Finding alternative techniques for eyes with failed PK

Cornea surgeons encountering a patient with failed penetrating keratoplasty (PK) will wonder if they should repeat the PK.

Although there are some exceptions to the rule, according to Donald T.H. Tan, MD, the answer is “no.”

Speaking at Cornea 2017, Dr. Tan provided an overview of studies evaluating outcomes of different techniques for managing eyes with a failed PK.

“As the learning curve for Descemet-stripping automated endothelial keratoplasty (DSAEK) and now Descemet membrane EK (DMEK) allows us to reduce complications and rejection rates of repeat grafting, it is likely that less repeat PKs will be performed,” Dr. Tan, professor of ophthalmology, National University of Singapore, Singapore.

“As we get better at EK, there is the potential for less rejection and enhanced long-term graft survival [with the EK procedures],” Dr. Tan said.

“In addition, the Boston Type 1 Keratoprosthesis (Boston KPro) now appears to be a truly viable alternative to repeat PKs, providing improved intermediate-term survival rates,” he said. “However, we need further follow-up to determine the long-term success rates of the Boston Kpro as compared to EK surgery.”

Depressing data

Findings from the Singapore Corneal Transplant Study (SCTS) and the Australian Corneal Graft Registry (ACGR) show how graft survival rates for PK decrease progressively over time. The data also show how the graft survival rates vary depending on the indication, but it is clear that repeat PK carries a poor long-term graft prognosis, Dr. Tan said.

“Furthermore, there are data showing that high-risk PKs (e.g., repeat PKs) have even lower survival rates than liver transplants,” he said.

“We always thought the cornea had immune privilege, but this is the reality check we have for PK,” he said. “What are we doing performing repeat PK?”

Pros and cons of DSAEK

DSAEK may be considered for failed PKs because it potentially has a lower risk of rejection and higher long-term graft survival.

As a caveat, however, DSAEK cannot be performed if there is significant pre-existing, irreversible stromal damage or distortion in the PK, Dr. Tan said.

Dr. Tan used the term “potentially” because there are some conflicting data in the literature. A study by Anshu et al. that included 60 cases that had DSAEK after failed PK reported a DSAEK graft survival rate of 74% at 4 years. The rate reached 96% in eyes without a glaucoma shunt, but it was only 22% in those with a glaucoma shunt.

In contrast, the ACGR showed PK may deliver a better graft survival rate than EK after a failed PK that was initially performed for keratoconus or pseudophakic bullous keratopathy.

“But you have to understand the setting of the Australian registry, and so we need to remove the effect of the DSAEK learning curve,” Dr. Tan said. “In our own study, comparing PK with DSAEK and anterior lamellar keratoplasty (ALK), DSAEK and ALK had significantly better graft survival than PK.”

Looking specifically at previous failed PKs, data from the SCTS showed DSAEK was far superior to repeat PK. At 5 years, graft survival rates were just 51% for repeat PK and 86.5% for eyes that had DSAEK.”

Role for the Boston Kpro

Results of study with mean follow-up of 24 months showed a much better graft survival rate after implantation of the Boston Kpro compared with repeat PK.

“Despite the good success rates with the Boston Kpro, AAO Ophthalmic Technology Assessment panel had cautious conclusions, noting that outcomes and complications worsened with increased follow-up time,” Dr. Tan said. “These complications can be quite severe and result in permanent visual loss, and so long-term results are needed.”

He added that in the AAO Ophthalmic Technology Assessment, it was also noted that patients with autoimmune conditions did worse with the Boston Kpro.

“Looking at that another way, the results are pretty good for eyes with the indication of failed PK,” Dr. Tan said.

Mitigating the cautions of the Ophthalmic Technology Assessment paper are some newer studies reporting on the Boston KPro.

Superimposing those data onto the Kaplan-Meier survival curves from the SCTS, Dr. Tan showed that the Boston Kpro has a much better graft survival rate than repeat PK, although the Boston KPro did not match DSAEK’s higher graft survival rate.

“As a caveat, the Boston KPro series includes eyes with all indications, not just failed PK, and that might mitigate for better results with the Boston KPro,” Dr. Tan said.

More relevant information may come from a systematic review and meta-analysis of the Boston Kpro versus repeat donor keratoplasty for corneal graft failure.

In this study, the probability of maintaining 20/200 or better visual acuity at 2 years was 42% in the PK group and 80% for the Boston KPro. At 5 years, 47% of eyes having repeat PK had a clear graft whereas the Boston KPro retention rate was 75%.

“These data indicate that survival with the Boston KPro is better than with repeat PK and similar to DSAEK, but still we need to see what happens with longer follow-up.

DMEK as another option

The opportunity to perform DMEK in eyes with failed PK also argues against repeating PK.

“If we are able to repeat a corneal transplant in an eye with a previous PK, most of the time it is because the PK failed due to allograft rejection. Then it makes sense to choose a procedure that clearly has the lowest risk of graft rejection,” said Dr. Tan, presenting data showing that DMEK and deep ALK have higher 5-year survival rates than PK and DSAEK.

“There are techniques that will help us to do DMEK for failed grafts, such as the hybrid DMEK,” he said.

Also telling are data from the SCTS that show a 5-year survival rate of 5 years after DMEK, which is higher than for the other keratoplasty alternatives.

In addition, the rate of secondary glaucoma after DMEK was just 10% compared with 24% for both DSAEK and PK.

The explanation is that DMEK eyes require less topical corticosteroid treatment postoperatively, Dr. Tan said.