LIMBAL DEFICIENCY

One thing Harminder S Dua MD, PhD does now that he didn’t in years past is delay surgery for limbal deficiency in patients who retain a central island of corneal epithelium and reasonable vision (Figure 1). While a good supply of limbal stem cells is essential to replace corneal epithelium once it is lost, his research suggests the limbus plays a minimal role in normal physiological corneal epithelial homeostasis.

“It is only in response to injury that the limbus comes in a big way,” Prof Dua, of Nottingham University, UK, told the 4th EuCornea Congress in Amsterdam. He has observed central corneal epithelium remain clear and stable for years in patients with total limbal stem cell deficiency and conjunctivalised epithelium in the peripheral cornea. What’s more, these patients often see quite well. “Limbal deficiency is compatible with good vision in patients where a central island survives. Even if I see it gradually shrinking over time, I will not rush in for surgery until vision is seriously compromised,” said Prof Dua. He presented several emerging limbal deficiency treatments based on a growing understanding of the roles the limbus and conjunctiva play in generating and maintaining corneal epithelium.

One challenge of autologous and living relative donor limbal grafts is they only partially surround the cornea, leaving wide paths for conjunctival epithelial ingrowth. This blocks corneal epithelium growing from limbal grafts from fully covering the corneal surface. Sequential sector conjunctival epitheliectomy (SSCE), can solve the problem. But while periodically scraping away competing conjunctival epithelium does help corneal epithelium spread, it can be difficult and painful, and risks complications such as bleeding, Prof Dua said.

Amnion assisted conjunctival redirection (ACER), can eliminate the need for SSCE, Prof Dua said. He places an amniotic membrane over the cornea after the limbal grafts are attached, tucking the membrane edge under the conjunctiva, but leaving the edge over the limbal grafts. This directs conjunctival epithelium growth over the membrane, allowing unobstructed transplanted limbus derived corneal epithelium growth beneath the membrane.

In two to four weeks, corneal epithelium covers the entire corneal surface while conjunctival epithelium covers the outside of the membrane, Prof Dua said. Often the membrane then separates spontaneously, but it also may be removed surgically.

“This is what I do now standard in all my autologous and living related grafts so I don’t have to repeatedly brush or scrape advancing conjunctival epithelium” he said.

Conjunctival alternatives

Keeping conjunctival epithelium off the cornea is usually desirable. But if the alternative is no epithelium at all, it may be better to let the conjunctiva grow, Prof Dua said. “Any epithelium is better than no epithelium. Even if it is very thin cornea, if it is epithelialised the risk of melting and perforation is low.” In some cases, the conjunctival epithelium may even develop into a phenotype indistinguishable from corneal epithelium, Prof Dua said. He discovered this in a patient with a chemical burn that had destroyed the entire corneal and limbal surfaces along with much of the conjunctiva. Epithelium began growing from a remote undamaged portion of the conjunctiva, so Prof Dua let it grow.

Within a month, the entire cornea re-epithelialised with what appeared under confocal microscopy to be normal corneal epithelium – even though the limbus was destroyed, Prof Dua said. “We saw it happen before our eyes and on long-term follow up the cornea remains completely clear.” Prof Dua has now seen the effect in five eyes of four patients. How it happens is not known. Conjunctiva transforming into corneal phenotype is a possibility. Or palisade-like structures found in bulbar conjunctiva of pigmented individuals may provide stem cells much as do palisades near the limbus. Or conjunctiva growing onto the cornea without vascularisation may produce epithelium of a more normal phenotype in the absence of limbus. “I look at the conjunctiva very differently now, as a friend rather than a foe,” Prof Dua said. Another use of conjunctiva is free autografts when corneal melting is not responding to other measures, and as an alternative to tenoplasty, Prof Dua said.

In a patient with a chemical burn and an amniotic membrane that fell off, Prof Dua took a conjunctiva graft from the fellow eye, and stitched it to pink (viable) tenon’s tissue. This resulted in vascular conjunctiva covering the corneal surface. The conjunctival graft vessels establish connections with the tenon’s vessels thus bringing blood to the ischaemic limbus. When the eye was quiet, he removed the conjunctiva and performed an auto limbal transplant. The result was a clear cornea with 20/60 vision unaided, and 20/30 with a rigid gas-permeable contact lens.

Peripheral cornea and limbus transplant?

In some cases, Prof Dua transplants peripheral cornea from donor cornea-scleral discs to preserve central cornea in patients for which limbal donors are not available. In one case he observed both the peripheral and central cornea survive with sustained epithelialisation. In another patient the epithelium over the transplanted peripheral corneal strip was lost, but conjunctival epithelium grew in to replace it and continued to migrate centrally to cover the corneal surface. Central corneal epithelium was maintained, apparently by cells migrating from the recipient conjunctiva across the donor peripheral corneal stroma.

“The question I am asking myself is, if I just have the niche, which is the stroma of the peripheral cornea and the distant sclera, and the cells come from the conjunctiva on it, do they change phenotype? Does the niche influence the cells to start coming off the other end as more cornea-like cells, or is it just conjunctiva coming off and behaving like cornea?” he said.