Incidence and Clinical Course of Immune Reactions after Descemet Membrane Endothelial Keratoplasty: Retrospective Analysis of 1000 Consecutive Eyes.

PURPOSE: To analyze the incidence and clinical course of graft rejection episodes after Descemet membrane endothelial keratoplasty (DMEK). DESIGN: Retrospective analysis of a consecutive, interventional case series. PARTICIPANTS: One thousand eyes that underwent DMEK from July 2011 through August 2015 at the Department of Ophthalmology, University of Cologne. METHODS: All cases with follow-up of at least 1 month were included (mean follow-up, 18.5 months). Patients with a graft rejection episode were followed up for 1 additional year. MAIN OUTCOME MEASURES: Incidence of graft rejection, best spectacle-corrected visual acuity (BSCVA), central corneal thickness (CCT), endothelial cell density (ECD), and need for regraft. RESULTS: Nine hundred five cases met the inclusion criteria. A graft rejection episode developed in 12 patients (estimated probability of rejection at 1 year, 0.9%; at 2 years, 2.3%; at 4 years, 2.3%). At time of rejection, 9 of 12 patients had stopped corticosteroids. Five patients were symptomatic and 7 did not note the rejection episode. Intensified topical corticosteroid therapy was started immediately after diagnosis of rejection. Two eyes decompensated and required a regraft, whereas the remaining 10 eyes required no regraft (BSCVA, 0.27±0.28 logarithm of the minimum angle of resolution [logMAR]; CCT, 554.1±39.1 µm at last visit before rejection vs. BSCVA, 0.21±0.15 logMAR; CCT, 540.0±15.0 µm 3 months after rejection). One year after the rejection episodes, BSCVA and CCT in these eyes remained unchanged when compared with the last visit before rejection (BSCVA, 0.15±0.11 logMAR; CCT, 533.8±26.0 µm). Significant changes were observed for ECD values (1741±274.5 cells/mm2 at last visit before rejection vs. 1356±380.3 cells/mm2 after 3 months [P = 0.04] and 1290±359.0 cells/mm2 after 1 year [P = 0.01]). CONCLUSIONS: The risk for graft rejection after DMEK is low, and an even smaller minority requires a regraft. After intensified local corticosteroid therapy, most patients show stable visual acuity and CCT, although ECD decreases. The occurrence of immune reactions up to 2 years after surgery predominantly in patients not receiving corticosteroids supports the prolonged use of corticosteroids after DMEK.