A round table on aqueous outflow

Glaucoma researchers and clinicians discuss collector channels, outflow and the future of MIGS.

Davinder S. Grover, MD, MPH: When MIGS (minimally invasive glaucoma surgeries) first came out, they were specifically [safety-focused] angle-based procedures, but now we’re seeing that role evolve with newer devices that aren’t just based in the trabecular meshwork and the Schlemm’s canal. Some suprachoroidal shunts and subconjunctival filtration surgeries are being termed MIGS. So it’s not just angle-based surgery anymore. Improved safety with MIGS has lowered our threshold to operate and allowed us to intervene at an earlier stage.

We are lucky to have a very impressive collector-channel brain trust here with us today, and I am curious to know their thoughts about MIGS and outflow. To start, from an outflow perspective, are there potential advantages to intervening early?

Alex Huang, MD, PhD: In terms of moving early, I think your introduction was spot on. Safety has to be a priority in order to intervene early because the vision is often good and all glaucoma interventions have potential complications. Then when you combine safety with something that is potentially effective in lowering IOP plus gaining greater topical medicine independence, then you have something special for the patient. Basically, I think that MIGS created, in the last 10 years or 15 years, a re-invigorated re-focus on the outflow problem. We always knew that outflow was there, that resistance is a problem and the trabecular meshwork mattered, but at the same time it was more complex. MIGS and earlier intervention brought greater clinical attention which then became the inspiration for deeper investigation into normal and diseased biology.

Dr. Grover: Murray Johnstone, what are your thoughts on that?

Murray Johnstone, MD: The whole issue of intervening early is one that has been around for a number of years. Even many years ago when people did gonioscopy showing that blood reflux in Schlemm’s canal, they realized that early in the disease it would move in and out of the canal very rapidly, but as the disease process increased or worsened, the movement was much slower and eventually the adhesions between the walls of Schlemm’s canal seemed to occur. They pointed out that probably what we’re dealing with is a vicious cycle where, as the pressure goes up, the meshwork comes to the external wall of the canal, becomes oppositional and then perhaps synechial. And even then, they suggested that probably if we had safe procedures we would be better off intervening much earlier, getting the pressure down so we didn’t get into this vicious cycle situation. So I think the conceptual framework is there and some very recent evidence suggests that we can actually look at what’s occurring, particularly with the new OCT technology, when the walls of the canal come into that position and develop adhesions.

Sayoko (Sy) Moroi, MD, PhD: I think the MIGS procedures have been transformative from the standpoint of safety. We all raise questions regarding efficacy and understanding the role of these newer techniques at different stages of disease. But I think foremost the Schlemm’s canal-based surgeries allow us to ask more questions about what structures contribute to the distal outflow resistance. A tremendous amount of knowledge has been gained about the trabecular meshwork, so now we can focus on the distal outflow. Now is the time to learn more about the distal outflow structures, especially at the early stage of disease, and then learn what happens in its late stage so that we can prevent those changes from happening.

Dr. Grover: Sy, you’re doing great work on imaging distal outflow with fluorescein tracers during canaloplasty. Have you seen a relationship with that and stage of disease?

Dr. Moroi: Anecdotally, yes. A colleague and I are in the process of quantitating vessel density revealed by the fluorescein canalograms recorded on high-definition video. Once we get the quantitative data, we will analyze if the vessel density correlates to IOP lowering outcome.

Ronald L. Fellman, MD: It’s a good point that Dr. Grover made about getting to things early because in the past, especially 30 to 35 years ago, our glaucoma surgeries could be detrimental to the patient’s vision and we were hesitant to do anything until we had to. And operating earlier because of MIGS safety, prior to advanced disease is just a total paradigm shift in thinking about how we handle this relentless disease. Now we are able to tell patients, early on, that we have a chance of increasing flow in their natural drain, and patients favor that — something natural without creating something artificial. We have to communicate that we are limited in determining the capacity of flow through their natural drain, and if the surgery doesn’t work, they may need more aggressive surgery. So it’s a total mind shift. We need to intervene earlier before the disease process gets so refractory that our MIGS outcomes are hindered by loss of elasticity, adhesions and ultimately atrophy of the collector system.

