Purpose To analyze the impact of performing premarking of the Descemet roll and using SF6 20% on a surgeon's Descemet membrane endothelial keratoplasty (DMEK) learning process. Methods A total of 30 consecutive eyes with endothelial dysfunction undergoing DMEK during the learning curve of a surgeon were retrospectively analyzed. Prior to the study, the surgeon had already performed 10 DMEKs. The first 15 consecutive patients were included in group 1 (no premarking and air tamponade) and the other 15 consecutive patients were included in group 2 (premarking and SF6 tamponade). Main outcome parameters were best-corrected visual acuity (BCVA), endothelial cell density (ECD) loss at 6 months, and intraoperative and postoperative complications. Results Among the 2 groups, BCVA and ECD loss at 6 months were similar. However, there was a statistically significant reduction in primary graft failure (40% vs 0%) and need of rebubbling due to complete or partial graft detachment (40% vs 6%) when comparing group 1 versus group 2. In group 1, half of the patients needing rebubbling had primary graft failure. Conclusions Based on our personal experience, premarking the graft to assess orientation and using a SF6 gas tamponade dramatically reduces the risk of primary graft failure and the need for rebubbling even during the first stages of the learning curve. These findings should encourage surgeons to safely change from Descemet stripping automated endothelial keratoplasty to DMEK.
Keywords: Descemet membrane endothelial keratoplasty; Endothelial keratoplasty; Primary graft failure; Rebubbling, SF6.