PURPOSE: To perform a fellow eye comparison of outcomes and complications when using air or sulfur hexafluoride (SF6) gas as a tamponade in Descemet membrane endothelial keratoplasty (DMEK). METHODS: One hundred thirty-six eyes of 68 consecutive patients who underwent uneventful DMEK in both eyes for Fuchs endothelial corneal dystrophy were included in this retrospective study. Inclusion criteria were air tamponade (80% of the anterior chamber volume) in the first eye and 20% SF6 gas tamponade (80% of the anterior chamber volume) in the second eye; and same donor tissue culture condition in both eyes. All eyes received laser iridotomy on the day before DMEK. Main outcome measures included preoperative and postoperative best-corrected visual acuity, endothelial cell density, corneal volume, rebubbling rate, and rate of postoperative pupillary block caused by the air/gas bubble. RESULTS: Thirteen of 68 eyes (19.1%) with an air tamponade needed rebubbling compared with 4 of 68 eyes (5.9%) with an SF6 gas tamponade (P = 0.036). Postoperative pupillary block necessitating partial release of air/gas occurred in 1 eye (1.5%) with an air tamponade and 3 eyes (4.4%) with an SF6 gas tamponade (P = 0.301). There were no significant differences in preoperative and postoperative best-corrected visual acuity, endothelial cell density, and corneal volume within 3-month follow-up. CONCLUSIONS: Our results confirm the previously reported better graft adhesion when using an SF6 gas tamponade in DMEK without increased endothelial cell toxicity. The rate of pupillary block in eyes with an SF6 gas tamponade was comparable to that with an air tamponade. As a consequence, we recommend using SF6 gas as the tamponade in DMEK.
PURPOSE: To perform a fellow eye comparison of outcomes and complications when using air or sulfur hexafluoride (SF6) gas as a tamponade in Descemet membrane endothelial keratoplasty (DMEK). METHODS: One hundred thirty-six eyes of 68 consecutive patients who underwent uneventful DMEK in both eyes for Fuchs endothelial corneal dystrophy were included in this retrospective study. Inclusion criteria were air tamponade (80% of the anterior chamber volume) in the first eye and 20% SF6 gas tamponade (80% of the anterior chamber volume) in the second eye; and same donor tissue culture condition in both eyes. All eyes received laser iridotomy on the day before DMEK. Main outcome measures included preoperative and postoperative best-corrected visual acuity, endothelial cell density, corneal volume, rebubbling rate, and rate of postoperative pupillary block caused by the air/gas bubble. RESULTS: Thirteen of 68 eyes (19.1%) with an air tamponade needed rebubbling compared with 4 of 68 eyes (5.9%) with an SF6 gas tamponade (P = 0.036). Postoperative pupillary block necessitating partial release of air/gas occurred in 1 eye (1.5%) with an air tamponade and 3 eyes (4.4%) with an SF6 gas tamponade (P = 0.301). There were no significant differences in preoperative and postoperative best-corrected visual acuity, endothelial cell density, and corneal volume within 3-month follow-up. CONCLUSIONS: Our results confirm the previously reported better graft adhesion when using an SF6 gas tamponade in DMEK without increased endothelial cell toxicity. The rate of pupillary block in eyes with an SF6 gas tamponade was comparable to that with an air tamponade. As a consequence, we recommend using SF6 gas as the tamponade in DMEK.