Dr. Grover: That’s a great point. Alex, you’ve touched on how the disease process can possibly change outflow patterns. In your experience with the anterior segment OCT and anterior segment angiography, have you seen evidence of that?

Dr. Huang: We’ve just started looking into the disease process using animal models of glaucoma. In the beginning when we developed the angiography, the spirit was to follow along your guys’ footsteps in terms of developing a method that was compatible for humans, like the episcleral venous fluid wave. To accomplish that using angiography, first we started in the lab with post-mortem animal and human eyes, then moved into live animals, but when we finally imaged live humans, we started with normal individuals who needed cataract surgery. They had a reason to be in the OR anyway. We haven’t imaged glaucomatous people yet, but that is part of the plan.

In parallel and as a separate offshoot I mentioned starting to look at the disease process using animals models. I met some really astute clinicians and vision scientists who were veterinarians. Animals get glaucoma too, and they’re treated by vets who are really sharp individuals. I went to the University of Wisconsin, it was Gillian McLellan in the vet school who takes care of glaucoma dogs and cats. So we started with glaucoma dogs and we saw a very marked impact where angiographic outflow was diminished in dogs with high pressures. Therefore, it seemed that disease did change the outflow pattern.

No more tricks: the need for a tracer

Dr. Huang: With anterior segment OCT, we’re lacking a reference function. In retina OCT, they have tracking and a reference function so that next week, two years, even later we can go back. The OCT finds itself and can scan the same location. And to try to get at the pathology maybe these various septa that we’re trying to name, you have to go back to the same place after you do something. But I can never be sure I’ve done so before and after an experimental manipulation.

Dr. Moroi: So what you want is image registration, right? That’s what we need.

Dr. Fellman: Yeah, collector channel image registration.

Dr. Moroi: Well, but those change.

Dr. Fellman: Oh, you mean, in other words, you could see it one time and not the next?

Dr. Moroi: Correct.

Dr. Huang: Sure, maybe there’s change from disease or natural biology.

Dr. Fellman: So, what would be a reference point?

Dr. Moroi: Perhaps you can lock into some unique iris or limbus anatomy, but if this changes post-surgically, then that’s a challenge. But we have observed that some individuals have very dominant aqueous vein, perhaps that’s something you can lock in and do image registration.

Dr. Fellman: That’s an idea.

Dr. Johnstone: With the newer OCT, they’re able to do a 3D volumetric scan so they see it all. You just do a little region and it’s almost like a fingerprint or an iris print. Probably within a couple years you’re going to see that fingerprint and 3D volumetric arrangement. I think we’ll get there with the OCT’s ability to do that and follow it.

Dr. Huang: I think it’s super-important because that’s what allows you to do the study; when you have the intervention, how do you know how it responded? You do your OCT a little bit differently and then you get tricked.

Dr. Johnstone: How do you tell what’s what? You go over with the technician and you think yes, it’s right there; that little river right there. And then you wonder, is that aqueous or blood? Is there a way to tell the difference?

Dr. Huang: That maybe back to the OCTA, right? OCTA allows you to see blood or something that moves. In the aqueous there’s no blood, so OCTA doesn’t work the same way as it does for retinal vasculature. So, if you wanted to have non-invasive aqueous imaging, what we need is to find something endogenous in the eye that’s at a high level. Then develop an imaging technique for that. In this case, you don’t need fluorescein or ICG anymore. The only thing that I can think of is vitamin C. Patients get ulcers, you give them vitamin C. If you had vitamin C imaging, you have noninvasive aqueous imaging because it’s at 100 times higher levels in the aqueous compared to blood. But vitamin C is a bland molecule that I think you can’t do a lot with. However, if you have something native that has high levels compared to blood, then you have it.

Dr. Fellman: So just hoping the technology will allow us to differentiate aqueous from the blood mixture in there is not going to cut it without something else helping us.

Dr. Moroi: You need a tracer.

Dr. Fellman: Otherwise, you don’t know, right?

