Endothelial Cell Loss Curve in Descemet Stripping Automated Endothelial Keratoplasty versus Descemet Membrane Endothelial Keratoplasty

Abstract

Purpose: To compare endothelial cell loss in adult patients with corneal endothelial disorders who were submitted to Descemet stripping automated endothelial keratoplasty (DSAEK) or Descemet membrane endothelial keratoplasty (DMEK). Methods: Retrospective, single-centre, observational cohort study. 51 eyes from 51 patients that underwent DSAEK (n=23 patients) or DMEK (n=28 patients) at Centro Hospitalar Universitário S. João (Porto, Portugal), between April 2015 and March 2020 were included. Patients without at least one endothelial cell density (ECD) determination after transplantation and those who experienced primary graft failure were excluded. Patient demographics, best corrected visual acuity (BCVA) with the logMAR scale before and one year after grafting, indication for transplantation, and postoperative complications were recorded. Specular microscopy with ECD determination (in cells/mm2) was performed on all donor corneas before grafting and regularly after transplantation. Results: Patients' demographics, indications for transplantation and BCVA before grafting were similar in both groups. BCVA 1-year after transplantation was better in the DMEK group (0.26 ± 0.19 vs. 0.47 ± 0.29 in the DSAEK group; p=0.003). ECD in donor corneas before grafting was similar in both groups (p=0.986). Graft ECD after transplantation was higher in the DMEK group at up to 5 months (p<0.001), 5 to 9 months (p=0.037) and 9 to 15 months follow-up (p=0.003), being similar in posterior determinations. ECD loss was lower in the DMEK group at up to 5 months (p<0.001), 5 to 9 months (p=0.004) and 9 to 15 months follow-up (p=0.016), being similar in posterior determinations. 2 DMEK eyes required rebubbling. 2 DSAEK eyes suffered graft rejection. Conclusion: In our cohort, DMEK presented better visual outcomes than DSAEK. The DMEK group showed higher mean ECD and lower ECD loss in the first 15 months of follow-up, but posterior measurements were similar in both groups. Therefore, both techniques had similar long-term mean ECD and ECD loss and other criteria should be used to determine which one is best suited for each case in our clinical practice.