Impact Of Host Descemetorhexis On Outcomes Of Dmek Following Penetrating Keratoplasty

Standard endothelial keratoplasty includes descemetorhexis, the removal of host Descemet’s membrane to promote donor graft adherence. However, this step is frequently omitted when performing endothelial keratoplasty after penetrating keratoplasty (PK), due to concerns about disrupting the graft–host junction or uncertainty regarding its benefit. This study evaluates the impact of host descemetorhexis on anatomical and survival outcomes of secondary DMEK following PK.

Tertiary referral corneal service, United Kingdom.

A retrospective audit was conducted of eyes undergoing DMEK after previous PK (2019 -2025). Outcomes included graft failure, rebubbling, logMAR best-corrected visual acuity (BCVA), and central corneal thickness (CCT). Categorical variables were analysed using Fisher’s exact test; continuous variables using independent t-tests or ANOVA. Graft survival was evaluated using Kaplan–Meier analysis with log-rank testing. Cox proportional hazards models were used for univariable and multivariable analysis.

Sixty nine eyes had DMEK after PK (44 with descemetorhexis; 25 without). Sixteen grafts failed in the descemetorhexis group and 17 in the non-descemetorhexis group. Kaplan–Meier analysis showed significantly improved graft survival with descemetorhexis (univariable Cox HR 0.39, 95% CI 0.20–0.78, p=0.008).At 1 year 27.4% eyes undergoing descemetorhexis failed compared with 48.0% eyes without descemetorhexis. At 2 years, failure proportions were 29.9% and 60.0%, respectively. In a reduced multivariable Cox model adjusting for surgeon grade, rebubbling requirement, and number of prior PKs (n=69; 33 events), descemetorhexis remained independently protective (adjusted HR 0.40, 95% CI 0.19–0.81, p=0.011).

In eyes with a previous PK undergoing subsequent DMEK, descemetorhexis was associated with significantly improved graft survival compared to not performing descemetorhexis. The protective effect persisted after multivariable adjustment and appeared stronger beyond the early postoperative period. This simple intraoperative modification may optimise outcomes in DMEK performed after PK and warrants consideration as standard practice.