Better Before or After?

Ophthalmologists know corneal neovascularisation increases the risk of graft rejection and failure after penetrating keratoplasty. Available evidence documenting the various treatments for inducing regression of existing corneal vessels are at least partially effective. Yet, treating corneal neovascularisation before keratoplasty remains controversial among cornea specialists who hold differing views.

In a recent debate, Irit Bahar MD, MHA acknowledged there is no gold standard approach for treating corneal neovascularisation. She observed that some methods are still experimental, so long-term follow-up and prospective studies are mandatory. However, Dr Bahar said she typically treats it before keratoplasty because she believes it simple, safe, and at least partially effective for inducing blood vessel regression and improving graft outcomes.

“Even partial regression is better than none, especially when the risk of the treatment is low and the side effects are negligible,” she said. “And I promise you that the vessels will not disappear by simply ignoring them.”

Lack of data?

Taking another view, Jennifer Li MD said she does not treat corneal neovascularisation before keratoplasty, citing the lack of long-term data and safety concerns.

“Most of what has been tried had limited and variable success long-term. In particular, there are limited data for newer or combination therapies, and some of the potential side effects of the various options terrify me,” she said.

“If the benefit of the treatment [for inducing vessel regression] lasts only three to six months, how much effect will we see that it has long term?”

Rationale for treating

Dr Bahar noted the long list of treatments tried over the years to induce corneal blood vessel regression before transplantation indicates there is no ideal method. However, each has induced some vessel regression and some to improve high-risk corneal graft surgery outcomes.

As an example, Dr Bahar cited a report describing five patients treated with corneal cross-linking before penetrating keratoplasty—among whom there were no cases of revascularisation, graft opacification, or immune rejection during 4 to 42 weeks of follow-up.

Graft survival results better than expected were also reported by investigators reviewing their experience with fine needle diathermy combined with subconjunctival bevacizumab. Very recently, promising results were obtained using mitomycin intravascular chemoembolisation to treat corneal neovascularisation.

For more than a decade, Dr Bahar said she has been injecting bevacizumab 2.5 mg/0.1 mL per affected quadrant at the limbus at the end of the graft procedure in high-risk cases. She reported that in a series of 20 patients, vessel regression was observed in all eyes, 19 grafts remained transparent for 24 months, and no side effects were observed.

In a recent case, she injected mitomycin-C into superficial vessels before placing ligation sutures over the two main feeder vessels because she observed only partial vessel bleaching. She presented images showing an “impressive outcome” after two months.

Reasons not to treat

Dr Li reviewed the efficacy and safety of various corneal neovascularisation treatments to support her reasons for not treating corneal vessels beforehand. Any benefit of laser thermal cauterisation, she said, is limited mainly to efferent vessels. A 37% recanalisation rate at three months has also been reported. The inherent destructiveness of the thermal procedure risks collateral damage to the cornea, limbal stem cells, crystalline lens, and iris. It also risks damage that can promote collateral vessel development.

She said photodynamic therapy appears to be the most effective of the laser treatment options and is relatively safe. However, it also induces a rebound effect secondary to enhanced VEGF expression and is very expensive and time-consuming.

A frequent need for retreatment is also a limitation of fine needle diathermy, which can be associated with significant adverse events and the potential to stimulate new vascularisation. Treatment with anti-VEGF agents has gained attention based on their use for retinal vascular diseases. But Dr Li warned of a limited window for therapeutic utility because they are not effective in clearing mature blood vessels.

“VEGF inhibitors used topically have limited penetration and so are not very effective on deeper vessels,” she said. “Subconjunctival injection is invasive, has limited effect on vessels in the central cornea, and associated with a 35% recurrence rate in one study. In addition, adverse events can occur, including corneal stromal cytotoxicity with thinning and melting.”

Adding intravascular mitomycin treatment of corneal vessels has featured heavily on the Cornea Society’s online re­source, Dr Li agreed the preliminary results are very promising.

“However, data on this modality is very limited,” she said. “There are questions about whether it can be used with corneal transplantation, and there is the potential for some severe adverse events with mitomycin.”

The debate took place at the AAO annual meeting in Chicago, US.

Irit Bahar MD, MHA is Chief of the Department of Ophthalmology, Rabin Medical Center, Petah Tiqwa, Israel, and Full Professor of Ophthalmology at Tel Aviv University, Israel. iritbahar@gmail.com or iritba2@clalit.org.il

Jennifer Li MD is Director, Cornea and External Disease Service and Professor, Department of Ophthalmology, UC Davis Health Eye Center, Sacramento, California, US. jyhli@ucdavis.edu