ESCRS - FP14.08 - Comparison Of Endo-In Technique Vs Endo-Out-Double-Bubble Technique In Descemet’S Membrane Endothelial Keratoplasty (Dmek) For Vitrectomized Eyes

Comparison Of Endo-In Technique Vs Endo-Out-Double-Bubble Technique In Descemet’S Membrane Endothelial Keratoplasty (Dmek) For Vitrectomized Eyes

Published 2026 - 44th Congress of the ESCRS

Reference: FP14.08 | Type: Free paper | DOI: 10.82333/6b0r-cz41

Authors: Dr Kyohei Fujiwara* 1 , Dr HIdeaki Yokogawa 2 , Dr Ami Igarashi 1 , Dr Akira Kobayashi 2 , Dr Tomomi Higashide 2 , Dr Satoru Yamagami 1 , Dr Takahiko Hayashi 1

1Division of Ophthalmology, Department of Visual Sciences,Nihon University School of Medicine,Tokyo,Japan, 2Department of Ophthalmology, Kanazawa University Hospital,Kanazawa University Graduate School of Medical Science,Kanazawa,Japan

Purpose

To compare the clinical outcomes of Descemet’s membrane endothelial keratoplasty (DMEK) for vitrectomized eyes using endothelium-in pull-through DMEK (ENDO-IN group) versus the endothelium-out-double-bubble technique (ENDO-OUT group).

Setting

Retrospective, comparative study conducted at two tertiary referral centers in Japan.

Methods

Consecutive vitrectomized eyes that underwent DMEK for endothelial dysfunction between August 2022 and October 2024 were reviewed and assigned to one of two groups, based on the surgical technique: patients in the ENDO-IN group underwent endothelium-in pull-through DMEK using EndoGlide, and patients in the ENDO-OUT group underwent DMEK using the double-bubble technique. Outcome measures, including best spectacle-corrected visual acuity, central corneal thickness, endothelial cell density (ECD), and intra- and post-operative complications, such as re-bubbling, were analyzed.

Results

Twenty eyes of 20 patients (mean age 68.5 ± 17.9 years; mean follow-up 5.5 ± 3.4 months) were included, with no instances of primary graft failure and no significant differences in baseline characteristics between groups (p > 0.05). The ENDO-IN group had a lower re-bubbling rate, with no eyes requiring re-bubbling, and less endothelial cell loss than the ENDO-OUT group (25.8 ± 15.2% versus 38.8 ± 16.3%), although the difference was not statistically significant (p = 0.099). Overall, post-operative ECD was significantly lower compared with donor ECD (p < 0.001).

Conclusions

Endothelium-in pull-through DMEK is feasible in vitrectomized eyes, achieving stable attachment with visual and anatomical improvement even when adjunctive maneuvers are required, demonstrating clinical utility in post-vitrectomy corneal edema.