Research supports use of corneal tissue from donors with diabetes for DMEK

medwireNews: Diabetes status among cornea donors does not impact the success of Descemet membrane endothelial keratoplasty (DMEK), or the likelihood of postsurgery endothelial cell loss (ECL) and morphometric changes, show primary and secondary analyses of the Diabetes Endothelial Keratoplasty Study (DEKS).

The two sets of findings, which were simultaneously published in JAMA Ophthalmology, “provide strong support for having no restrictions on the use of tissue from donors with diabetes for DMEK,” write Jonathan Lass (University Hospitals Cleveland Medical Center, Ohio, USA) and colleagues who co-authored both studies.

Impact of cornea donor diabetes status on 1-year DMEK success rate

The primary analysis included data for 1097 individuals (1421 study eyes; median age 71 years; 58% women) with a clinical indication for low to moderate risk DMEK due to endothelial dysfunction (95% for Fuchs endothelial corneal dystrophy [FECD]). They were randomly assigned to receive a donor cornea from people without (64.2%) or with (35.8%) diabetes.

The researchers report that, at 1 year, the cumulative probability of graft success was 96.3% among the study eyes given tissue from donors without diabetes and 97.1% for those receiving tissue from donors with diabetes, resulting in a nonsignificant difference of 0.7 percentage points. Graft failure was defined as the need for cornea regrafting for any reason or having a recipient cornea classified as cloudy, without clearing, based on a standardized grading scale.

The results were similar when the corneal donors were stratified by diabetes severity. Specifically, the 1-year cumulative probability of graft success was 96.5% in the subgroup that received corneas from donors with mild diabetes (n=173) and 97.3% when the donors had moderate to severe diabetes severity (n=336).

The rates of primary donor failure, early failure related to surgical complications, and subsequent failure were 2.5%, 0.7%, and 0.3%, respectively, in recipients of tissue from a donor without diabetes. The corresponding rates among recipients of tissue from a donor with diabetes were 2.6%, 0.4%, and 0.0%.

There were no failures due to graft rejection and there was no significant impact of recipient age, indication for keratoplasty, or recipient diabetes status on the 1-year graft success rate.

However, the investigators note that DMEK graft preparation was less likely to fail in donors without diabetes versus those with diabetes (1.6 vs 6.1%), which they say represents “a small but potentially relevant difference with respect to eye bank costs in procuring corneas from donors with diabetes.”

Endothelial cell loss and morphometric changes at 1 year after successful DMEK

In the secondary analysis, Lass and team investigated the impact of donor diabetes status on ECL and morphometric changes at 1 year after successful DMEK using data for 1274 eyes of 982 DEKS participants (mean age 70 years; 58% women).

Prior to surgery, mean central endothelial cell density (ECD) was 2676 cells/mm² in tissue from donors without diabetes and 2671 cells/mm² in that from donors with diabetes.

At 1 year, the researchers found that mean ECL was 28.3% and 28.0% in the donor groups without and with diabetes, respectively, resulting in a corresponding mean 1-year ECD of 1927 cells/mm² and 1920 cells/mm², with no significant difference between the two groups.

In both groups, the majority of ECL occurred within the first month after DMEK. At this time, mean ECL was 23.5% in the group without diabetes and 23.8% in the group with diabetes. For the remainder of the follow-up period, both groups experienced an additional ECL of approximately 5%.

There were no differences in ECD according to donor diabetes severity score, or the presence of concurrent donor and recipient diabetes.

There were also no significant morphologic differences between the tissue from donors without diabetes and that from donors with diabetes. Specifically, the mean coefficient of variation in cell area at 1 year was 31.5% in the group given tissue from donors without diabetes and 31.4% in the group with diabetes, while the mean percentage of hexagonal cells was 57.7% and 57.2%, respectively.

In an accompanying comment, Kathryn Colby and Andrea Blitzer, both from New York University Langone Health in the USA, say the study “implications are clear.”

They write: “The results of the DEKS should reassure eye banks and surgeons that a donor having diabetes does not compromise DMEK success through at least 1 year among patients with FECD.”

They acknowledge that although graft preparation failure rates were higher for donors with diabetes, “eye banks can confidently use tissue from donors with diabetes for EK, and surgeons should incorporate the DEKS findings when considering their specific preferences.”

Colby and Blitzer say: “This expands the donor pool at a critical time, as US Food and Drug Administration recommendations may further restrict supply by excluding tissue with broad risk factors for sepsis and tuberculosis.

“Given that nearly one-third of the corneal tissue recovered by US eye banks is exported to international recipients, adopting this level 1 evidence may help to ease the global shortage of donated corneal tissue.”

medwireNews is an independent medical news service provided by Springer Healthcare Ltd. © 2025 Springer Healthcare Ltd, part of Springer Nature

JAMA Ophthalmol 2025; doi:10.1001/jamaophthalmol.2025.4253
JAMA Ophthalmol 2025; doi:10.1001/jamaophthalmol.2025.4261
JAMA Ophthalmol 2025; doi:10.1001/jamaophthalmol.2025.4254