PURPOSE: To study the histopathologic features of Descemet stripping automated endothelial keratoplasty (DSAEK) grafts removed after graft failure and to correlate with relevant clinical features. METHODS: Retrospective study of 8 cases in 7 patients who experienced DSAEK graft failure during the surgeons' learning phase. Perioperative clinical findings were correlated with pathologic findings of the explanted graft specimens. Imaging software was used to measure the central and peripheral thickness of each graft. RESULTS: Graft failure occurred in association with graft detachment or excessive surgical manipulation. In 4 cases, the donor tissue detached, including 2 cases with routes for bubble escape. In 3 cases, the graft remained attached but failed to clear. The mean preoperative DSAEK button thickness was 205 +/- 61.8 microm. Average death to preservation time was 8 hours 47 minutes, and average preservation to surgery time was 2 days 22 hours. By light microscopy, explanted donor tissue showed varying degrees of keratocyte degeneration and marked endothelial cell loss. Two specimens with bubble escape had less endothelial loss, and 1 specimen showed epithelial ingrowth at the interface. Repeat keratoplasty (5 DSAEK, 2 penetrating keratoplasty) was successful in 6 of 7 cases. CONCLUSIONS: This series correlates factors that may play a role in DSAEK failure with histopathologic features of explanted DSAEK lenticules. Marked endothelial loss was common in cases with surgical trauma but was less in cases with bubble escape. One specimen showed an epithelial membrane on the stromal interface. Outcomes of repeat DSAEK in these patients seem promising.
PURPOSE: To study the histopathologic features of Descemet stripping automated endothelial keratoplasty (DSAEK) grafts removed after graft failure and to correlate with relevant clinical features. METHODS: Retrospective study of 8 cases in 7 patients who experienced DSAEK graft failure during the surgeons' learning phase. Perioperative clinical findings were correlated with pathologic findings of the explanted graft specimens. Imaging software was used to measure the central and peripheral thickness of each graft. RESULTS:Graft failure occurred in association with graft detachment or excessive surgical manipulation. In 4 cases, the donor tissue detached, including 2 cases with routes for bubble escape. In 3 cases, the graft remained attached but failed to clear. The mean preoperative DSAEK button thickness was 205 +/- 61.8 microm. Average death to preservation time was 8 hours 47 minutes, and average preservation to surgery time was 2 days 22 hours. By light microscopy, explanted donor tissue showed varying degrees of keratocyte degeneration and marked endothelial cell loss. Two specimens with bubble escape had less endothelial loss, and 1 specimen showed epithelial ingrowth at the interface. Repeat keratoplasty (5 DSAEK, 2 penetrating keratoplasty) was successful in 6 of 7 cases. CONCLUSIONS: This series correlates factors that may play a role in DSAEK failure with histopathologic features of explanted DSAEK lenticules. Marked endothelial loss was common in cases with surgical trauma but was less in cases with bubble escape. One specimen showed an epithelial membrane on the stromal interface. Outcomes of repeat DSAEK in these patients seem promising.
Authors: Qing Zhang; J Bradley Randleman; R Doyle Stulting; W Barry Lee; Donald U Stone; Alan M Kozarsky; Hans E Grossniklaus Journal: Arch Ophthalmol Date: 2010-08