Factors Affecting the Success Rate of Preloaded Descemet Membrane Endothelial Keratoplasty With Endothelium-Inward Technique: A Multicenter Clinical Study


  • Preloaded Descemet membrane endothelial keratoplasty (DMEK) with endothelium-inward must be preserved for less than 4 days to avoid endothelial cell loss and early postoperative complications.

  • Rebubbling rate following preloaded DMEK grafts is multifactorial.


To evaluate factors affecting the outcomes of preloaded Descemet membrane endothelial keratoplasty (pl-DMEK) with endothelium-inward.


Retrospective clinical case series and a comparative tissue preparation study.


Participants : Fifty-five donor tissues for ex vivo study and 147 eyes of 147 patients indicated with Fuchs endothelial dystrophy or pseudophakic bullous keratopathy with or without cataract.

Intervention : Standardized DMEK peeling was performed with 9.5-mm-diameter trephination followed by second trephination for loading the graft (8.0-9.5 mm diameter). The tissues were manually preloaded with endothelium-inward and preserved for 4 days or shipped for transplantation. Live and dead assay and immunostaining was performed on ex vivo tissues. For the clinical study, the tissues were delivered using bimanual pull-through technique followed by air tamponade at all the centers.

Main outcome measures : Tissue characteristics, donor and recipient factors, rebubbling rate, endothelial cell loss (ECL), and corrected distance visual acuity (CDVA) at 3, 6, and 12 months.


At day 4, significant cell loss ( P = .04) was observed in pl-DMEK with loss of biomarker expression seen in prestripped and pl-DMEK tissues. Rebubbling was observed in 40.24% cases. Average ECL at 3, 6, and 12 months was 45.87%, 40.98%, and 47.54%, respectively. CDVA improved significantly at 3 months postoperation (0.23 ± 0.37 logMAR) ( P < .01) compared to the baseline (0.79 ± 0.61 logMAR). A significant association ( P < .05) between graft diameter, preservation time, recipient gender, gender mismatch, and recipient age to rebubbling rate was observed.


Graft loading to delivery time of pl-DMEK tissues in endothelium-inward fashion must be limited to 4 days after processing. Rebubbling rate and overall surgical outcomes following preloaded DMEK can be multifactorial and center-specific.

I t is widely known that Descemet membrane endothelial keratoplasty (DMEK) supports faster visual rehabilitation and improved visual outcomes , with long-term graft survival rates. Given the challenges that this technique has introduced, it is undeniable that there is a steep learning curve to achieve best results both in terms of graft preparation and delivery. Standardization of DMEK for graft preparation and delivery has therefore been a key area of interest. Performing DMEK, from preparation to delivery, requires skilled maneuvers increasing the overall surgical time and higher chances of tissue wastage, thus making it a relatively complicated surgery. We, therefore, described the preparation and use of preloaded DMEK (pl-DMEK) tissues , with an aim at reducing the efforts, time, and tissue wastage in addition to providing a quality-certified graft to the surgeon.

We also noted that folding a DMEK graft with endothelium-inward and preloading it in an intraocular lens cartridge could significantly reduce the surgical time. , It has been previously noted that graft detachment and rebubbling rate is one of the routine postoperative complications noticed following DMEK surgery. , However, we have observed that factors such as donor characteristics, graft preparation and preservation time, delivery methods, and surgical and postoperative complications could all contribute to a successful DMEK surgery. Being a relatively new technique and with limited available data, early postoperative complications may have been overlooked following a pl-DMEK. Therefore, we aim at investigating the efficacy and factors related to success of pl-DMEK with endothelium-inward prepared by a single eye bank in a retrospective multicenter clinical study.



Tissues for research (n = 55) and clinical purpose (n = 147) were obtained with written consent from the donors’ next of kin. The tissues were deemed suitable for research only because of poor endothelial cell count (typically <2000 cells/mm 2 ). All the pl-DMEK grafts with endothelium-inward were prepared by the eye bank technicians at Fondazione Banca degli Occhi del Veneto Onlus, Italy.

For the retrospective clinical case series, records from the patients treated for endothelial dysfunction with a pl-DMEK between March 2017 and January 2020 were included. Royal Liverpool University Hospital (Liverpool, United Kingdom), SS Giovanni e Paolo hospital (Venice, Italy), San Bortolo Hospital (Vicenza, Italy), and University Hospital of Verona (Verona, Italy) received pl-DMEK tissues. The tenets of the Declaration of Helsinki were followed with informed consent for surgery as part of routine clinical care. The retrospective data collection was approved by the respective local institutional review boards.


