Surgical Technique

All donor tissue used was stored in corneal storage solution (Optisol; Bausch & Lomb, Rochester, New York, USA) and received from the Eye Bank of Canada, Ontario division. The mean donor age was 66.8 ± 4.8 years. All procedures were performed by a single experienced corneal surgeon (D.S.R.) and were not among the first 50 DMEK surgeries performed by him. DMEK grafts were prepared as previously described. Graft preparation was done according to the modified Melles technique using an “F” marking through a scleral window.

Our DMEK technique has been described previously. Briefly, descemetorhexis size was marked on the cornea, and 2 limbal paracenteses were performed at 2 and 10 o’clock. A temporal 2.4-mm clear corneal incision was performed. An anterior chamber maintainer was inserted inferotemporally into the anterior chamber. In previously vitrectomized eyes, a pars plana infusion was used to better control anterior chamber depth. A descemetorhexis was created using a reverse Sinskey hook under balanced salt solution (BSS) infusion followed by removal of the recipient Descemet membrane. Vision Blue (D.O.R.C., Zuidland, Netherlands) was injected into the anterior chamber to ensure complete removal of Descemet membrane remnants. The 7 mm-9 mm donor Descemet membrane (size was chosen according to recipient white-to-white measurements) was loaded into either a glass pipet (Geuder AG, Heidelberg, Germany) or an intraocular lens injector (Monarch D, Alcon Labs Inc, Fort Worth, Texas, USA), and injected into the anterior chamber through the clear corneal incision. The anterior chamber infusion was turned on and off as needed to keep the anterior chamber shallow but was removed after injection of the donor tissue into the anterior chamber. The tapping technique was used to unfold and position the graft, and the anterior chamber was then filled with air. BSS was injected into the anterior chamber, between the air bubble and the iris, to reduce the air bubble size up to a diameter slightly larger than that of the graft. No peripheral iridectomy was performed. Cyclopentolate hydrochloride 1% (MINIMS Cyc 1.0; Chauvin Pharmaceuticals Ltd, UK) and phenylephrine hydrochloride 2.5% (MINIMS PHNL 2.5; Chauvin Pharmaceuticals Ltd) were instilled to prevent pupillary block.

All patients remained supine for 2 hours and were then instructed to remain so as much as possible at home until the next morning. All patients were examined 2 hours and 1 day after surgery. Four patients required air release due to either elevated intraocular pressure (IOP) or a total anterior chamber air fill. All eyes underwent pressure patching overnight. The following day, 0.1% dexamethasone sodium phosphate and 0.3% tobramycin antibiotic (Tobradex; Alcon, Mississauga, Ontario, Canada) eye drops were administered 4 times daily for 1 week. Then, the antibiotic steroid drops were discontinued and 0.1% dexamethasone sodium phosphate (Maxidex; Alcon Labs Inc) eye drops were tapered down from 4 times daily to once daily over a 3-month period and continued once daily thereafter for a prolonged period of time. Hypotensive drops were maintained as preoperatively and changed according to clinical indication. Patients were examined at 1 week, 1 month, quarterly for the first postoperative year, semiannually for the second postoperative year, and annually thereafter.

Graft detachment was defined as any total or partial separation of the graft from the host cornea. Rebubbling was performed within 24 hours in eyes with Descemet membrane detachment of more than one-third of the DMEK graft area if no air bubble was left in the anterior chamber. Rebubbling was also performed 7-45 days postoperatively if there was unresolved Descemet membrane detachment that was causing persistent corneal edema either limiting rapid visual recovery or causing significant ocular surface discomfort. In cases of uncertainty, anterior segment optical coherence tomography (Optovue, Fremont, California, USA) was performed to determine whether there was graft detachment. Primary graft failure was defined as persistent, nonclearing corneal edema 2 months after DMEK. Secondary graft failure was defined as corneal decompensation after an initially functional DMEK graft. Endothelial graft rejection was defined as the presence of inflammation as evidenced by anterior chamber cells, keratic precipitates or endothelial rejection line, or the presence of corneal edema with conjunctival injection and symptoms of pain or light sensitivity.

Study Outcomes

Primary outcomes included graft survival and graft rejection rates. Secondary outcomes included rates of detachment/rebubble, visual acuity, endothelial cell loss, IOP, and glaucoma medications/surgeries.

Data Collection and Statistical Analysis

The data collected in this study included patient demographics, best spectacle-corrected visual acuity (BSCVA), associated operative procedures (including details on timing and indication for glaucoma surgeries), IOP, number of hypotensive ocular medications, intraoperative and postoperative complications, corneal donor characteristics, and endothelial cell density using a noncontact specular microscope (Robo, KSS 300; Konan Medical, Hyogo, Japan). In eyes where graft failure occurred, follow-up data were included up to the point of graft failure. Data after graft failure was not included in endothelial cell loss and BSCVA analyses. Data were also recorded on the management of failed grafts. In addition to comparison of the study group with a control group of patients with Fuchs dystrophy, a subgroup analysis within the study group compared survival, detachment, and rebubble rates between eyes that had a GDD (n = 32) and those that did not have a GDD (n = 19).

Data were recorded in Microsoft Excel 2016 (Microsoft Corp, Redmond, Washington, USA) and analyzed using the XLSTAT add-in (v 2019.1.2; Addinsoft, New York, New York, USA). Continuous paired variables were compared using the Wilcoxon nonparametric test. Continuous nonpaired variables were compared with the Mann-Whitney U test. Repeated measures analysis of variance and mixed effect models were used where appropriate. Categorical variables were analyzed using the Fisher exact test. Graft survival was analyzed using Kaplan-Meier survival analysis and was compared between groups using the log-rank test. All tests were 2-tailed, and the threshold for statistical significance was defined as a P < .05.