Impact of Surgical Learning Curve in Descemet Membrane Endothelial Keratoplasty on Visual Acuity Gain.

PURPOSE: To evaluate the learning curve for graft preparation and graft unrolling during Descemet membrane endothelial keratoplasty (DMEK) and to assess the evolution of visual acuity gain and percentage cell loss with experience. METHODS: The first 109 DMEK procedures performed by a single surgeon (A.S.) at the Rothschild Foundation Ophthalmology Hospital in Paris, France, between March 2012 and November 2014 were included. Best-corrected visual acuity and endothelial cell density were recorded preoperatively and again 1 week, 1 month, 3 months, and 6 months after DMEK. Donor age and ECC were registered. Graft preparation time and graft unrolling time were assessed using video recording. Incidence and types of complications were noted. RESULTS: The number of cases necessary to reach 90% of the plateau of the learning curve was 68 for preparation time and 46 for unrolling time in this model. There was no correlation between the best-corrected visual acuity gain at 6 months postsurgery and the learning curve. The percentage cell loss was found to be significantly lower with experience (R = 0.17, P = 0.0011). CONCLUSIONS: Surgical experience allowed faster graft preparation and faster unrolling time in DMEK. Neither experience nor percentage cell loss influenced postoperative visual acuity gain. The number of procedures needed to reach a good standard of care was estimated to be 50 in our patient database.

Intraoperative OCT-Assisted DMEK: 14 Consecutive Cases.

PURPOSE: To describe the utility of a new intraoperative optical coherence tomographer (OCT) to evaluate endothelio-Descemet graft orientation during Descemet membrane endothelial keratoplasty (DMEK) procedures. METHODS: Prospective, observational, and single-center pilot case series including 14 eyes of 14 patients consecutively scheduled for DMEK surgery. After injecting the graft into the anterior chamber, the graft orientation was assessed with the help of anterior segment OCT. The surgical time and unfolding time were measured. The postoperative measurements included best-corrected visual acuity, central pachymetry, and specular microscopy at 1 month. RESULTS: Using the OCT images, it was possible to evaluate the graft orientation in all cases. The mean unfolding time was 6.1 ± 3.0 minutes, the mean best-corrected visual acuity was 0.3 ± 0.3 logarithm of the minimum angle of resolution, the mean decrease in central pachymetry was 213 ± 177 µm, and the mean central endothelial cell count was 1906 ± 319 cells per square millimeter. CONCLUSIONS: Live intraoperative OCT is useful to visualize and assess graft orientation in DMEK surgery. It enables faster graft positioning with less graft manipulation in the presence of severe corneal edema.

Serological viral testing of cadaveric cornea donors.

BACKGROUND: Cornea graft recipients are exposed to viral transmission from the donor. Cadaveric donor serum is often of poor quality and frequently yields falsely positive results in serological assays that may result in the graft being needlessly discarded. OBJECTIVE: We examined the influence of the time of blood collection after death, and the macroscopic aspect of serum, on serological test results in cadaveric cornea donors. METHODS: Five hundred sixty-five consecutive cadaveric cornea donors were systematically tested for serological markers of human immunodeficiency virus type 1 and 2, human T-cell leukemia virus type 1, hepatitis B and hepatitis C viruses (HCV). We studied the influence of the macroscopic aspect of the donor's serum and the time of blood collection after death on the results of serological testing and on the subsequent decision to use or discard the graft. RESULTS: Twenty-one and a half percent of corneas were rejected on the basis of virological test results. We found significant relationships between the macroscopic aspect of serum at the time of testing and: (i) a positive, equivocal or discrepant result of immunoassays, for all markers except anti-HCV antibodies, (ii) non acceptance of cornea grafts, and (iii) the time of blood sampling after death. CONCLUSIONS: The macroscopic aspect of postmortem blood samples is the best predictor of the specificity of serological testing in cornea donors. Serological results should be interpreted with care when serum is macroscopically abnormal, and cadaveric donors should not be sampled more than 12 hr after death.

Expression of HLA-G in human cornea, an immune-privileged tissue.

Human leukocyte antigen (HLA)-G retains the capacity to modulate immune responses, favoring the establishment of tolerance in solid-tissue allotransplants. To better understand the mechanisms that promote corneal allograft survival, we investigated whether HLA-G was an immunoregulatory factor involved in corneal immunology. We therefore sought HLA-G expression in corneal tissues. Corneal transplantation consists in replacing the center of a diseased cornea with normal corneal tissue. Two corneal parts are not used in such surgery: diseased central corneal tissue and peripheral normal cornea. For this study, we used healthy corneas obtained from deceased donors and diseased corneas obtained from patients with pseudophakic bullous keratopathy or keratoconus who had undergone corneal transplantation. Immunohistochemical analysis carried out on the cryopreserved corneas showed a positive immunohistochemical staining with anti-HLA-G, anti-HLA-A, -B, and -C, and anti-HLA class I monoclonal antibodies. Staining was obtained for keratocytes, epithelial cells, and endothelial cells from both healthy and pathologic human corneas, revealing the presence of HLA class I proteins, including HLA-G. HLA-G transcripts were detected in normal cornea by reverse transcriptase-polymerase chain reaction with a classical pattern of alternative splicing. The detection of HLA-G protein in adult corneas leads to the conclusion that this protein may contribute to the maintenance of the privileged immune status of cornea.