For ex vivo study, prestripped tissues immediately stained for live and dead assay (n = 3) and immunostaining (n = 2 for ZO-1 and n = 2 for Na + /K + -ATPase) were used as controls. Prestripped and pl-DMEK tissues preserved for day 1, 2, 3, and 4 were stained for live and dead assay (n = 32) and immunostaining (n = 4 at day 1 and n = 4 at day 4 for ZO-1 and; n = 4 at day 1 and n = 4 at day 4 for Na + /K + -ATPase) for comparative analysis. For clinical study, data from 147 eyes of 147 patients were collected.


The pl-DMEK with endothelium-inward were prepared using the technique described earlier with minor modifications. , The endothelial cells and mortality were counted using trypan blue stain under light microscope. In brief, the Descemet membrane-endothelial complex was peeled (9.5 mm diameter) from the corneoscleral rim ( Figure 1 , A) leaving a hinge and marked with an “F” stamp ( Figure 1 , B).


Preparation of a preloaded DMEK graft with endothelium-inward technique. A. Peeling of Descemet membrane-endothelial complex from the corneoscleral rim. B. Marking with “F” stamp. C. DMEK tissue trephined to a desired diameter (8.0-9.5 mm) and manually trifolded on the stroma. D. Trifolded membrane gently pulled from the stroma using an aluminium foil. E. A micro glider’s preservation chamber (funnel) using a Tweezer Grip 23G from the funnel opening. F. Sealing the rear end with a silicone plug. G. The membrane within the cartridge placed inside an I-Glide transport container. H. Filled with the organ culture media supplemented with dextran and left for 1, 2, 3 or 4 days at RT. DMEK = Descemet membrane endothelial keratoplasty.


The tissues prestripped with a hinge anchored to the cornea were repositioned back onto the stroma. The stamped DMEK tissues were retrephined (8.25 mm) and stained (controls at day 0) immediately or preserved in organ culture media supplemented with dextran for 1, 2, 3, and 4 days at room temperature (RT).

For pl-DMEK, the tissues were manually trifolded on the stroma ( Figure 1 , C) and gently transferred from the tissue base using an aluminium foil ( Figure 1 , D) into a micro glider’s (00267 Micro B, MDJ SAS; La Monnerie) preservation chamber and pulled inside the funnel using a Tweezer Grip 23G (AK-1670G; Aktive S.R.L.) from the funnel opening ( Figure 1 , E). The funnel was prefilled with organ culture media supplemented with dextran. Once the membrane was inserted inside the cartridge, the rear end was sealed with a silicone plug ( Figure 1 , F). The membrane within the cartridge was placed inside an I-Glide transport container (CMXGLIDE-UN; Eurobio Scientific) ( Figure 1 , G) filled with the same medium and left for 1, 2, 3, or 4 days at RT ( Figure 1 , H). Donor characteristics such as age, gender, endothelial cell density (ECD), and tissue preservation time were recorded.


All the tissues (prestripped/preloaded) were gently pulled out from the cartridge using a 23G tweezer grip and placed on a glass slide with the endothelium facing upward. The tissues were stained with LIVE/DEAD Viability/Cytotoxicity Kit (Molecular Probes, Invitrogen; cat. 03224) following manufacturer’s protocol.

In brief, the tissues were washed twice in phosphate-buffered saline and incubated for 40 minutes at RT with 250 µL of 2 µM Calcein AM and 4 µM ethidium homodimer-1 in phosphate-buffered saline. A relaxing cut was made on the membrane allowing a flat mount on the glass slide. Slides were sealed with 4′,6-diamidino-2-phenylindole (DAPI) Fluoromount-G (Electron Microscopy Sciences; 17984-24) and analyzed at 10 × magnification with a Nikon Eclipse T i fluorescence microscope. Acquired images were imported into the FIJI open software ( imagej.net/ ) for further analysis. The images were analyzed using WEKA (Waikato Environment for Knowledge Analysis) Segmentation procedure as described in our earlier publication.


