A step-wise approach to producing femtosecond-assisted hybrid DMEK grafts with improved optical and adhesion characteristics

Session Details

Session Title: Cornea Surgical I

Session Date/Time: Monday 15/09/2014 | 08:00-10:30

Paper Time: 09:58

Venue: Boulevard B

First Author: : M.Bhogal UK

Co Author(s): :    B. Allan              

Abstract Details

Purpose:

Descemets membrane endothelial keratoplasty is has distinct advantages over other forms of endothelial keratoplasty in terms of visual recovery, best corrected visual acuity and rejection. At present the learning curve remains steep and the technique is unsuitable for cases such as aphakia or anterior chamber lenses. Previous attempts at hybrid DMEK (DMAEK) techniques have resulted in high tissue wastage and high dislocation rates. Here we present a step-by-step approach to produce femtosecond assisted DMAEK graft with minimal tissue wastage and sloped stromal rim contours that reduce fluid trapping.

Setting:

Moorfields Eye Hospital London, Institute of Ophthalmology, University College London

Methods:

16 corneoscleral rims unsuitable for transplantation were received from Moorfileds Lions Eye Bank and Bristol Eye bank. Corneas were stored in organ culture for 20+/-10 days. Each cornea was placed in a punch block and a 27 g needle inserted into the deep stromal as close to the endothelial surface as possible. Air was injected and a big bubble produced. The corneas were then placed into dextran containing media for 3 days prior to femtosecond laser dissection using the IFS 150 Intralase. In group A (n=6) a flat profile with a step was cut. In group B (n=8) a deep lamellar dissection, passing through the bubble centrally, was used to produce a graft with sloped contours. Femtosecond DMAEK grafts were imaged prior to trephination using anterior segment optical coherence tomography. They were subsequently punched and inserted under another cornea mounted on an artificial anterior chamber prior to reimaging. Trapped fluid was quantified and compared to standard profile DMAEK grafts. Interface properties were measured using atomic force microscopy and compare to microkeratome produced DSAEK grafts and peeled DMEK grafts & FS-DMAEK grafts.

Results:

Air injection was successful in big bubble production in 14/16 samples. In the 2 unsuccessful cases peeling to produce a DMEK graft was successful. Storage of grafts in dextran for 2 days at room temperature was long enough to allow dissipation of the opaque bubble layer in all cases. All grafts had uniform central profiles with a thickness of 20-30 microns and no observable variations in thickness within the optical zone. Fluid trapping was 92% lower in the sloped contour FS-DMAEK grafts compared to the standard profile grafts (p<0.001). Surface roughness (Ra), as measured by atomic force microscopy, was significantly lower in FS-DMAEK (0.1) grafts compared to DASEK grafts (0.55) and comparable DMEK (0.05) produced by peeling.

Conclusions:

FS-DMAEK produces endothelial grafts with smooth and uniform optical interfaces. The preparation schedule is suitable for use in an eye bank setting and tissue can be reallocated for DMEK is big-bubble separation is unsuccessful. Problems with graft dislocation are due to fluid trapping and can be overcome adjusting the contour of the stromal rim using the methods described.

Financial Interest:

NONE