEK is an entirely different surgical skill set than DSEK. So, the surgeon has to commit himself to learning another surgery skill. During that learning curve, the surgeon’s patients will have a higher complication rate than they have with his DSEK surgery.”

Dr. Terry described a clear trade-off between the superior visual outcomes and steep learning curve.

“With such good DSEK results right now, with DSEK yielding good vision of 20/25 over 65% of the time and 20/20 vision over 25% of the time, the average surgeon is not as compelled to take on another learning curve because most of his patients are really happy with DSEK and they’re not complaining,” Dr. Terry said. “We know that while DMEK will give them better vision and an even higher percentage of 20/20 results, the surgeon has to make the decision whether or not he is willing to go through the ordeal of learning a new skill set in order to get that little bit of extra vision when his patients are already happy with DSEK.”

Additionally, learning DSEK or DMEK is even more difficult in cases involving a glaucoma bleb or previous insertion of a drainage implant, Dr. Tan said.

“Pre-existing glaucoma blebs or seton implants are not a contraindication for EK, but it is recognized that surgery may be more challenging and a higher rate of donor dislocation exists due to inability of the anterior chamber to attain adequate air tamponade during or immediately after surgery,” Dr. Tan said.

In a study published in Cornea, Kim and colleagues reported that DSAEK proved to be a viable alternative to PK for corneal decompensation after previous implantation of a glaucoma drainage device.

“Although surgery may be more challenging, the presence of [a glaucoma drainage device] is not a contraindication to DSAEK surgery,” the researchers wrote.

Graft insertion, placement

DMEK does not require specialized equipment, even for the graft harvesting phase.

“You can manually do the stripping of the Descemet’s. You don’t need a microkeratome,” Dr. Terry said.

DMEK researchers are focused on improving graft harvesting techniques, Dr. Hardten said.

“There are a lot of people working within that area to try to make the process for harvesting more reproducible. There’s continued progress in that area,” he said.

Unlike DSEK, DMEK does not require a 5-mm incision for graft insertion, Dr. Terry said.

“With DSEK, when you insert the tissue through a smaller 3-mm incision, you crush the endothelium and you destroy it. We’ve always advocated a 5-mm or larger scleral incision for DSEK if you are not using an injector,” Dr. Terry said. “With DMEK, you can put the tissue in through a smaller clear corneal incision because the tissue rolls up so much that you can put it into an inserter that can go through a 3-mm, 2.5-mm or even a 1.8-mm incision. You just have to have an inserter that’s appropriate.”

Some surgeons use IOL injectors or glass pipettes to insert DMEK grafts, and newer delivery devices for DMEK are on the horizon, Dr. Terry said.

Thomas John

The ideal incision size for graft injection is 3 mm to 3.2 mm, Dr. John said.

“Keeping the incision size smaller retains the strength of the patient’s globe or eye compared to a larger incision. Because the tissue rolls on itself, we really don’t need a very large incision,” he said.

Dr. John described tissue manipulation in DMEK as a “no-touch” technique that largely depends on fluidics and careful handling of the graft.

“If all else fails and you’re still stuck in terms of an inability to unroll the donor scroll of Descemet’s membrane, there are special instruments that can be used to unroll it,” Dr. John said.

Dr. Tan said it is difficult to position the fragile DMEK graft without damaging it.

“Descemet’s membrane, when separated, also scrolls up, endothelial side out, and it is difficult to unscroll and position the donor Descemet’s membrane sheet right side up in the anterior chamber of the recipient eye without too much direct manipulation and damage to donor endothelial cells,” Dr. Tan said. “Wrinkling can also occur, and areas of detachment of Descemet’s membrane after DMEK surgery necessitate a much higher incidence of postoperative re-bubbling to reattach these areas, as compared to DSEK. During the steep learning curve of DMEK, there is therefore also a higher rate of endothelial cell loss. All these surgical factors currently limit the amount of DMEK surgery performed today.”

Detachment and re-bubbling

DSEK and DMEK involve the injection of an air bubble to force the graft tissue to adhere to the recipient tissue. DSEK requires only a small bubble that remains at the conclusion of the surgery to support the graft, while in DMEK, the bubble is significantly larger, presenting a new set of potential problems.

In their study, Guerra and colleagues reported DMEK re-bubbling rates ranging from 52% to 77%.

Dr. Terry said that re-bubbling rates in DMEK are high, even for surgeons who are well-versed in the technique.

“The conundrum that we have right now is that, even in the best of hands, by people who have a lot of experience with DMEK, the re-bubble rate is much higher with DMEK than it is with DSEK,” Dr. Terry said.

Dr. Terry cited published studies showing re-bubbling rates of 50% to 80% in several series by good surgeons.

“It’s harder to get the tissue to stick with DMEK than it is with DSEK,” he said. “My opinion is that it has to do with the difference in the interface between DSEK and DMEK. We’ve always advocated roughening the peripheral recipient interface in DSEK. That rough surface of the periphery combined with the rough surface of the donor from the microkeratome cut — those two rough surfaces help with adherence in DSEK. With DMEK, you have only glassy surfaces for the interface.”

In DMEK, an inferior peripheral iridotomy is needed to prevent the bubble from causing pupillary block, Dr. Terry said.

“With a standard DSEK surgery, you do not have to do an iridotomy because you can leave a very small bubble inside the eye, and that small bubble will not cause pupillary block,” Dr. Terry said. “That’s what we’ve been doing with hundreds and hundreds of eyes. With DSEK, we leave a small air bubble that is freely mobile, and you avoid pupillary block entirely. With DMEK, on the other hand, you have to leave a large air bubble in order to get this very fragile tissue to stick along the edges. So, because you have to leave a very large air bubble, we routinely do a peripheral iridotomy on every single case of DMEK that we do.”

