Corneas from donors with diabetes did as well in DMEK, whatever the donor's HbA1c

Design

Secondary analysis of a multicentre, double-masked randomised clinical trial

Population

1,421 eyes of 1,097 adults having low- to moderate-risk DMEK at 28 US sites

Primary outcome

One-year graft success and endothelial cell outcomes by donor diabetes severity

Effect

Graft success 97.1% (HbA1c <6.5%) vs 93.0-97.4% in higher groups; P = 0.72

The Diabetes Endothelial Keratoplasty Study was a double-masked US trial at 28 sites in which donor corneas with or without diabetes were allocated, by a minimisation method comparable to randomisation, for low- to moderate-risk Descemet membrane endothelial keratoplasty (DMEK). This secondary analysis (JAMA Ophthalmology, 8 October 2026) asked whether the severity of donor diabetes mattered.

Of 1,421 eyes in 1,097 recipients, 95% were operated on for Fuchs dystrophy. One-year graft success was 97.1% with donor HbA1c below 6.5%, and 97.4%, 93.0% and 95.8% in the 6.5-7.4%, 7.5-8.9% and 9.0% or higher groups (P = 0.72). Endothelial cell density, cell-size variation and hexagonality did not differ by donor HbA1c, diabetes severity score or advanced glycation and oxidation markers.

The authors conclude that corneas from donors with diabetes can be used for endothelial keratoplasty without extra testing or stratification by diabetes severity. That matters for supply: diabetes is common among donors, and excluding their tissue shrinks the pool.

• Accept corneas from donors with diabetes for low- to moderate-risk DMEK; one-year outcomes matched non-diabetic tissue.
• Do not request donor HbA1c or glycation testing to select endothelial keratoplasty tissue.
• Use usual eye bank endothelial criteria for tissue selection.
• Recognise that the highest-HbA1c groups were small, so very poorly controlled donor diabetes is less well studied.

Why it matters

It removes a reason some surgeons and eye banks have used to decline tissue, widening the donor pool.

Don't overread it

Results are at one year and in low- to moderate-risk eyes; longer-term graft survival was not reported here.

The statistics, in plain English

The confidence intervals for the higher-HbA1c groups are wide because they included 48 to 194 eyes; P = 0.72 means no detectable trend across groups.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.