Newer Corneal Transplant Techniques Linked to Worse Outcomes

Veronica Hackethal, MD

May 15, 2014

Penetrating keratoplasties have better survival and visual outcomes than newer lamellar techniques, without much evidence for better outcomes among experienced surgeons, according to a study of the Australian Corneal Graft Registry (ACGR), published in the May issue of Ophthalmology.

"The new procedures of lamellar keratoplasty...have been adopted enthusiastically by corneal surgeons worldwide," the authors write. "Although outcomes have been promoted as being significantly better than those of the well-established alternative of penetrating keratoplasty, the evidence for this claim is unconvincing outside of single-center studies."

Using the ACGR, the researchers looked at 13,920 penetrating keratoplasties, 858 deep anterior lamellar keratoplasties (DALKs), and 2287 endokeratoplasties performed between January 1,1996, and February 21, 2013.

The investigators found an increasing number of corneal grafts annually, with changes in practice patterns occurring after 2006. From 1996 to 2006, the number of corneal grafts remained stable (mean, 926). However, the number increased over the course of the next 6 years, reaching 1482 in 2012, revealing a need for 264 additional corneal donors throughout Australia.

The increased number of procedures correlates with a shift in practice, with surgeons performing an increasing number of DALKs, endothelial grafts, and pseudophakic bullous keratopathy after 2006 and a declining number of penetrating grafts.

However, when the investigators looked at outcomes, they found little benefit with the newer techniques when the procedures were matched by era and indication. Penetrating grafts had significantly better survival (P < .001) and visual outcomes (P < .001) compared with DALKs performed for the same indication during the same period.

Endokeratoplasties had significantly worse survival than penetrating grafts performed for the same indication during the same time (P < .001). Compared with penetrating grafts, endokeratoplasties had significantly worse visual outcomes when performed for Fuchs' dystrophy (P < .001), but they had significantly better visual outcomes when performed for pseudophakic bullous keratopathy (P < .001).

Surgeons with more than 100 registered keratoplasties had better survival of endokeratoplasties (P < .001) compared with less-experienced surgeons. However, keratoplasty failure occurred even after 100 grafts among experienced, high-volume surgeons.

Limitations included those inherent in registry studies, which affect the ability of this study to provide evidence equivalent to a randomized controlled trial. In addition, the study took place in Australia, which could limit the generalizability and acceptance of the results. Nevertheless, the ACGR affords opportunities to overcome some of the limitations of existing keratoplasty studies, such as small sample size and short follow-up. Established 30 years ago, the advantages of the ACGR include its long-term follow-up, prospectively recorded data, large number of patients, and real-world data drawn from a broad variety of surgical practices.

"When the ACGR began reporting the results of penetrating corneal transplantation in the 1980s, there was some disquiet around the world about the results," commented first author Douglas Coster, DSc, FRANZCO, emeritus professor in the Department of Ophthalmology at Flinders University, Bedford Park, Australia. Dr. Coster, a corneal surgeon, has more than 40 years of experience with keratoplasty.

"The usual explanation was that Australian surgeons were not good at the procedure, but as registries were set up in other countries, the same results were reported as those from Australia," he told Medscape Medical News.

In an invited editorial, Sanjay Patel, MD, from the Mayo Clinic, Rochester, Minnesota, and colleagues suggest past studies may not reflect what really happens in clinical practice.

"[These] are potentially the best outcomes data for comparing penetrating keratoplasty to endokeratoplasties, and penetrating keratoplasties to DALK, given the lack of large, multicenter randomized trials," Dr. Patel told Medscape Medical News. "The data tell us that every surgeon should assess their own outcomes and not assume that their outcomes will be the same as those reported by the surgeons who have pioneered the techniques and strived for the best outcomes."

Dr. Patel also noted that despite the lack of a definitive learning curve effect in this study, variability between surgeons likely does play a role in outcomes.

"Penetrating keratoplasty is still a good procedure and, for some surgeons, could be a better procedure than endokeratoplasty or DALK," Dr. Patel noted, "Endokeratoplasty is accepted as standard of care in the US, and probably most of the rest of the world now, for endothelial disease. DALK is accepted in some parts of the world as the standard, but potentially not in the US.

"I would say the cornea community has moved towards endokeratoplasty as a procedure of choice over penetrating keratoplasty, but the type of endokeratoplasty probably requires a RCT to know which is better, and we do not have that RCT," he continued.

Dr. Coster, doubtful this study will change clinical practice, also thinks surgeons should continue to make their own choices.

"This study will not influence the view of ophthalmologists about the standard of care or how they advise their patients," Dr. Coster said. "There is a widely held belief that the new lamellar procedures are an advance on penetrating keratoplasty. There is not the accumulated evidence[, however,] to ascertain whether the newer lamellar procedures are more effective than what has come before."

"Hopefully the study will encourage clinicians to begin to think about two things," Dr. Coster emphasized, "First, that single-center, noncontrolled observational studies from well-resourced experienced surgeons, who have encouraged the uptake of these procedures, may not extrapolate to the broader community. Second, there may well be a place for some form of postsurgical surveillance of new surgical procedures, [which] does not exist at present."

Supported by DonateLife and the National Health and Medical Research Council, Canberra, Australia. The authors and editorialists have disclosed no relevant financial relationships.

Ophthalmology. 2014;121:977-987. Article full text


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: