• Descemet membrane
• DMEK
• endothelial keratoplasty
• donor preparation
• cornea
• treatment surgery
• experimental & laboratory
• eye (tissue) banking

We are pleased to note that Dr Zarei-Ghanavati and his team have replicated our technique (originally presented at the American Academy of Ophthalmology meeting in Atlanta 2008 and subsequently published)1 and agree that it is a viable, quicker and easier alternative to conventional donor preparation for Descemet membrane endothelial keratoplasty (DMEK) surgery.2 The process of pneumatic dissection described is essentially similar to ours; however, they have overlooked some key adaptations which further maximise the potential of this novel technique.

We performed a superficial keratectomy using a 300 μm microkeratome head prior to performing the pneumatic dissection. In the event of failure to separate Descemet and endothelium (5% in our series), this additional step ensures that the tissue can still be used for Descemet stripping automated endothelial keratoplasty surgery and thus eliminates wastage. Furthermore, the anterior lamella obtained may be utilised for other cases.

Advancing the needle to a distance of approximately 2 mm from the limbus is sufficient to obtain the correct depth for separating Descemet and endothelium. The authors advocate advancing the needle into the central cornea. In our experience, this increases the chances of penetrating stroma or perforating through the endothelium. This results in either failure to separate Descemet and endothelium or tissue wastage.

The dissected tissue can be prepared and stored for up to 7 days in tissue culture medium without significant endothelial cell loss. We performed cell counts prior to preparation and 7 days after storage. Average endothelial cell loss was 4.44±4.3% which is lower than that reported using other techniques.

Staining the area of dissected Descemet and endothelium by injecting trypan blue into the bubble allows delineation of the dissected tissue. Although the authors have stained the endothelium directly, they advise against its use due to endothelial toxicity. While there is lack of convincing evidence that a low concentration of trypan blue and short exposure is toxic, injecting it into the bubble offers the advantage of (a) preventing direct contact with endothelium and (b) allowing eccentric punching to obtain a rim of stromal support.3

A key advantage of this technique is that the Descemet membrane (DM) is prevented from coiling into a roll as the surface tension between it and the underlying stroma keeps it flat. This facilitates correct orientation and eases delivery. Lifting the DM off the underlying stroma or filling the punching block with fluid would result in loss of surface tension, allowing the DM to form a roll. In our series, the DM tissue remained flat in all cases. Although the authors do not elaborate on how the roll of DM was formed at the end of their preparation, we advocate avoiding excessive fluid within the punch block and lifting of the graft from the underlying stromal bed.

We encourage readers to adopt the above variations which enhance the advantages that this technique offers and allow eye banks and surgeons greater flexibility and ease.