To Galen, Mundinus and Tulp, Add Dua

What exactly is Dua’s layer?

It’s a thick, tough, acellular layer of collagen found just before (anterior to) Descemet’s membrane in the cornea. When you blow air through the cornea, it travels throughout the whole stroma as tiny bubbles until it reaches Dua’s layer, which is lifted up as a dome: it is impervious to air. The collagen fibers here are smaller in diameter than in the adjacent corneal stroma, meaning that there is more space between them for gel-like proteoaminoglycans. This probably explains why it does not let air through.

Did you have an inkling that this additional layer was present?

Yes. Let me give you a bit of background. Corneal transplantation had been around for over a hundred years. Initially, no matter what the problem, the whole disc was cut out and replaced. The biggest risk of this procedure was rejection, which involves the endothelium. The logical response was that when the endothelium was not involved in the disease, it should be retained, not replaced. A gentleman called Mohammed Anwar introduced the use of air to separate the inner lining of the cornea, the Descemet’s and the endothelium, from the rest. The idea was that by replacing only the rest of the cornea, you would never get graft failure due to endothelial rejection. Bingo!

It was a great idea and even though results were unpredictable as the bubble that separates the layers did not always form, when it did separate the Descemet’s membrane the outcome was good. This procedure was referred to as a ’Descemet’s baring technique’. Every textbook says that the technique separates Descemet’s membrane. On occasions, big bubbles in different planes were observed and the explanation offered was that this was due to a split between the banded and non-banded zones of the Descemet’s membrane. However there were certain observations that did not add up and provided clues that the big bubble separation was occurring in a different plane. When the air bubble doesn’t extend to the trephine mark one has to separate the stroma from the underlying Descemet’s membrane manually in that sector. During this separation one often sees strands of tissue extending between the stroma and the membrane. Yet, when you transplant the donor in this operation, you take off the donor’s Descemet’s membrane, which peels off very smoothly and there are never any strands. That distinction was an important clue.

Another was that when you stitch the tissue of a full-thickness cornea into the recipient, you see the edge of the Descemet’s standing very proudly as the needle passes just anterior to the Descemet’s through deep stroma and we avoid puncturing the Descemet’s. But in deep lamellar transplantation, the donor’s Descemet’s is removed and we still see the edge when placing the sutures: there must be something else producing this sharp edge. In addition, many surgeons have commented that after a deep lamellar graft the eye is much stronger than following a full-thickness graft. Descemet’s is not strong enough to impart this strength, so that was another clue. Enough for us to decide to investigate that.

Was there a race to characterize this layer?

Not quite. It didn’t click with others that there was another layer. The stromal tissue attached to the Descemet’s membrane after failed operations was commented upon, but no-one latched on to the idea that it was very different to the rest of the stroma and might be a different layer relevant to the surgical anatomy of the cornea. 

What was the reaction when you announced your findings?

We submitted the paper for publication in August 2012, and in September 2012 I presented it for the first time at the EuCornea meeting in Milan. It was the EuCornea medal lecture, in front of an audience of 800 corneal specialists from across the world, including some leading names. Prominent specialists told me, “I have learned so much about my operations which I didn’t understand” and that it was “So clear, the clinical implications, the anatomy, and the science”. Donald Tan, president of the Cornea Society and of the Asia Cornea Society, who was in the chair that day commented “anatomy text books will have to be rewritten”. At another talk, I was told, “You know, I’m not surprised you discovered it, I’m disappointed I didn’t, because I see this under my eyes every week but I didn’t catch on to what it was”. People who perform the operation relate to it immediately: people who don’t, take a little while to grasp it.

The media interest has been phenomenal. I’ve had a lot of emails of congratulations from all over the world. For example, a group from Namsos, Norway (Anita Blixt Wojciechowski and Astrid Meistad) sent me a picture of a six-layer cake that they baked to honor the discovery. To the cornea’s known five layers they added a sixth, in green, to represent Dua’s layer.

Extraordinary that we are still making discoveries about the morphology of the eye!

Yes, it is. Some people argued, and rightly so, that since we only showed it in adults (the mean age group of our patients is 77 years) it may not be a true layer: is it present in children? they asked. When I was in Thiruvananthapuram in India a young corneal surgeon, Vinay Pillai, showed me OCT and histology images from the cornea of a nine-year-old girl who had a failed deep lamellar graft, and you can clearly see this layer, beautifully illustrated. Personally I’m convinced. It is different from the rest of the cornea and is very, very relevant to surgical anatomy.

Will it change clinical practice?

I think it changes clinical understanding tremendously. Where it could change clinical practice is in the cases where we get mixed bubbles and Descemetic bubbles. In two out of ten occasions we get the bubble between this layer and the Descemet’s membrane. The Descemet’s membrane bubble is very delicate and less able to withstand pressure and simple handling maneuvers. Previously, we didn’t recognize the issue and during the operation the Descemet’s bubbles have burst, forcing conversion to a full-thickness graft. It should now be possible to tell intra-operatively which type of bubble we’ve got and take necessary precautions to prevent a Descemetic bubble bursting. It will actually improve outcomes, it will make the operation a little safer.

Clinically, there are certain diseases of the cornea where this layer may have a role to play. One is keratoconus, where the cornea becomes more and more conical in shape and in some instances the Descemet’s membrane tears and suddenly the cornea gets hydrated, a condition called acute hydrops. We hypothesize that the tear is not just in the Descemet’s, but also in this layer. Two cornea colleagues from India, Rajesh Fogla and Mohammed Shahbaaz, have sent me images and videos of cases of macular dystrophy of the cornea, where they have performed a successful big bubble separation and demonstrated that the stromal opacities also involve this layer. In such situations, one may consider peeling this layer off as well, which has been accomplished though it is tricky. I’m getting videos sent from colleagues all over the world where they are encountering things. This will condense the time it will take for this layer to affect our understanding of diseases and their treatment.

How does it feel to join Achilles, Fallopio and Langerhans in having an eponymous body part?

I think it is mixed feelings, because in a way it’s embarrassing. When I presented the early data in 2007 and when we wrote the paper, I called it “A Novel pre-Descemet’s Stromal Layer”. My colleagues suggested “Dua’s layer” and we included it in the title in brackets. Everybody just picked up on that. Normally it is for your peers to ascribe the eponym.

Do you think we now know the complete anatomy of the eye, or are there other surprises in store?

Well, we are preparing a follow-up paper, an extension of the anatomy... It won’t be as exciting for the lay press, but will interest the scientific community. There’s always something new. Simple things are still there to be discovered. Our understanding – what we take for granted now – is, I think, not complete.

How do you balance research, clinical work, administrative work and family life?

I’ve found myself sleeping less, four or five hours a night. The thing is, if you enjoy what you’re doing, then you want to do more. On Saturdays, I spend the morning doing BJO (British Journal of Ophthalmology) work, then I get on the treadmill for some exercise, go into the office and in the evening, I have my social bit. To quote, “If you enjoy what you’re doing, you don’t have to work a single day in your life”.

Harminder Dua is the President of the Royal College of Ophthalmologists, Chair and Professor of Ophthalmology at the University of Nottingham and head of the Division of Ophthalmology and Visual Sciences.