Repeat Descemet Membrane Endothelial Keratoplasty

Purpose: To evaluate the outcomes of secondary Descemet membrane endothelial keratoplasty (DMEK) after failed primary DMEK. Design: Retrospective, interventional case series. Participants: Fifty-five DMEK recipients 42 to 89 years of age. Methods: An initial consecutive series of 1655 DMEK surgeries was reviewed to identify cases of secondary DMEK after failed primary DMEK (n 1⁄4 55). A paired fellow-eye analysis was performed with a subgroup of 29 patients who underwent secondary DMEK in 1 eye and successful primary DMEK in the fellow eye. Main Outcome Measures: Corrected distance visual acuity (CDVA), central corneal thickness, and 1-year endothelial cell loss. Results: The median follow-up after DMEK regraft was 18 months (range, 3e61 months). All 55 regrafts cleared, 8 (15%) had air reinjected to promote attachment, 1 eye (2%) with trabeculectomy and progressive synechiae demonstrated late endothelial failure, and no rejection episodes occurred (0%). In the paired analysis, the median duration of endothelial decompensation before the regraft was 21 days (range, 2e133 days). At 1, 3, 6, or 12 months, CDVA did not differ between the primary and secondary grafts in fellow eyes (mean difference, 2 Snellen letters; P > 0.05 at all examinations). At 1 year, the visual acuity was 20/20 in 61%, 20/25 in 81%, and 20/40 in 100% of the secondary grafts in the paired analysis, excluding 1 eye with retinal problems. Vision differed by 1 line between fellow eyes in all but the 1 patient with the longest time to regraft (133 days), who demonstrated central haze and irregular astigmatism from anterior stromal scarring during that period. At 1 year, CDVA associated with the scarring was 20/40 versus 20/20 for the fellow-eye primary graft. The central corneal thickness was comparable between fellow-eye primary and secondary grafts at 3, 6, and 12 months (mean difference at 1 year, 2 mm; P 1⁄4 0.57). The 1-year endothelial cell loss was comparable in primary and secondary grafts (27% vs. 31%, respectively; P 1⁄4 0.58). Conclusions: In patients who received prompt intervention to minimize the duration of central corneal decompensation, the visual outcomes with secondary DMEK matched the fellow-eye visual outcomes with primary DMEK. Ophthalmology 2015;-:1e6 a 2015 by the American Academy of Ophthalmology.