Ability of routinely collected clinical factors to predict good visual results after primary Descemet membrane endothelial keratoplasty: a cohort study

Background: A comprehensive analysis of routinely collected pre/perioperative demographic/clinical factors that could predict final visual acuity after primary Descemet membrane endothelial keratoplasty (DMEK) has not been conducted previously.

Methods: A retrospective monocenter cohort study was performed with consecutive patients with Fuchs endothelial corneal dystrophy (FECD) who underwent DMEK or triple-DMEK (DMEK combined with cataract surgery) in 2016-2020 in a French tertiary-care hospital. DMEK-only patients were pseudophakic. Patients were followed for 12 months. Surgery was considered successful when 12-month best-corrected visual acuity (BCVA) was ≤0.1 logMAR (≥0.8). Exploratory multivariate analysis was conducted with the following routinely collected variables to determine their ability to predict 12-month BCVA: patient age and sex; graft donor age; triple DMEK; preoperative values of BCVA, endothelial cell density (ECD), central corneal thickness (CCT), and mean anterior keratometry; and rebubbling.

Results: Of 100 eyes (100 patients; mean age, 72 years; 61% female), 81 achieved a 12-month BCVA of ≤0.1 logMAR. Logistic regression analysis showed that older age was a significant prognosticator for 12-month BCVA > 0.1 logMAR (Odds Ratio = 0.914, 95% confidence intervals = 0.846-0.987; p = 0.02).

Conclusions: An older age associated with worse visual acuity outcomes after DMEK. This was confirmed by our analysis of the literature and supports the notion that DMEK should be conducted without delay once symptoms appear. Patient sex, donor age, triple-DMEK, and anterior keratometry also did not predict final BCVA in the literature. Preoperative CCT, ECD, and BCVA, and rebubbling occasionally appear in the literature as BCVA predictors, possibly reflecting an underlying ECD-BCVA axis.