Cornea
Dissecting DSEK’s devices and techniques

View chart - Keratoplasty instrumentation

Surgeons try procedures and equipment that reduce endothelial cell loss

Information on Dr. Tan's EndoGlide
Source: Donald T.H. Tan, F.R.C.Ophth.

The devices and techniques used for Descemet-stripping endothelial keratoplasty (DSEK) are evolving. In addition to performing the surgery with forceps, more and more surgeons are trying glides and injectors to lower endothelial cell loss and aim for better outcomes.
This DSEK evolution is important as nearly half of the 40,000 patients in the U.S. who undergo corneal transplantation each year are candidates for DSEK, said Terry Kim, M.D., professor of ophthalmology, Duke University School of Medicine, Durham, N.C.

Using forceps

According to Gerald D’Aversa, M.D., Rockville Centre, N.Y., surgeons who use forceps commonly take the donor tissue that is on a disk, fold it into the shape of a taco for the so-called “taco fold,” then use the forceps to insert the tissue into an opening in the anterior chamber. Once the tissue is in the eye, the “taco” unfolds, and the forceps are removed. “A number of forceps are used, included Goosey, Olga, Kelman, and MacPherson forceps,” Dr. D’Aversa said.
However, forceps and the taco-fold technique seem to cause a higher rate of endothelial cell loss—reaching as high as 30% to 35% at six months, said Donald T.H. Tan, F.R.C.Ophth., head and senior consultant, Singapore National Eye Centre, Singapore.
Those percentages can reach 36% to 54% at two years in some studies, Dr. Kim said. Additionally, the forceps may catch the tissue, pulling it back again toward or into the incision, according to a report in the 2009 issue of the Journal of Cataract and Refractive Surgery written by Chandra Balachandran, M.D., Netherlands Institute for Innovative Ocular Surgery, Rotterdam, the Netherlands.

The pull-through technique

Concerns over minimizing endothelial cell loss has led to the use of the pull-through technique in DSEK. With the pull-through technique, donor tissue is pulled into the anterior chamber with the use of a glide that is placed through an incision at the limbus on the opposite side. “I think that has opened a slew of new devices,” Dr. D’Aversa said.
Dr. D’Aversa has designed a microforceps and insertor that will premiere at this year’s ASCRS meeting in Boston. With his surgical tools, surgeons will avoid folding of the donor disk and will insert the donor tissue through a 4.5-mm to 5-mm incision.
Although pull-through devices are thought to have some advantages over the use of forceps, they still “may induce additional trauma to the recipient tissues; for example, by requiring a secondary incision,” Dr. Balachandran wrote.
The report from Dr. Balachandran and co-investigators describes a technique in which a 5-mm scleral tunnel incision is made to insert a folded donor posterior corneal disk over a plastic guide with the use of a 30-gauge needle. The investigators report this technique as combining the benefits of a single incision push-in technique with the advantages of a pull-through technique. Their so-called needle graft insertion technique was performed in 15 eyes of 15 patients. At six months, the mean endothelial cell density was 1,940 cells/mm2, compared with a pre-op mean endothelial cell density of 2,675 cells/mm2.
A number of glides also are available to assist with DSEK. For example, a glide from Dr. Tan, which premiered at the 2009 American Academy of Ophthalmology meeting in San Francisco, includes a cartridge design that has double-coil loading for the donor tissue. In 20 patients, the three-month endothelial cell loss was 14% and the six-month rate was 19.6%, Dr. Tan said, adding that full data are not yet available for all 20 patients. “These are quite favorable results, and these are all Asian eyes with shallow chambers,” he said. Dr. Tan’s Tan EndoGlide is marketed in the United States by Angiotech Pharmaceutics (Vancouver).

