Document resource
Purpose To compare the 6-month post-graft results in recipients of surgeon-dissected DMEK (SD-DMEK) versus eye bank technician-predissected (TD-DMEK), with organ cultured corneasMethods All consecutive patients who underwent DMEK at St-Etienne University Hospital by a single surgeon were analyzed. Patients operated on before Jan 2023 received a SD-DMEK, (n=141), and those operated between Jan and Dec 2023 received TD-DMEK, pre-dissected by 3 trained technicians at the St-Etienne eye bank (n=71) using the same SCUBA technique but leaving a central Descemet attachment. For donor corneas, we analyzed: donor age, storage time, dissection failure rate, and pre-dissection endothelial cell density (ECD). For recipients, we analyzed: age, sex, indication, comorbidities, best-corrected visual acuity (BCVA), and pachymetry. Post-surgery analysis included a minimum follow-up of 6 months and covered: rebubbling rate, other complications, failures, BCVA, ECD and cell loss, and pachymetry variation. All corneas were stored in CorneaMax (Eurobio-Scientific, Les Ulis, France) and not deswelled.Results SD-DMEK initially differed from TD-DMEK for storage time with respectively 19.1±3.6 (mean±SD) vs 21.5±5.2 days (p<0.01), donor age with 71±10 vs 76±10 years (p=0.002), initial ECD with 2568±270 vs 2692±296 cells/mm² (p=0.001) and dissection failure rate with 1/142 (0.7%) vs 4/75 (5.3%) (p=0.0496). TD-DMEK were stored for 2.2±1.1 days (1 to 4). In recipients, rebubbling rates were 14/141 (9.9%) vs 11/71 (15.5%) (p=0.257) and primary graft failures occurred in 1/141 (0.7%) vs 2/71 (2.8%) (p=0.220). At 6 months postoperatively, ECD did not differ with 1344±429 vs 1476±424 cells/mm² (p=0.054), nor the decrease in pachymetry, nor the BCVA.Conclusions Grafts pre- dissected by technicians from corneas stored in organoculture without deturgescence and left with a small central attachment had a survival comparable to that of grafts dissected in the operating room on the same day by the surgeon. Technicians choose grafts with a higher DCE than those sent to the surgeon for fear of increased cell loss, which we did not observe. Predissection by technicians, under the conditions we have described, is therefore a safe practice.