Endothelial cell loss after descemet stripping with endothelial keratoplasty - Influencing factors and 2-year trend

Purpose: To identify factors correlated with early endothelial cell loss after Descemet stripping with endothelial keratoplasty (DSEK) and to document cell loss over a 2-year period. Design: Cross-sectional and longitudinal retrospective analyses. Participants: A series of 263 eyes in 216 patients who were treated with DSEK and were examined at the same center after surgery. Methods: Six-month endothelial cell density (ECD) measurements were analyzed by multivariate analysis of variance to identify preoperative, operative, and postoperative factors that significantly influenced early cell loss. In addition, ECD measurements were compared longitudinally at 6, 12, and 24 months after grafting. Main Outcome Measures: Central ECD after transplantation. Results: Mean ECD was 2000 +/- 550 cells/mm(2) 6 months after DSEK, representing cell loss of 34 +/- 18%. Factors associated with lower 6-month cell loss were use of single-point fixation forceps that compressed the donor tissue only at the tip during graft insertion (P = 0.025) and absence of any secondary donor reattachment procedure (P = 0.035). Six-month cell loss also was significantly less when DSEK was performed as a combined procedure; in combined cases, the graft was inserted through a clear corneal incision, whereas in standard cases, it was inserted through a scleral tunnel incision, which likely exerted more compression. Cell loss was comparable with manual and microkeratome donor dissection techniques (P = 0.91). Postoperative ECD was correlated positively with donor ECD (P<0.0001) and was correlated inversely with donor age (P<0.0001), although these donor characteristics explained only 10% of the total variance in 6-month ECD. Factors not significantly correlated with 6-month ECD or cell loss included donor death-to-preservation time (range, 1-22 hours; P = 0.45), donor death-to-use time (range, 2-8 days; P = 0.86), and recipient demographics. In 34 DSEK eyes analyzed longitudinally, mean ECD was 3100 +/- 250 cells/mm(2) before surgery, 2000 +/- 540 cells/mm(2) at 6 months, 1900 +/- 480 cells/mm(2) at 1 year, and 1800 +/- 490 cells/mm(2) at 2 years. Conclusions: Cell loss 6 months and 1 year after DSEK was higher than in a recent penetrating keratoplasty series performed at the same center, consistent with more donor tissue manipulation in DSEK. Cell loss was reduced with certain variations in surgical technique.