Dr. Johnstone: I think there is a way. With OCT, it follows blood on the surface; phase-based (PhS) OCT with Doppler OCT follows that. But, in association with that, the intrascleral vessels, and Ricky Wang, PhD, has done this, you can see there’s a void, a tissue void that’s clearly vessels. They show that in the lymphatics. So, they can look at the void and see that as an area of vessels without blood. Also, the vessel walls change shape with every pulse wave. The problem is we don’t have the time resolution or the other features yet to make that happen. But, I can envision that, since you can see the image void, you could see the walls changing as technology advances.

Dr. Huang: To tie the structure to the flow.

Dr. Fellman: If you don’t see the wave, if you don’t see the pulse wave in the wall, then it’s probably not aqueous. Is that how you would think of it?

Dr. Johnstone: Dr. Wang’s lab has shown a tissue void by SD-OCT and by registration of PhS-OCT showing a connection with blood-containing vessels. But you can see with PhS-OCT the blood moving through the vessels. And they can see vascular voids. They’ve shown this in lymphatics. But they don’t have the time resolution available to look at vessel wall changes; these are sort of more single shot deals.

Dr. Fellman: Okay, so, it’s not there yet.

Dr. Johnstone: No, it’s not there. I’m talking about where can we go.

Dr. Huang: Talking about funny tracers, there was a paper or a poster at ARVO a couple years ago where they used whole milk as an OCT contrast agent. Whole milk and then OCT. Interesting approach, but I don’t think we’re really going there.

Now animal glaucoma is separately interesting because high pressures is still practically part of the definition. You almost can’t take it out because we don’t have a clear normotensive glaucoma situation, particularly for domestic pets. One doesn’t usually see cupping or find a field defect on “routine exam.” That’s why it was easier to start with animals to look for an impact for disease on outflow.

Please peer into the crystal ball ...

Dr. Grover: I was wondering if we could do a short synopsis of where you think we’re going with outflow and our understanding of outflow over the next five years, and where you want us to go.

Dr. Huang: Two directions: surgeries and drugs. On the surgery side, it was lower-hanging fruit. Let’s just talk about conventional because that’s the therapeutics that we have available to us, at least initially before the uveoscleral ones came along — how we do the surgery and how you place the surgery manner and either structural information or functional information of fluid flow. Will that lead you to better placement?

But then there’s still the pharmacological side that we could think of, especially with these involutional topics we’re talking about. Some of the changes that we saw or observations that we had with angiography were that it isn’t a static set of pipes. That thing could actually change. That makes sense. These are living things that we can take advantage of. Are the drugs that we use all along, if we’re going to talk about the scleral spur and the trabecular meshwork complex, let’s say, with just pilocarpine, was that just homogenizing flow and just taking segmental things and opening everything up?

And then, through the technologies that we develop, can these become screening endpoints for new drugs? Only recently are we looking at FDA-approved drugs for conventional flow, and we haven’t had any forever. Is there further expansion there possible with other medicines in that we actually now have tools to test for them?

Dr. Johnstone: I think probably the first thing is to understand the system. It’s beginning to look like just an absolutely beautiful organ system that maintains a normal homeostasis. Understanding how it normally regulates and how it functions is important. Then, we can ask what’s going on in glaucoma, how that changes in response to the glaucoma problem.

The next step, of course, is to figure out how to reverse the changes. And I think that by developing better imaging technologies, we should be in a position to better understand the mechanisms and when people are getting into trouble and treat them more effectively earlier.

And then as Alex says, we ultimately want to move on to when we should do surgery: when we should do laser, when we should do meds. That will fall out best by developing the diagnostic tools to identify when problems are beginning.

Dr. Moroi: Aside from the structural and the imaging tests, I’m hoping we can understand more of the genetic regulation and age-related changes in the sclera. I believe this knowledge will give us opportunities for new treatments. Not only IOP-based treatments, but looking at connective tissue-based treatments that can counter or slow these bigger changes occurring in glaucoma and make things more viscoelastic or “better shock absorbers.” I think that’s the right direction rather than these “cardboard-stiff TM” responding to excess TGF-beta. Those are cell biology studies I look forward to and hope to see that evolve as a new therapeutic armamentarium and with tools for understanding the biomechanics of the tissue.

