Epithelial Thickness Changes After Descemet Membrane Endothelial Keratoplasty (DMEK): An Observational Study

Aims: The aim of this study is to characterise corneal epithelial thickness profiles after Descemet membrane endothelial keratoplasty (DMEK) and compare it with healthy controls, focusing on inferior-superior (I-S) epithelial thickness differences and their relationship with age. Methods: This single-centre observational study included 36 post-DMEK eyes with at least 6 months' follow-up and 36 healthy control eyes. High-resolution spectral-domain anterior segment OCT maps were analysed for central epithelial thickness (CET, defined as the mean epithelial thickness within the central 2 mm zone [E2.0]) and peripheral sectors to derive inferior (E-I) and superior (E-S) values (between 2 and 7 mm), with the I-S difference computed at a 3 mm radius; group comparisons used t-tests and correlations used Pearson's r (α = 0.05). Central corneal thickness (CCT) was also compared between groups. Results: Post-DMEK eyes had significantly lower mean CCT than controls (525.7 ± 98.4 μm vs. 544.71 ± 27.8 μm, p = 0.04). Central epithelial thickness did not differ between groups (post-DMEK 53.7 ± 5.5 μm vs. controls 52.7 ± 3.3 μm, p = 0.62), but the I-S epithelial difference was greater after DMEK (5.9 ± 4.3 μm) than controls (3.0 ± 2.2 μm, p < 0.01), indicating a more pronounced inferior thickening pattern. Age showed no significant relationship with epithelial thickness in controls, and only very weak or non-significant correlations with central thickness and I-S difference in post-DMEK eyes, indicating no clinically meaningful age effect postoperatively. Conclusions: DMEK restores central epithelial thickness to values comparable to normal eyes, while accentuating the physiologic inferior-superior epithelial gradient, consistent with localised postoperative epithelial remodelling rather than global epithelial thickening or thinning. Corneal stromal remodelling may result in lower CCT post-DMEK versus controls, and age does not meaningfully influence epithelial distribution after surgery.