Modified procedure for Loading "Flat" DMEK Grafts Into an Injector.

PURPOSE The aim of this study was to determine whether loading a Descemet membrane endothelial keratoplasty (DMEK) graft using a drop-in procedure results in more endothelial cell loss (ECL) than the standard suction procedure. METHODS Pairs of donor corneas with equivalent preprocessing endothelium were prepared using the standard protocol of our eye bank. One member of each pair was loaded into an injector using the standard suction protocol. The mate graft was loaded using a drop-in protocol, in which the edge of the graft was gently grasped with a forceps, lifted to the edge of the injector, and dropped inside. Grafts were evaluated for ECL and examined for grab marks or other loading-associated damage. RESULTS There was no difference in mean ECL of grafts prepared for DMEK using the standard protocol (20.6% ± 4.5%) compared with that of mate grafts prepared using the drop-in loading protocol (19.5% ± 4.8%, P = 0.59). There was no consistent pattern of damage in the drop-in-loaded grafts, as grab marks or other tissue damage associated with the drop-in loading protocol were not consistently identified by a trained corneal surgeon. CONCLUSIONS ECL was not significantly different in grafts prepared using a drop-in loading procedure compared with grafts prepared using the standard suction protocol. The drop-in loading protocol may be particularly useful to surgeons who load their own grafts and eye bank processing technicians who encounter a "flat" DMEK graft that does not scroll or a loosely scrolled DMEK graft.