Dr. Fellman: When I read there are 200 genes involved, I needed an antidepressant right away because it’s so complicated. But I think we can do it.

Dr. Grover: That’s fascinating. As Alex alluded to, Ron, you were the first to describe the episcleral venous fluid wave. How have you used that to evaluate characteristics of outflow and to give you insight into the disease state?

Dr. Fellman: We’re learning more and more about it. For example, during Trabectome surgery where we cleave open the meshwork and the inner wall of the canal, thereby decreasing outflow resistance, during the I&A we saw in some patients remarkable flow through their episcleral venous system, which created a marked distal blanching of the episcleral vasculature distal to the surgical site.

Then, there were mixtures in between and there were some patients who had no flow, thus no blanching of their adjacent episcleral vasculature. We went back and looked at the video of a series of patients. We saw a correlation between four to six hours of blanching and a lower pressure with less medicine. The pressure I think was about 13 mm Hg in that group. It was 18 mm Hg in patients who had minimal flush or episcleral venography, if you want to call it that. But we lost that correlation at around six months because I think there are alterations in the outflow system based on wound healing, and other factors that we just don’t understand.

It’s exciting to do a canal-based procedure, see a big wave, then tell the family, “We saw a fluid wave and we think that’s a favorable intraoperative sign that we increased outflow through your natural drain.” To the patients who don’t have a wave at all, we say, “We didn’t see a lot of flow so that’s not great, but we’ll see how you do, because some cases still have a reasonable IOP drop.” However, in our study, one-third of the patients who had no wave had another glaucoma procedure within a year.

So, the episcleral venous fluid wave is part of the evolution of trying to understand what’s going on with the outflow system. For example, during an iStent, there’s not much circumferential flow in the canal that we can see with the wave. So we are learning, how else can you look into the system at this stage in the operating room? It’s certainly not something commonplace yet. Exploring the wave during subsequent surgery after failed canal procedures also teaches us what happens with wound healing.

Panel participants (from left): Alex Huang, Murray Johnstone, Davinder Grover, Sayoko (Sy) Moroi, Ronald Fellman.

“An increase in pulsatile flow [an inherent phenomenon] occurs in response to the miotics, to adrenergics and even latanoprost.” — Murray Johnstone

“Operating earlier because of MIGS safety, prior to advanced disease, is just a total paradigm shift in thinking about how we handle this ... disease.” — Ronald Fellman

Dr. Grover: Have any of you experienced that too, struggles with wound healing in the canal?

Murray Johnstone, MD: I think that’s an issue. Your group has made a great contribution in terms of identifying the fluid wave at the time of surgery. That’s very valuable, and it gives you a sense of what’s happening.

But there is also a question of mechanisms and how the distal outflow system is functioning. For many years, it has been known that aqueous outflow is pulsatile. And interestingly, as the glaucoma process progresses, this pulsatile flow phenomenon gradually slows, is less often seen and is completely absent in advanced glaucoma, suggesting there’s some sort of an underlying process occurring that stops flow. Interestingly, you can look at the aqueous veins and kind of watch the pulsatile flow and identify where it’s occurring. Most of flow occurs inferiorly, in the inferior nasal quadrants, and it is possible to identify where the vessels are. There are usually only two to three aqueous veins working at one time. You can see whether there is active pulsatile flow, get a sense of what the distal outflow channels are doing. I think that’s potentially beneficial as well because it can all be done before going to the OR.

But, Ron, I loved your comment a while back saying you sit down and start studying the aqueous veins and 10 minutes later the patient says, “Jeez, doc, what are you doing?”

Dr. Grover: The patients are probably thinking “Did my doctor fall asleep?”

Dr. Johnstone: Exactly. I’ve had that happenover and over.

So, all of this sounds good on paper, to sit and study the aqueous veins.

It’s very tedious. But the hope here is that we’ll have an OCT system, or perhaps other imaging technology and the sort of thing that Alex and Sy are working on too, that will allow us to identify the functioning veins in a simple way. An instrument will look at it, identify what’s going on, where the flow is and be able to calculate or quantitate how much flow is occurring in different areas. I think we’ll have that kind of technology in the not-too-distant future.

