Endothelial keratoplasty: the influence of preoperative donor endothelial cell densities on dislocation, primary graft failure, and 1-year cell counts.

Purpose: The purpose of this study was to determine if an association exists between preoperative donor central endothelial cell density (ECD) and the complications of donor dislocation, iatrogenic primary graft failure (IPGF), and endothelial survival at 1 year after endothelial keratoplasty (EK) surgery.

Methods: A prospective, nonrandomized, interventional case study was conducted evaluating 629 consecutive EK procedures. The preoperative donor ECD was recorded for each case. The mean preoperative ECD of the group of EK cases with postoperative donor dislocation was compared with the mean ECD of the group of EK cases that did not suffer donor dislocation. The subset of eyes that underwent Descemet stripping automated endothelial keratoplasty (DSAEK) was also evaluated for dislocation, IPGF, and ECD at 1 year.

Results: There were 31 eyes that suffered a dislocation in the overall group of 629 eyes (4.9% dislocation rate). The mean preoperative ECD of the donor tissue in this dislocation group was 2769 cells per square millimeter (range = 2147-3454 cells/mm). The mean preoperative ECD of the donor tissue in the group that did not dislocate (n = 598) was 2818 cells per square millimeter (range = 2110-4209 cells/mm). There was no significant difference in preoperative ECD between these groups (P = 0.428). There was a subset of 350 cases of DSAEK, of which 9 cases dislocated (2.6% dislocation rate). The mean preoperative ECD of the donor tissue in this dislocation group was 2604 cells per square millimeter (range = 2323-3175 cells/mm), with 5 of the 9 dislocation donors with a preoperative ECD above 2500 cells per square millimeter. The mean preoperative ECD of the donor tissue in the group that did not dislocate (n = 341) was 2825 cells per square millimeter (range = 2110-4209 cells/mm). There was no significant difference in preoperative ECD between these groups (P = 0.069). There was no significant correlation between preoperative ECD and the ECD at 1 year after DSAEK (n = 90; Pearson correlation = 0.184; P = 0.082). There were no IPGFs in the entire series of 350 consecutive DSAEK cases, and therefore, no statistical analysis is possible for IPGF.

Conclusions: Preoperative donor ECD was not associated with donor dislocation for any form of EK surgery. Tissue with donor cell counts below 2500 cells per square millimeter can attach, and tissue with donor cell counts above 2500 cells per square millimeter can detach. Higher preoperative donor ECD was not correlated with higher ECD at 1 year postoperatively. Surgeons' requests for donor tissue with an ECD above 2500 cells per square millimeter for DSAEK surgery for the purpose of avoiding dislocations, IPGF, or improving 1-year ECD are not supported by this data.