Focus on Cornea
In endothelial keratoplasty, advantage DMEK
A more challenging procedure, but outcomes are demonstrably superior.
By Mark S. Gorovoy, MD, with Ian R. Gorovoy, MD
The alphabet of cornea endothelial transplants continues to expand. The past 10 years have seen more advances in corneal transplant for endothelial dysfunction then perhaps the preceding 50 years. Despite the notion that Descemet’s membrane endothelial keratoplasty (DMEK) may represent the ideal anatomical form of endothelial keratoplasty, another new and difficult learning curve is hampering universal adoption.1
Moreover, whereas Descemet’s stripping automated endothelial keratoplasty (DSAEK) was revolutionary compared to penetrating keratoplasty (PK) and quickly became the standard of care for endothelial pathology, DMEK is best described as an evolutionary advancement. So it is reasonable to expect that DMEK may be slower in becoming the dominant endothelial keratoplasty procedure, especially because DMEK requires the use of extremely thin and fragile donor tissue.
Mark Gorovoy, MD, preformed this DMEK graft, shown here at day one postoperatively.
COURTESY MARK S. GOROVOY, MD
DMEK is worth learning
As with any new procedure, the learning curve for DMEK presents a significant investment in time and resources for the surgeon, as well as an increased risk for the initial patients.2
DMEK has two learning curves; the donor tissue stripping must be performed as skillfully as the procedure itself. Donor tissue loss can result in scheduling issues and economic loss. As DMEK advocates, we concede the incremental advantages of DMEK over DSAEK may be a difficult “sell” to both the patient and the surgeon.
All the advantages DSAEK has over PK for endothelial disease also apply to DMEK, but advance to a higher level. These include quicker rates of postoperative visual recovery, more predictable refractive outcomes, fewer instances of postoperative rejection and improved vision quality.
Four advantages of DMEK
• Faster rate of postoperative visual recovery.
• More predictable refractive outcomes.
• Lower incidence of postoperative rejection.
• Improved quality of vision.
DMEK vs. DSAEK outcomes
Specifically, studies have shown that fewer than 20% of DSAEK patients achieve 20/20 BSCVA compared to nearly 50% of DMEK patients. But Snellen acuity only captures part of the visual superiority of DMEK compared to DSAEK and may in fact underestimate the true difference. All of these qualities ensure that the popularity of DMEK will undoubtedly grow.3–5
Here, we will share our experience and pearls regarding DMEK and highlight some of the major differences between the two procedures.
DONOR TISSUE SELECTION
Eye banks could spur DMEK adoption
While eye banks cut the vast majority of DSAEK donor tissue, surgeons have prepared most DMEK tissue.
Tissue preparation remains a barrier to surgeon transition. Although experienced surgeons have reported successful peel rates well over 95%, eye banks will hopefully see the opportunity and the need to supply donor tissue for DMEK to eliminate the risk of donor loss.6
Harvesting DMEK tissue should not be a daunting task for eye banks. Standards for DMEK and DSAEK donor tissue are similar. They are:
• Endothelial cell count more than 2,300 cells/mm2.
• No more than 12 hours death to preservation time (which can extend to 24 hours with refrigeration of the tissue).
• Tissue age no more than seven days and no previous intraocular surgery.
Hazards of harvesting
One quality unique to DMEK donor tissue is that, when harvested from a person age of 50 or younger, it is more likely to develop tight spontaneous scrolls and subsequent graft detachment, and does not offer significantly better postoperative endothelial cell counts.
Another important quality is the strong correlation between peel times and the risk of graft tear for the two eyes from the same donor, which speaks to some inherent adhesive qualities of individuals’ Descemet’s membrane that undoubtedly exist but have not been elucidated. We believe that tissues from donors age 50 or younger or tissue from the second eye when the first eye had a peeling complication are better suited for DSAEK, PK or lamellar grafts.
DMEK PATIENT SELECTION
Identifying candidates
Like DSAEK, DMEK is strictly indicated for patients with endothelial dysfunction. Fuchs’ dystrophy remains the most common etiology followed by failed prior DSAEKs, bullous keratopathy and ICE syndromes.
Currently, bullous keratopathy patients comprise a smaller group suited for DMEK because they have had prior vitrectomies for anterior segment complications, a group better served with DSAEK. Vitrectomized eyes are poor candidates for DMEK because the transplant surgeon can have difficulty shallowing the anterior chamber while keeping the cornea convex.
In addition, many patients with histories of complicated surgeries may not have BSCVA potential of 20/20, further reducing the benefit-risk ratio of DMEK compared to DSAEK. Eyes with blebs and/or shunts are still reasonable candidates for both procedures — emphasizing again that prior vitrectomy rules out DMEK.
Some patients can avoid cataract surgery
DMEK patients with clear lenses undergo a similar but more liberal algorithm regarding the ability to forgo cataract surgery.
For DSAEK, the requirements to leave a patient phakic are:
• Age 50 or younger.
• No significant refractive error after calculating the 1.00-1.50 D hyperopic shift that occurs after DSAEK.
• An anterior chamber depth of 2.8 mm or greater.