The prestripped and pl-DMEK tissues collected at different time points were fixed in 4% paraformaldehyde (30 minutes at RT) and permeabilized using 0.5% Triton-X 100 (30 minutes at RT). The tissues were blocked in goat serum (10%) for 1 hour at RT and incubated with primary antibodies (ZO-1 rabbit polyclonal antibody [40-2200, dilution 1:100 in blocking buffer; Invitrogen] and Na + /K + -ATPase α1 mouse monoclonal antibody [sc-21712, dilution 1:200 in blocking buffer; Santa Cruz Biotechnology]) overnight at 4°C followed by incubation with secondary antibody (Alexa Fluor 488 goat antirabbit [A11008, dilution 1:500 in blocking buffer; Invitrogen and Rhodamine Red goat antimouse IgG [R6393, dilution 1:500 in blocking buffer; Invitrogen]) for 1 hour at room temperature. The flat-mounted tissues following relaxing cuts were sealed using DAPI Fluoromount-G and cover slips followed by analyses at 400 × magnification with a Nikon Eclipse T i fluorescence microscope.


At all the centers, the patients were locally anesthetized by peribulbar injection. The main entrance using the corneoscleral tunnel (2.8 mm width) or clear cornea approach was created. The cartridge with pl-DMEK tissue (8.0-9.5 mm diameter based on surgeons’ request) was removed and unplugged at the front end. The medium was washed out gently with balanced salt solution using a 25G needle mounted to a 1-mL syringe from the front end ensuring the architecture (endo-in) of the graft is maintained. The presence of a rear plug ensured that the graft did not aspirate out of the funnel. Tissues were delivered using bimanual pull-through technique , followed by gentle tapping on the cornea to unfold the graft. The tissues were attached following air tamponade in the anterior chamber and the tunnel was sutured.

Antibiotics were administered subconjunctivally at the early postoperative days. The grafts were rebubbled immediately in the events when the tissue showed detachment or only >30% detachment or involvement of the pupillary area (center-specific). Data such as recipient gender, primary diagnosis, ECD during postoperative care, surgery details (graft diameter and combination with phacoemulsification), corrected distance visual acuity (CDVA), and rebubbling procedures were collected.


Statistical analysis was performed using Stata, version 14.0 (StataCorp), and a P value of less than .05 was considered statistically significant. Quantitative variables were tested for normality using the Shapiro-Wilk test. Two-sample t test, Mann-Whitney U test or 1-way analysis of variance followed by post hoc Tukey test were used with 95% CI according to data distribution. Paired tests were used to compare pre- and postoperative variables within the same group.



Fifty-five human corneas were obtained from 30 donors (mean age: 70 ±7 years; range: 50-76 years). Prior to graft preparation, the average storage time of corneas was 14 (±4) days (range 6-20 days). Postmortem time (time from death to procurement) was <30 hours. Cell viability was observed in all groups at all the time points following live and dead assay ( Figure 2 , A). The control tissues, which were prestripped and evaluated immediately at day 0, showed 73.8% (±7.5%) viable cells. Statistical difference was not observed between prestripped and pl-DMEK tissues at all time points ( P = .1377; Figure 2 , B).


Cell viability analysis using live and dead assay. A. Control group stained immediately after peeling. B. Comparative analysis using percentage viability of prestripped and pl-DMEK tissues preserved from day 0 to day 4. C. Global view of the viability staining on prestripped and pl-DMEK tissues preserved from day 0 to day 4 using calcein-AM (green) and ethidium homodimer (red) stain.

Prestripped tissues did not show significant difference in cell viability between any time point ( P = .5099; Figure 2 , C); however, cell viability dropped significantly in the pl-DMEK group ( P = .0401; Figure 2 , D) between day 0 (control) and day 4. The data at respective time points are shown in the table 1 . Tight- and gap-junction biomarkers were well preserved when the tissues were peeled and immediately stained ( Figure 3 , A). However, loss of ZO-1 as well as Na + /K + -ATPase markers was observed at day 4 both in prestripped and pl-DMEK groups compared with day 0 and day 1 ( Figure 3 , A through E).

Sep 11, 2022 | Posted by in OPHTHALMOLOGY | Comments Off on Factors Affecting the Success Rate of Preloaded Descemet Membrane Endothelial Keratoplasty With Endothelium-Inward Technique: A Multicenter Clinical Study

Full access? Get Clinical Tree

Get Clinical Tree app for offline access