DMEK trial opens the door to a bigger cornea donor pool

A donor cornea’s diabetes status does not need to keep it out of the transplant pool. The Diabetes Endothelial Keratoplasty Study (DEKS), a multicenter, randomized trial (NCT05134480), found that corneas from donors with diabetes achieve 1-year outcomes in Descemet membrane endothelial keratoplasty (DMEK) comparable to those from nondiabetic donors.¹

The study enrolled 1421 eyes from approximately 1100 donors across 28 clinical sites and 13 participating eye banks, with all clinical sites, surgeons, patients, and technicians preparing donor tissue masked to donor diabetes status.¹ The study was named an Editor’s Choice: Clinical Trial of 2025 by JAMA Ophthalmology.² One-year results, published in 2 papers in the journal, showed comparable graft success and endothelial cell loss between diabetic and nondiabetic donors.^1,3 The research team is now reenrolling patients for 3- and 5-year follow-up data, funded by the National Eye Institute (NEI), with results expected by approximately 2030. Jonathan H. Lass, MD, of Case Western Reserve University School of Medicine and the University Hospitals Eye Institute in Cleveland, Ohio, discussed the findings—and what comes next—with Ophthalmology Times (OT).

Note: Transcript edited lightly for clarity and length.

OT: What does this recognition mean to you and the broader research team behind the study?

Jonathan H. Lass, MD: I think it’s a recognition of the importance of the question that we studied. Diabetes is incredibly common among the population. Probably at least 20 million people are affected by the disease, and many are undiagnosed. This is affecting the pool of donors that eye banks have. People with diabetes tend to die more frequently than those without diabetes, so the donor pool of corneas is now represented by between 30% and 40% of all the donors going for corneal transplants having diabetes.¹

The NEI recognized, when we applied for this grant, that the question of which corneas from donors with diabetes can be used, and which cannot, needed to be answered. Many eye banks had decided to avoid using donors with diabetes. The trial was [designed to assess] whether donors with diabetes could be used. We were very excited that when the results came out last fall at the American Academy of Ophthalmology meeting in October, we were able to present the data, with 2 papers also coming out in JAMA Ophthalmology.^1,3 With all the great clinical trials that JAMA Ophthalmology publishes, they recognized the importance of our results, and it was selected as the Trial of the Year.²

OT: Your study challenges a long-standing concern in corneal transplantation, that donor corneas from people with diabetes may be less suitable for DMEK surgery. What motivated your team to test that assumption in such a large randomized clinical trial?

Lass: There were concerns about the effect of the donor pool. DMEK has become the most common endothelial keratoplasty procedure, yet eye banks were having to avoid using donors with diabetes. The concern was not only the long-term effects but also preparation of the donor tissue. A study done more than 10 years ago showed that the preparation failure rate for preparing the lenticule for transplant was maybe 3 times as high in diabetic donors, which is why many eye banks avoided using them.

We put together a trial to look at both the long-term effects, at least out to 1 year, and the preparation failure question. Two-thirds of the donors did not have diabetes, and one-third did. All the clinical sites, surgeons, and patients and even the technician preparing the donor tissue were masked as to the diabetes status of the donor. The study was powered to detect about a 5% difference in the failure rate at 1 year between donors with diabetes and without. We enrolled 1421 eyes from approximately 1100 donors, in about 1100 patients, across 28 clinical sites and 13 participating eye banks.¹

What we found was somewhat surprising: The success rate was incredibly similar across groups. It was 96% in the nondiabetic donors at 1 year and 97% in the diabetic donors—exactly the same.¹ We also looked at endothelial cell loss for the first time very early on, at 1 month post transplant, and found the loss was the same in both groups: 24% at 1 month and only another 4% by 1 year, for 28% total.³ At least at 1 year, diabetic donors did as well as nondiabetic donors.

The one difference we did find was in donor preparation: The failure rate was about 6% in diabetic donors compared with 1% to 2% in nondiabetic donors.¹ That’s actually much better than 10 years ago, when it was closer to 15% for diabetic donors. We believe this is more of an adhesion issue between the Descemet membrane and the stroma during stripping; diabetes appears to affect that basement membrane, making the tissue more fragile during preparation. But once the donor tissue is successfully prepared, the endothelial cells behave similarly. We also looked at whether diabetes severity mattered—participants with mild diabetes on oral medication only vs those with more severe diabetes with retinopathy or kidney disease—and severity did not appear to affect cell loss or graft survival at 1 year.^1,3

OT: How could these findings change clinical practice for corneal surgeons, eye banks, and patients waiting for transplants?

Lass: At least with the 1-year data, we hope we can convince both surgeons and eye banks that the entire spectrum of donor diabetes—if they’re willing to accept a slightly greater risk of preparation failure—performs as well as nondiabetic donor tissue. We’re hoping this will expand the donor pool, particularly among eye banks that were avoiding diabetic donors, now that we have strong scientific evidence that diabetes status didn’t make a difference at 1 year.¹

We were able to convince the NEI to look at the donor diabetes question even further. We were funded last year to follow a subset of patients out to 5 years, so we’ll be looking at 5-year cell loss and graft survival. We’re currently reenrolling patients to gather 3-year and 5-year data, with results expected around 2030.

OT: As your NEI-funded follow-up studies continue, what are the next big questions you hope to answer about long-term graft survival and outcomes?

Lass: We’re basically going to be doing the same thing we did at 1 year, in terms of graft survival and cell loss. We’re also adding another component: looking at the genetic influence of Fuchs dystrophy and diabetes on outcomes. Fuchs dystrophy was the most common reason for transplant in this study; 95% of all recipients had it.¹ We’re going to study the genetics of both the donor and recipient for diabetes and Fuchs dystrophy and develop what is called a polygenic risk score and relate that score to graft success and cell loss outcomes, similar to how risk scores are used in areas like breast cancer research.

Jonathan H. Lass, MD

E: [email protected]

Lass is a professor in the Department of Ophthalmology and Visual Sciences at Case Western Reserve University School of Medicine and is affiliated with the University Hospitals Eye Institute in Cleveland, Ohio. He served as study chair of the Diabetes Endothelial Keratoplasty Study.

References

1. Price FW Jr, Szczotka-Flynn LB, Price MO, et al. Donor diabetes and 1-year Descemet membrane endothelial keratoplasty success rate: a randomized clinical trial. JAMA Ophthalmol. 2025;143(12):1043-1051. doi:10.1001/jamaophthalmol.2025.4253

2. Research by ophthalmology’s Jonathan Lass selected as Clinical Trial of the Year. News release. Case Western Reserve University. May 1, 2026. Accessed July 13, 2026. https://case.edu/news/research-ophthalmologys-jonathan-lass-selected-clinical-trial-year

3. Lass JH, Benetz BA, Verdier DD, et al. Endothelial cell loss 1 year after successful DMEK in the Diabetes Endothelial Keratoplasty Study: a randomized clinical trial. JAMA Ophthalmol. 2025;143(12):1053-1060. doi:10.1001/jamaophthalmol.2025.4261