Surgeons make the case for DSEK vs. DMEK
Parag A. Majmudar, MD, and Kenneth A. Beckman, MD, FACS, weigh the pros and cons of switching from DSEK to DMEK.
CEDARS Debates is a monthly feature in Ocular Surgery News. CEDARS — Cornea, External Disease, and Refractive Surgery Society — is a group of cornea, cataract and refractive surgery specialists, here to discuss some of the latest hot topics in ophthalmology.
This month, Parag A. Majmudar, MD, and I discuss the advantages and disadvantages of Descemet’s membrane endothelial keratoplasty and Descemet’s stripping endothelial keratoplasty. While endothelial keratoplasty has become the surgical treatment of choice for endothelial disease, there are ongoing debates as to which procedure is preferred. DSEK has been around for several years and has widespread acceptance, but the newer DMEK appears to be gaining acceptance and may have superior results. We hope you enjoy this discussion.
Kenneth A. Beckman, MD, FACS
OSN CEDARS Debates Editor
I got DSAEK down; why should I switch to DMEK?
Right before our eyes, we are seeing the evolution of endothelial keratoplasty as a refractive procedure. Ten years ago, we were performing penetrating keratoplasty for Fuchs’ corneal dystrophy; the refractive results were abhorrent. Cataract surgery, LASIK and other refractive surgery were delivering outstanding visual outcomes. But our Fuchs’ patients were getting left behind.
Enter Descemet’s stripping automated endothelial keratoplasty. In 2007, this procedure started to gain momentum. No longer were patients left to deal with 4 D and 5 D of surgically induced cylinder postoperatively. Refractive accuracy had finally come to the field of corneal transplantation. In 2012, 23,049 endothelial keratoplasty cases were performed in the United States. Of those, nearly 97% were DSAEK.
So, if so many corneal surgeons are experiencing great results with DSAEK, why should they make the transition? In a word, outcomes. Just as innovation spurred the development of LASIK over PRK and the relegation of radial keratotomy to a footnote, DMEK refractive and visual outcomes will be the main reason why surgeons adopt this technique.
The advantages of DMEK are numerous: more anatomic result with no optical interface, faster visual recovery, lower rejection rates and better refractive outcomes. Most patients who undergo DMEK experience functional vision within 3 to 4 weeks vs. up to several months in DSAEK. The presence of the optical interface in DSAEK likely limits their final acuity. In a recent article, the mean best corrected visual acuity was 20/24 in DMEK patients vs. 20/34 to 20/66 in DSAEK patients. In another series, 95% of DMEK patients were 20/40 or better and 50% were 20/25 or better vs. 43% of DSAEK patients better than 20/40 and only 6% of DSAEK patients better than 20/25. DMEK has been shown to induce fewer higher-order aberrations and induce less of a hyperopic shift than DSAEK (0.7 D to 1.5 D in DSAEK vs. 0.24 D in DMEK).
Images: Majmudar PA
Despite some surgeons performing ultra-thin DSAEK, the visual results in DMEK are still better. At the upcoming American Society of Cataract and Refractive Surgery meeting, Terry et al are scheduled to report on the results of a contralateral eye study of ultra-thin DSAEK eyes (mean thickness: 89 µm) vs. DMEK in 16 patients. The mean preoperative acuity for DSAEK eyes (20/50) and DMEK eyes (20/40) was not statistically different. But the mean best postop vision of ultra-thin DSAEK (20/32) was statistically worse than vision after DMEK (20/25) (P = .05); 31% of DMEK eyes vs. only 13% of ultra-thin DSAEK eyes had acuity of 20/20, and 75% of patients preferred their DMEK eye to their DSAEK eye.
Endothelial cell loss has not been shown to be markedly different between the two techniques, but the overall rejection rate may be lower with DMEK. Price et al showed a 1% rejection rate in DMEK vs. 12% in DSAEK, and Anshu et al showed that one out of 141 DMEK grafts experienced rejection (0.7%) vs. 54 out of 598 DSAEK grafts (9%).