Moving toward injectors and thin EK

The newest kind of device to aid DSEK is an injector in which the donor graft is slowly retracted into a tube and then injected into the anterior chamber. This kind of device also has a port to insert fluid from a phaco machine as necessary. There is some variation in the incision size required to use the injectors, with some only requiring a 4-mm incision while others require a 5.5-mm incision, Dr. Kim said. However, the real advantage of the newest generation of DSEK instrumentation is in lower endothelial cell loss, he added. “Anecdotally with the Endosaver [Ocular Systems, Inc., Winston-Salem, N.C.], which requires a 4 mm limbal corneal incision, endothelial cell loss is 15% to 25% at six months,” he said. The Endosaver is one of the injectors under development now for use with DSEK.
The injectors do have some disadvantages. “They cost more than forceps, and there’s a learning curve, but it’s not a large one,” Dr. Kim said.
Edward J. Holland, M.D., professor of ophthalmology, University of Cincinnati, and director, Cornea Service at the Cincinnati Eye Institute, Cincinnati, said that companies are currently working on the design of injectors that do not require such large incisions. Dr. Holland said that surgeons will continue to keep an eye on long-term endothelial cell counts and graft survival rates with the use of injectors.
The other trend underway to maximize DSEK outcomes is performing the surgery with thinner tissue, Dr. Holland said. Donor tissue given by eye banks may typically be around 160 to 180 microns, making it thicker and easier to handle. “We’ve found [this tissue] doesn’t give as good visual acuity,” he said. On the other hand, patients who have received tissue that is 130 microns or less have all had a visual acuity of 20/20 or 20/25 in his experience. “If the tissue was greater than 130 microns, only 70% were 20/20 or 20/25,” he said.
For this reason, Dr. Holland is aiming for donor tissue that is 115 to 120 microns. Dr. Kim also prefers to use thinner tissue, even 100 microns or lower. “The thicker tissue gives more of a refractive shift, and there’s more light scattering.” However, both Drs. Holland and Kim acknowledge that the thinner tissue is much harder to handle.

Moving toward DMEK?

As surgeons who perform DSEK continue to evaluate their instrumentation and ways to make the surgery safer, their next challenge is evaluating Descemet’s membrane endothelial keratoplasty, or DMEK, a procedure during which only Descemet’s membrane is removed. The key will be a comparison of how the results with DMEK compare with thin-tissue DSEK, Dr. Holland said.
A study reported in the May 2009 issue of Archivos de la Sociedad Española de Oftalmología regarding two-year results with DMEK in 10 patients found average endothelial cell density of 2,039 cells/mm2 at six months, 1,925 cells/mm2 at one year, and 1,730 cells/mm2 at two years. Three eyes had complete detachment of the tissue, and DSEK was then performed.
In another report from the same investigators in the July 2009 issue of Current Opinion in Ophthalmology, they wrote, “Compared with DSEK/DSAEK, DMEK may have higher clinical potential with 75% of cases reaching 20/25 or better within 1-3 months. Furthermore, preparation of isolated Descemet grafts does not require large investments and may increase overall donor tissue availability.”
However, they also added that the surgery is accompanied by lower visual acuity and relatively slow visual rehabilitation, some limited accessibility to surgical equipment, and an early pre-op drop in endothelial cell density.
“In theory, DMEK is the perfect procedure. In practice, there are major challenges with it,” Dr. Tan said, citing the difficulty of performing the procedure and a current lack of consensus on how to perform it.

Editors’ note: Dr. Kim has financial interests with OSI Pharmaceuticals (Melville, N.Y.). Dr. D’Aversa has no financial interests related to his comments. Dr. Tan is the inventor of the Tan EndoGlide. Dr. Holland has financial interests with Alcon, Allergan, and Bausch & Lomb (Rochester, N.Y.), among other ophthalmic companies.

Contact information

D’Aversa: 516-374-4199, gdaversa@ocli.net
Holland: eholland1@fuse.net
Kim: 919-681-3568, kim00006@mc.duke.edu
Tan: +85 6227 7255, snecdt@pacific.net.sg