Murat V. Kalayoglu, M.D., Ph.D.
Contributing Editor
The first experiments on keratoplasty took place in early 19th century, with Franz Reisinger attempting to suture corneal grafts after excising host corneal tissue in rabbits. The experiments failed and the grafts were rejected; it would be decades until the first successful human keratoplasty. The procedure has evolved dramatically since then, and today an expert ophthalmic surgeon can perform penetrating keratoplasty in less than 30 min, significantly improve visual acuity, and prevent rejection with local steroids for many diseases. However, the procedure is far from perfect; a host of autoimmune diseases and chemical injuries that cause corneal scarring have high rates of rejection after transplantation. Furthermore, the operation requires sutures, with the potential of poor wound healing, irregular astigmatism, and suture-induced vascularization leading to vision loss. Since the entire cornea is removed, there is a chance of intraoperative expulsive hemorrhage, as well as infection. Advances in keratoplasty research have focused on improving the technique to reduce these risks.
One emerging technique – deep lamellar endothelial keratoplasty (DLEK) – relies on the observation that many causes of corneal decompensation occur as a result of endothelial dysfunction, and that replacing the endothelium without transplanting the entire cornea may be sufficient to restore vision. The technique offers the potential for highly predictable corneal power for extended periods without astigmatism. Furthermore, since the endothelium can be inserted through a small incision instead of an open-sky approach, the relative chance for infection compared with penetrating keratoplasty is minimal.
Endothelial replacement research dates back to 1993 when Ko et al. performed posterior lamellar transplantation experiments in rabbits. Posterior lamellar keratoplasty (PLK) was then attempted in cadaver eyes and primates through the mid-1990s, and Melles et al. reported the first successful human PLK in 1999. Since then, the technique has been further refined by Terry et al., who termed the procedure deep lamellar endothelial keratoplasty (DLEK). Theoretically, DLEK can be performed for nearly half of the 38,000 transplants performed each year in the United States, since nearly half of these procedures are due to vision loss from endothelial dysfunction. DLEK is a technically challenging procedure that relies on the surgeon’s ability to dissect into and through the corneal stroma, microns away from perforating through the cornea or into the anterior chamber. Both the corneal surface and limbus are preserved in order to maintain the integrity of the eye.
A small scleral wound is created 2 mm posterior to the limbus, and a Devers semi-sharp stromal dissector is used to dissect into the stroma at ~80% depth. Once the dissection is completed, a “Terry trephine” is inserted into the stroma and used for a trephination cut from the posterior stroma, through the endothelium and into the anterior chamber. The host stromal-endothelial interface is then stripped away and removed. Viscoelastic is used to facilitate the trephination and removal of host tissue, which is then evacuated from the anterior chamber. Finally, the donor stromal-endothelial interface, which has been dissected and prepared for transplantation, is inserted through the small opening and gently rotated to flip, endothelial side posteriorly.
Care must be taken to remove viscoelastic completely from the anterior chamber, since residual viscoelastic may prevent adequate adhesion of donor transplant to host stroma. The tissue self-adheres due to hydrophilic attraction between two hydrated tissues; adherence is facilitated by inserting air into the anterior chamber, which presses against the endothelial side to exert anterior pressure. The endothelial pump then creates a stromal suction to complete the adherence. The wound is closed with few sutures, and does not result in significant astigmatism since it was created 2 mm posterior to the limbus. To date, available data suggest that donor endothelial counts between DLEK and penetrating keratoplasty are similar at 3, 6 and 12 months post-procedure; however, long-term data for this new procedure are not yet available. Graft displacement or wrinkling is also a concern, but occurs surprisingly rarely.
DLEK represents a new frontier in corneal “component surgery”, replacing only the essential part of the diseased tissue. This minimally invasive procedure is a product of the drive to further refine an already delicate operation in order to obtain even better results. Technological advances will continue to advance the technique; in fact, surgeons are already examining if the femtosecond laser can be used effectively to dissect the deep stroma for DLEK, thereby obviating the need for mechanical dissection.