So, what are the factors limiting DMEK adoption?
- “The procedure is too hard.” I would submit that the learning curve from PK to DSAEK is much harder and longer than the curve of DSAEK to DMEK.
- “Harvesting the donor is too complex.” I believe that the rise of DSAEK was directly linked to the ready availability of precut donor tissue. This took the financial risk of damaging the corneal button out of the equation, not to mention the time savings in the OR for the physician and the facility. The anticipated rise in DMEK will similarly be linked to eye banks providing pre-stripped tissue.
- “The dislocation rate is too high.” I personally have been using an inferior iridotomy and sulfur hexafluoride (20%) SF6 gas tamponade in all cases of DSAEK and DMEK, and the re-bubble rates have decreased dramatically.
To be sure, not all patients will be ideal candidates for DMEK. DSAEK will still be the preferred technique in eyes with aphakia, prior vitrectomy, prior failed grafts, blebs and valves, anterior chamber IOLs, peripheral anterior synechiae and aniridia. I therefore do not believe that DSAEK will go the way of RK. But I firmly believe that DMEK is the present and the future.
Endothelial keratoplasty was first performed in the United States in 2000. It is becoming the treatment of choice for endothelial cell dysfunction associated with failed penetrating keratoplasty, Fuchs’ corneal dystrophy and bullous keratopathy. While DSEK has gained significant traction in the U.S., DMEK is just beginning to become more widely used.
For more information:
I do well with DSEK, so do I need to abandon ship?
As Majmudar points out, DMEK is an elegant procedure that can lead to outstanding results. There are reports of better final visual acuity and lower rejection rates already in the literature. Unfortunately, DMEK is not without its own drawbacks, as well. As with any new procedure, there is a learning curve. This learning curve comes with some caveats that other surgeries do not. The donor tissue is fragile, and proper handling is critical. Particularly for surgeons who do not use eye bank-prepared tissue, there is a significant risk of wasting the tissue during preparation. Even eye banks that prepare the donor in advance may have tissue wasted during handling.
For the DSEK procedure, I place a peripheral iridectomy inferiorly, which allows me to leave a complete air fill following surgery. Since starting this, my graft detachment rate is negligible. Obviously, this can be done for DMEK, as well. Unfortunately, graft detachments in DMEK have significantly greater ramifications than for DSEK. With DSEK, the graft can typically be refloated, but with a complete detachment in DMEK, an entirely new piece of tissue may be required.
For my technique, I insert the DSEK graft unfolded on a lens glide with a bed of Healon (sodium hyaluronate, Abbott Medical Optics), pushing the graft with a bent needle rather than folding or using an inserter. I use a 5.5-mm scleral incision. Smaller incisions have been associated with increased endothelial cell damage in the donor graft. By using a scleral incision, I frequently need no sutures and can easily maintain a tight air fill, while using a longer incision. Once inserted, being that the button is already unfolded and in good position, it takes little time to inflate the chamber and center the graft. There is little manipulation of the graft. It is not uncommon for a DMEK procedure to require extensive manipulation, particularly in the early stages of the learning curve. This combination is critical to long-term endothelial cell survival.
With my current technique, the entire procedure typically takes 10 to 15 minutes. Graft detachments are rare, and patients tend to be satisfied. Although there have been reports of lower rejection rates with DMEK, I have found my rejection rate with DSEK to be low, as well. By virtue of having a quick procedure with minimal graft manipulation and low detachment and rejection rates, I have found that DSEK is an excellent procedure for endothelial disease. The benefits of short procedure time, low detachment rates, low risk for wasting tissue or needing new tissue for detachments, and minimal tissue manipulation have limited my concern for long-term endothelial cell survival.
Majmudar compares the transition from DSEK to DMEK to the transition from PRK to LASIK. We have seen in recent years that there is still a place for PRK, and this is actually the treatment of choice for many surgeons. While DMEK is a wonderful procedure that will only gain popularity over time, I do not feel that the corneal surgeon needs to abandon DSEK yet.