In DMEK, the age cutoff and refractive criteria are the same (accounting for no significant post-DMEK refractive change), but anterior chamber depth is not a consideration as a shallow chamber may in fact actually aid in surgery.
Lastly, the abnormal pupil is not a barrier to successful DSAEK. However, a pupil that does not constrict or has large iris defects, or both, are barriers to successful DMEK. The choice to do a combined cataract-DMEK vs. a staged procedure involves the same decision-making process as DSAEK. We prefer the latter in both instances.
PERFORMING THE OPERATION
The first step in DSAEK is to place the graft in the eye with forceps through a 3.2-mm incision. In DMEK, insert the graft into the eye with an IOL injector through a much smaller 2.4-mm incision. This smaller corneal incision in DMEK provides for less postoperative astigmatism.
The next step in DMEK is to strip the host Descemet’s tissue as in DSEAK; although the area stripped in DMEK is slightly larger, the donor tissue, on average, is smaller — 9 mm in DSAEK, 8 mm in DMEK.
Clearly unfolding the graft is quite different between the two procedures, as DMEK involves a much thinner and fragile tissue. Surgeons have used various techniques to unscroll DMEK donor tissue, including an air bubble under or over the donor tissue. We prefer no air bubbles and rely instead on corneal dome compression to position the graft. This technique avoids any instrument-donor touch.
Immediate postoperative complications
Primary failure in DSAEK procedures are rare (less than 1%), but DMEK in our clinic has a failure rate of approximately 3%.
Dislocations in DSAEK are anatomically completely different than DMEK. In DSAEK, detachments are almost always complete and can be treated with re-bubbling in the vast majority of cases. Delayed re-bubbling does not provide for worse outcomes.
In contrast, DMEK dislocations are typically peripheral and rarely complete. These peripheral detachments can be simply observed without treatment, but a time line exists for successful re-bubbling if the detachment becomes more central. At approximately six weeks postoperatively, the tissue scroll is fibrotic and not amenable to re-bubbling. Re-bubbling is less frequently successful with DMEK than DSAEK and also involves uncurling the graft to avoid a permanently folded donor tissue edge.
Postoperative medications
Regarding the typical postoperative drop regimen, postoperative steroids can be tapered more quickly with DMEK due to decreased rates of graft rejection. Typically, after two months of QID steroid drops, DMEK patients can be tapered to once a day, which also decreases the rate of steroid-responders
Another important advantage of DMEK: The initial DMEK cell count data exceeds those of DSAEK. In addition, the standard deviation of cell loss appears to be much smaller.
DMEK: challenging but superior
In recent years DMEK has become the preferred technique in the appropriate eyes. The better quality of vision is the driving motivation. As a secondary positive factor, preliminary cell counts are significantly higher in DMEK than DSAEK.
The same controversies that exist in DSAEK, such as increased rates of endothelial loss in the setting of shunts or blebs and when to perform a staged or combined procedure, exist with DMEK.
This DMEK graft at three months postoperatively exhibits a peripheral scroll and clear central cornea. Vision is 20/20 and endothelial cell count is 2,000.
COURTESY MARK S. GOROVOY, MD
A group of transplant surgeons are now promoting ultra-thin DSAEK (UT-DSAEK) as an alternative to DMEK, but this procedure still falls short of the unique anatomical advantage of DMEK surgery. Leaving stromal fibers behind can only degradate the final visual acuity. Given the advantages DMEK offers the patient, we anticipate that DMEK will become the standard of care for endothelial pathology until a biological treatment is discovered. OM
REFERENCES
1. Terry MA. Endothelial keratoplasty: why aren’t we all doing Descemet membrane endothelial keratoplasty? Cornea. 2012;31:469-471.
2. Gorovoy, MS. DMEK complications. Cornea. 2014;33:101-104.
3. Melles GR, Ong TS, Ververs B, van der Wees J. Preliminary clinical results of Descemet membrane endothelial keratoplasty. Am J Ophthal. 2008;145:222-227.
4. Guerra FP, Anshu A, Price MO, et al. Descemet’s membrane endothelial keratoplasty; prospective study of 1-year visual outcomes, graft survival, and endothelial cell loss. Ophthalmology. 2011;118:2368-2373.
5. Guerra FP, Anshu A, Price MO, Price FW. Endothelial keratoplasty: fellow eyes comparison of Descemet stripping automated endothelial keratoplasty and Descemet membrane endothelial keratoplasty. Cornea. 2011;30:1382–1386
6. Price MO, Giebel AW, Fairchild KM, et al. Descemet’s membrane endothelial keratoplasty; prospective multi-center study of visual and refractive outcomes and endothelial survival. Ophthalmology. 2009;116:2361-2368.
About the Authors | |
Mark S. Gorovoy, MD, (top) is founder of Gorovoy MD Eye Specialists in Fort Myers, Fla. His e-mail address is mgorovoy@gorovoyeye.com. | |
Ian R. Gorovoy, MD, is a third-year ophthalmology resident at the University of California San Francisco. | |
Disclosures: The authors disclosed they have no conflicts with any company mentioned in this article. |