High endothelial cell count in donor tissue not needed to prevent dislocations

The rate of dislocation of grafts does not seem to be related to the donor endothelial cell count when performing endothelial keratoplasty. Grafts with endothelial cell counts lower than 2,500 cells/mm² performed well with no dislocations, according to Mark Terry, MD, Devers Eye Institute, Portland, OR.

Dr. Terry has had extensive experience with epithelial keratoplasty: 32 cases of Descemet's stripping endothelial keratoplasty (DSEK), 275 cases of deep lamellar endothelial keratoplasty (DLEK), and 220 cases of Descemet's stripping and automated endothelial keratoplasty (DSAEK).

The visual acuity results are better with DSAEK compared with the other two procedures, he said. Six months postoperatively, 92% of cases had a visual acuity of 20/40 or better, 30% 20/25 or better, and 12% 20/20 or better. Dislocation rates have also improved over time. DLEK resulted in a dislocation rate of 8%, DSEK 6%, and DSAEK 1.37%, according to Dr. Terry.

When accepting tissue from an eye bank, Dr. Terry commented that he will accept tissue with an endothelial cell count of 2,000 cells/mm² or higher, a donor patient age of 75 years or less, and tissue that is aged 11 days or less.

"When we consider all of our cases of endothelial keratoplasty and the preoperative endothelial cell counts in the 27 cases that dislocated and the vast majority that did not, the mean preoperative endothelial cell counts did not differ," Dr. Terry said. "In fact, there were eyes in which dislocation did not occur that had cell densities of about 2,100 cells/mm² and eyes that dislocated in which the cell densities were as high as 3,454 cells/mm².

"Dislocation did not seem to depend upon the preoperative endothelial cell count. I believe that dislocation in endothelial keratoplasty is not tissue dependent but technique dependent," he emphasized.

Dr. Terry believes that the technique used in his institution may be responsible for the low rate of dislocation. He and his colleagues do peripheral recipient bed scraping and surface compression sweeping to remove interface fluid. They do not use corneal stab incisions or drain fluid from corneal stab incisions. In addition, they use a 10-minute rest time for the air bubble and leave only an 8-mm residual air bubble, which has also eliminated papillary block.

"The preoperative endothelial cell counts do not influence whether the donor tissue will or will not dislocate with our technique," he said. "Donor endothelial cell counts below 2,500 cells/mm² did not dislocate using our technique and even donor endothelial cell counts above 3,000 cells/mm² can dislocate. Donor preoperative cell counts cannot correlate with 1-year endothelial cell counts, but may correlate with the 6-month endothelial cell counts.

"Surgical technique is likely the most important factor for preventing dislocation, primary graft failure, and long-term survival," Dr. Terry said. He urged corneal surgeons not to disregard available corneal tissue that has lower endothelial cell counts.