Discussion

Endothelial keratoplasty has become a popular procedure for treatment of corneal endothelial disorders. This leaves a secure eye with less postoperative astigmatism when compared with penetrating keratoplasty. In recent years, various techniques for endothelial keratoplasty have been described, namely, deep lamellar EK (DLEK), Descemet stripping (automated) endothelial keratoplasty (DSEK/DSAEK) and DMEK. Preliminary results have shown that DMEK offers a faster visual rehabilitation and better optical quality than DSAEK.2 Although donor tissue preparation for DMEK is not dependent on expensive equipment, the manual dissection is difficult and tedious for many surgeons. Use of the air injection into the cornea with the aim to separate the DM from the stroma, first described by Anwar and Teichmann3 in their big-bubble technique, can be used to harvest donor DM as well. The big bubble technique was originally described for deep anterior lamellar keratoplasty, but we have used it for preparation of donor DM. Interestingly, this reverse big bubble formation is much easier in a donor button than in the living eye of a patient during the deep anterior lamellar keratoplasty procedure.

In comparison with manual dissection of donor DM, it can be done faster, easier and without endothelial cell touch during the detachment. In addition, the technique is familiar to many surgeons and has a short learning curve. Lie JT et al4 reported 4% to 7% endothelial cell loss after stripping DM. We did not perform specular microscopy, but we suppose less endothelial loss rate would be less because it requires minimal touch and is a less traumatising procedure.

As this was an experimental study, we used trypan blue to visualise the endothelium. The procedure can be performed without trypan blue staining because this dye can be toxic to the endothelium and would not be recommended in actual surgical practice.