Early Results of Small-Incision Descemet’s Stripping and Automated Endothelial Keratoplasty

Purpose To describe our technique and early results of Descemet’s stripping and automated endothelial keratoplasty (DSAEK), identify perioperative complications, and discuss their management. Design Prospective, noncomparative, surgical case series. Participants Twenty-six eyes of 26 patients who had corneal edema from Fuchs’ endothelial dystrophy, pseudophakic bullous keratopathy, or aphakic bullous keratopathy. Intervention The donor corneal lenticule was prepared using a microkeratome and an artificial anterior chamber maintainer. Stripping of the diseased host endothelium was performed under viscoelastic using a 2.75-mm clear corneal temporal incision. The incision was enlarged to approximately 4.2 mm to allow placement of a folded 8.5-mm-diameter donor corneal lenticule. The donor graft was positioned using a temporary air bubble that was partially evacuated after 7 minutes. The corneal wound was closed with a single 10-0 nylon suture. Main Outcome Measures Preoperative and postoperative best spectacle-corrected visual acuity (BSCVA), manifest refraction, keratometry, topographic astigmatism, corneal pachymetry, and perioperative complications. Results Three months postoperatively, all donor grafts remained clear. The average 3-month postoperative BSCVA was 20/45 (range, 20/20–20/150). The average change in refractive astigmatism was 0.12±1.15 diopters (D) (range, −1.50 to 3.25). In patients who underwent simple DSAEK (i.e., no intraocular lens implantation), the average postoperative shift in spherical equivalent refractive error was 1.15±1.35 D (range, −0.25 to +4.25). Nine of 26 initial grafts dislocated postoperatively and had to be repositioned. Three of the repositioned grafts dislocated again and were replaced with new donor corneal lenticules, all of which remained clear. Conclusions Descemet’s stripping and automated endothelial keratoplasty uses a mechanical microkeratome to harvest the donor corneal lenticule and mechanical stripping of the diseased host endothelium and Descemet’s membrane. Despite a smooth graft–host interface, only 2 patients in the series achieved ≥20/25 vision. The average visual results were comparable to vision after deep lamellar endothelial keratoplasty. Although patients experienced excellent postoperative acuity with minimally induced surgical astigmatism, nearly one third of the donor lenticules needed to be either repositioned or replaced. Nevertheless, DSAEK offers significant advantages over standard penetrati