Corneal graft manipulation spurs endothelial cell loss

Mounting and dismounting tissue from delivery devices was the leading cause of damage in a laboratory study, surgeon says.

Mark A. Terry

FORT LAUDERDALE, Fla. — Manipulation of graft tissue poses some risk of endothelial damage in Descemet stripping automated endothelial keratoplasty, a speaker said here.

Mark A. Terry, MD, presented work with two common tissue insertion devices at the Association for Research in Vision and Ophthalmology meeting.

“It’s not as easy as the animation videos make it seem,” Dr. Terry said. “There’s a learning curve to using these devices. You likely need at least 10 practice sessions.”

Endothelial cell damage in DSAEK results from tissue handling, mounting and dismounting, in addition to wound compression during graft insertion, Dr. Terry said.

“The damage with common techniques of DSAEK is largely caused by compression of the tissue by the wound during insertion,” he said. “That’s responsible for most of the damage, although folding and compressing it with other instruments can cause damage as well.”

Devices and tissue preparation

Dr. Terry and colleagues set out to compare endothelial cell loss resulting from graft manipulation with two graft delivery devices, the Tan EndoGlide (Angiotech Pharmaceuticals) and Neusidl Corneal Inserter (NCI, Fischer Surgical). The only aspect of manipulation analyzed was the mounting and dismounting of tissue from either device.

The Tan EndoGlide requires pulling the donor lenticule through the coiling delivery tube and then pulling it out the other end into the anterior chamber, while the NCI requires mounting the graft onto a platform that is then retracted into a delivery tube, rolling the tissue into a circle as it is retracted by the platform.

Dr. Terry and colleagues mounted nine cadaver corneas on the NCI and nine corneas on the Tan EndoGlide. All grafts were punched out with an 8-mm trephine.

Corneas were then delivered into a basin of balanced salt solution, removed from the basin and stained with trypan blue for 90 seconds to mark endothelial damage. Investigators analyzed digital photographs of the entire endothelial surface of each cornea to gauge the extent and pattern of endothelial damage stemming from tissue manipulation.

“I’m very pleased that this consistent method we designed for quantifying endothelial damage is becoming more widespread, as it’s a more uniform way of looking at endothelial cell loss in the laboratory setting,” Dr. Terry said. “When you do the Adobe Photoshop, you can also get rid of the damage that’s caused by the trephination process itself, and you just look at the specific manipulation damage.”

Cell death rates and patterns

Mounting the donor lenticule onto the NCI resulted in a mean endothelial cell loss of about 14%. Analysis also showed a unique pattern of endothelial cell damage, Dr. Terry said.

“It’s a very interesting pattern,” he said. “There’s a kind of a crinkling effect on this tissue where you see these multiple little lines of cell damage, and there’s not the grasp point of damage that you would see if you do a forceps insertion. But this diffuse ‘crinkling’ damage is a typical pattern with NCI tissue that we have not seen with other manipulations or devices.”

Using the Tan EndoGlide yielded a mean endothelial cell loss of 13.5%, Dr. Terry said.

“When you look at the Tan Endoglide, what’s interesting is that now you do see a different pattern of endothelial damage than what you had with the NCI,” he said. “You’ll see a distinct grasp point and you’ll see radiating lines out from the grasp point on this. This pattern is very similar to the damage pattern we saw in our published study looking at DSAEK ‘pull-through’ techniques, which pull the tissue through the incision. Importantly, however, the total amount of damage using the Tan EndoGlide appears to be less than pulling the tissue through the incision without the glide.”

The manufacturer of the Tan EndoGlide recommends placing viscoelastic material onto the endothelial surface to protect it. For these laboratory experiments, the viscoelastic must be washed off before vital dye staining can be done. Viscoelastic has not been recommended for grafts inserted with the NCI and so was not used for the laboratory experiments with the NCI, Dr. Terry said.

The study has shown that endothelial tissue damage can occur with these new insertion devices from simply mounting and dismounting the tissue, Dr. Terry said. “Whether these devices will offer better long-term endothelial cell survival or not will require data from a carefully controlled clinical trial,” he added. – by Matt Hasson

• Mark A. Terry, MD, can be reached at 1040 NW 22nd Ave., Suite 200, Portland, OR 97210; 503-413-6223; fax: 503-413-6937; e-mail: mterry@deverseye.org.