Suture pull-through insertion of graft donor in Descemet stripping automated endothelial keratoplasty: Results of 4-year follow-up

PURPOSE: To report our clinical experience and 4-year follow-up results of Descemet stripping automated endothelial keratoplasty (DSAEK) with the suture pull-through insertion technique. METHODS: This is a retrospective study of 195 eyes in which a posterior lamellar keratoplasty was performed between 2007 and 2011. The insertion of a folded donor lenticule was performed with a double-armed 10-0 suture using a straight transchamber needle and half-circle needle. Endothelial cell density was measured annually up to 4 years after the surgery, and cell loss was calculated based on the median preoperative donor endothelial cell density. Postoperative complications, primary graft failure, pupillary block, and dislocation of the donor tissue were assessed. RESULTS: All patients underwent uncomplicated DSAEK. Data were available for 195 eyes (100%) at 1 year, 186 eyes (95.3%) at 2 years, 176 eyes (90.2%) at 3 years, and 160 eyes (82%) at 4 years. Median preop-erative donor endothelial cell density was 2688 cells/mm(2) [interquartile range (IQR) 207.5 cells/mm(2)], which decreased by 27% at 1 year (1956 cells/mm(2), IQR 264.8 cells/mm(2)), 31% at 2 years (1855 cells/mm(2), IQR 320.5 cells/mm(2)), 35% at 3 years (1756.5 cells/mm(2), IQR 306.5 cells/mm(2)), and 36% at 4 years (1709.5 cells/mm(2), IQR 288,0 cells/mm(2)). Nine patients (4.6%) had a dislocation of donor tissue; all were successfully reattached with a second air injection. Only three eyes (1.5%) developed graft failure. Pupillary block was present in 15 eyes (7.7%). CONCLUSION: DSAEK with suture pull-through insertion of donor graft represents a simplified and safe technique that has endothelial cell loss comparable with other techniques and low rates of intraoperative and postoperative complications.