DMEK outcome after one year – Results from a large multicenter study in Germany

Abstract

Purpose Descemet membrane endothelial keratoplasty (DMEK) accounts for >50% of all corneal transplants in Germany. So far, no data from such a large multicenter study have been published. Methods This retrospective study included 3200 DMEKs at seven departments performed for Fuchs endothelial corneal dystrophy (FECD) or bullous keratopathy (BK). We evaluated best corrected visual acuity (BCVA, logMAR), endothelial cell density (ECD, cells/mm2), minimal corneal thickness (CT, μm), rebubbling‐, primary transplant failure‐ and immune reaction‐rate. Changes over time were evaluated by linear mixed models for repeated measures and correlation with case number by center by weighted linear regression. Results For patients without vision‐limiting comorbidities (74% of all analysed eyes, n = 2270), mean BCVA improved from 0.6 ± 0.4 logMAR to 0.2 ± 0.2 logMAR 6 months (p < 0.001, n = 1441) and 0.1 ± 0.2 logMAR 12 months (p = 0.001, n = 1402) postoperatively. BK‐ had a worse BCVA compared to FECD‐patients (0.3 ± 0.5 vs. 0.1 ± 0.2 logMAR [p < 0.001] at 1 year). ECD declined from 2465 ± 259 cells/mm2 (n = 2876 preoperatively) to 1587 ± 433 cells/mm2 after 12 months (p < 0.001, n = 1237). Mean rebubbling rate was 0.4 ± 0.7/eye. 784 eyes (25%) received at least one rebubbling. More rebubblings correlated with a lower ECD, a worse BCVA, a higher CT, and higher transplant failure and rejection rates (p < 0.001, p = 0.013 for BCVA at 12 months). A single rebubbling did not influence the BCVA (p = 0.785). Graft failure rate was 3% (n = 67), rejection rate 1.5% (n = 48). Conclusion Descemet membrane endothelial keratoplasty increases visual acuity with low transplant failure‐ and rejection‐rates. FECD has a better outcome than BK. Since a quarter of all patients need a rebubbling, this should be included in the informed consent. Remarkably, one rebubbling has no influence on the outcome.

Preserflo MicroShunt: Efficacy and Endothelial Cell Density

“…This is of particular relevance since patients after DMEK and PK tend to have a lower ECD and loose endothelial cells faster than healthy corneas. 14,[22][23][24] In conclusion, our study shows no substantial ECL in a 20-month follow-up with a cohort including eyes with endothelial alteration. Therefore, Preserflo MicroShunt can currently be considered safe even in patients with endothelial cell alteration, for example, after DMEK.…”

“…In their study, 30.4% of patients underwent shunt implantation in combination with phacoemulsification, which is known to cause ECL, requiring lamellar corneal transplantation (DMEK). 14,15 Patients with isolated shunt implantation also showed an ECD reduction of 7.5%. A possible explanation for the different results shown in our study could be the larger distance between the shunt tip and corneal endothelium.…”

“…However, considered a shunt location of <200 µm is a risk factor. In their study, 30.4% of patients underwent shunt implantation in combination with phacoemulsification, which is known to cause ECL, requiring lamellar corneal transplantation (DMEK) 14,15 . Patients with isolated shunt implantation also showed an ECD reduction of 7.5%.…”

“…The cohort evaluated in this study provides additional aspects with regard to corneal safety, as 3 eyes had previously undergone keratoplasty, in which implantation of the Preserflo MicroShunt was not only technically feasible but also did not result in an enhanced ECL. This is of particular relevance since patients after DMEK and PK tend to have a lower ECD and loose endothelial cells faster than healthy corneas 14,22–24 …”

Descemet Membrane Endothelial Keratoplasty (DMEK) Anesthesia Analysis

“…Currently, it is the most commonly used technique for corneal transplantation in Germany and is also increasingly favored over DSAEK (Descemetʼs stripping automated endothelial keratoplasty) in the USA, as it is the preferred method for treating endothelial diseases such as Fuchs endothelial dystrophy (FED) or bullous keratopathy (BK) 3 , 4 , 5 , 6 , 7 .…”