Donor corneal tissue is most likely to detach from the recipient cornea within 1 week of surgery, Dr. John said.

“The surgeon really has to monitor the patient’s cornea and keep a close watch on how uniform and well the Descemet’s membrane is attached,” Dr. John said. “Any early detachment should be aggressively approached with the re-bubble technique to really attach the Descemet’s membrane. Once it’s well-attached and stays in place for at least a week or longer, then it’s smooth sailing, so to speak. But a close watch is essential.”

The high graft detachment and re-bubbling rates of DMEK may hamper its adoption, Dr. Hardten said.

“One of the current frustrations with DMEK is still the need to go back and add air in many patients, as well as the technical difficulties in preparing the tissue, and yet [I] have confidence that there will be continued improvement in the ability to overcome those barriers as more work and more time are invested in that procedure,” he said. “[Graft detachment] is one of the current limitations to that technique and tends to slow its adoption among many cornea users.”

New techniques

In Cornea, Dirisamer and colleagues described Descemet membrane endothelial transfer, in which a “free-floating” donor Descemet’s roll was identified in the anterior chamber of an 80-year-old patient who underwent DMEK for Fuchs’ endothelial dystrophy.

For the first few weeks after surgery, the researchers identified diffuse corneal edema with central corneal thickness of more than 1,000 µm in the transplanted cornea. However, after 1 month, the transplanted cornea showed signs of clearing. At 6 months, remnant edema was identified in the superior far periphery; central corneal thickness and endothelial cell density were normal.

The study authors suggested Descemet membrane endothelial transfer as a potential alternative treatment for corneal endothelial disorders.

In addition, Dr. Hardten noted that corneal collagen cross-linking is likely to reduce the demand for anterior lamellar corneal transplantation as a treatment for keratoconus and ectasia. The increased use of anterior lamellar transplants, though, may allow eye banks to separate tissue into anterior lamellar for keratoconus and then use the Descemet’s membrane from those corneas for DMEK.

Dr. Tan is working on a new surgical technique for DMEK.

“We are now evaluating a new surgical device for DMEK surgery, which we feel will make the surgery much easier to perform, as the device enables better control of the donor DMEK tissue,” he said. “Clinical trials on the prototype device are in progress.” – by Matt Hasson

Because of the potential for better vision with thinner EKs, is the continued effort to overcome the tissue and endothelial cell loss associated with DMEK still warranted?

Lindstrom's Perspective
Surgeons should embrace challenge of improving outcomes with DMEK

References:

• Anshu A. Price MO, Price FW. Risk of corneal transplant rejection significantly reduced with Descemet’s membrane endothelial keratoplasty [published online ahead of print Jan. 2, 2012]. Ophthalmology. doi:10.1016/j.ophtha.2011.09.019.
• Dirisamer M, Ham L, Dapena I, van Dijk K, Melles GR. Descemet membrane endothelial transfer: “free-floating” donor Descemet implantation as a potential alternative to “keratoplasty.” Cornea. 2012;31(2):194-197.
• Guerra FP, Anshu A, Price MO, Giebel AW, Price FW. Descemet’s membrane endothelial keratoplasty: prospective study of 1-year visual outcomes, graft survival and endothelial cell loss. Ophthalmology. 2011;118(12):2368-2373
• Kim P, Amiran M, Lichtinger A, Yeung SN, Slomovic AR, Rootman DS. Outcomes of Descemet stripping endothelial keratoplasty in patients with previous glaucoma drainage device insertion. Cornea. 2012;31(2):172-175.
• McCauley MB, Price MO, Fairchild KM, Price DA, Price FW Jr. Prospective study of visual outcomes and endothelial survival with Descemet membrane automated endothelial keratoplasty. Cornea. 2011;30(3):315-319.
• Price MO, Giebel AW, Fairchild KM, Price FW Jr. Descemet membrane endothelial keratoplasty: prospective multicenter study of visual and refractive outcomes and endothelial survival. Ophthalmology. 2009;116(12):2361-2368.
• Rudolph M, Laaser K, Bachmann BO, Cursiefen C, Epstein D, Kruse FE. Corneal higher-order aberrations after Descemet’s membrane endothelial keratoplasty [published online ahead of print Dec. 22, 2011]. Ophthalmology. doi:10.1016/j.ophtha.2011.08.034.

• David R. Hardten, MD, can be reached at Minnesota Eye Consultants, 710 E. 24th St., Suite 100, Minneapolis, MN 55404; 612-813-3600; email: drhardten@mneye.com.
• Thomas John, MD, can be reached at TJVI, 120 Oak Brook Center, Professional Building, Suite 204, Oak Brook, IL 60523; 708-429-2223; fax: 708-429-2226; email: tjcornea@gmail.com.
• Francis W. Price Jr., MD, can be reached at Price Vision Group, 9002 N. Meridian St., Suite 100, Indianapolis, IN 46260; 317-844-5530; fax: 317-844-5590; email: fprice@pricevisiongroup.net.
• Donald T.H. Tan, MBBS, FRCSG, FRCSE, FRCOphth, FAMS, can be reached at Singapore National Eye Center, 11 Third Hospital Avenue, Singapore 198751; 65-6227-7255; fax: 65-6323-1903; email: snecdt@pacific.net.sg.
• Mark A. Terry, MD, can be reached at Devers Eye Institute, 1040 NW 22nd Ave., Suite 200, Portland, OR 97210; 503-413-8202; fax: 503-413-6937; email: mterry@deverseye.org.
• Disclosures: Drs. Hardten, John and Terry have no relevant financial disclosures. Dr. Price receives travel grants from Moria. Dr. Tan receives royalties from Network Medical, UK, for the Tan EndoGlide DSEK donor insertion device.