Dr. Grover: Do you think certain MIGSbased surgeries can alter a pulsatile flow, or that that’s an inherent component to the eye?

Dr. Johnstone: The pulsatile flow is an inherent phenomenon that occurs. An increase in pulsatile flow occurs in response to the miotics, to adrenergics and even latanoprost. In the course of pressure reduction, the alterations in the pulsatile amplitude increase before any reduction in pressure. There’s good evidence that not only does pulsatile flow increase, it probably is altering something that’s going on in glaucoma that can be reversed to some extent transiently at least with our medications. So, I think it’s very real, yes.

Aqueous angiography performed in the left eye of a 60-year-old Asian female during cataract surgery. 0.4% indocyanine green (ICG) was introduced into the anterior chamber. Angiographic images were taken using a Spectralis OCT+HRA (Heidelberg Engineering). Centrally, the anterior chamber was filled with ICG. Posterior to the limbus, green arrows show areas of distal angiographic outflow; the red shows a region without.

Dr. Grover: Specifically latanoprost or what kind of medications?

Dr. Johnstone: It has been shown with pilocarpine and adrenergics. A couple of years ago, we reported that latanoprost does the same thing. We have some very definitive videos showing this. We looked at the response in five patients, all of whom had a pressure reduction on average 30%. All of them had marked increases in pulsatile flow. The pulse amplitude increase preceded the pressure reduction, which gave you a good sense that that’s what’s going on with a likely cause and effect relationship.

Dr. Moroi: How do you put together the complex mechanisms of these drugs and their effects on the TM (trabecular meshwork), the scleral spur and the ocular pulse?

Dr. Johnstone: A lot of studies show that miotics increase the tension on the scleral spur. That opens up the trabecular spaces. The process enlarges Schlemm’s canal, and, as the canal expands, it even enlarges the collector channels because there are attachments between the trabecular meshwork and collector channel entrances.

The adrenergics is more of a mystery, although there is some thought that it alters in some way the ciliary body functions as well. And latanoprost is even more interesting. There are actually quite a series of reports indicating that it is not just a uveoscleral flow agent. A number of papers report the flow increase occurs within an hour or two of the instillation and actually much quicker. The only way increased flow is going to occur is through some sort of vascular and muscular phenomenon. It can’t be a result of transcription that requires many hours to days.

Dr. Moroi: And just to make sure, you’re not talking about aqueous inflow. You’re talking about outflow, correct?

Dr. Johnstone: Yes, aqueous outflow. The reduction in IOP occurs very quickly, much more quickly than can be explained by the concept of the transcription of metalloproteinases with later changes in extracellular matrix permeability. So it seems pretty clear that there is another phenomenon going on. Johan Stjernschantz, the medicinal chemist who was perhaps the most instrumental in developing latanoprost, reviewed the quantitative evidence and associated video at an ARVO presentation. He suggested the early mechanism was unrelated to transcription and suggested a vascular phenomenon. That doesn’t mean that there might be a whole different series of phenomena occurring, so it’s only one part of the picture.

Dr. Huang: So, you’re thinking about pulsatile flow in terms of drug effects. Do you think the pulsatility comes from an actual outflow pathway component, or is it upstream? What I mean by this is that I always saw your data showing the time delay between outflow pulse and the digital pulse as evidence that the pulse amplitude came from the heart thereby linking the heart to the eye. Then, when we did the monkey studies in terms of angiographic outflow pulsatility, while we didn’t have cardiovascular data synched to outflow data, we saw that the angiographic pulse matched average monkey heart rates which supported your finding. So do you think these drugs impact pulsatility by working upstream on the cardiovascular side, somehow altering ocular blood flow into the eye or by directly acting downstream outflow pathways themselves?

Dr. Johnstone: There are four studies that show that when you introduce latanoprost, you greatly increase the ocular pulse amplitude, and people have wondered how that occurs. The investigators who report increased pulsatile flow have proposed that there is a relaxation of the choroidal vasculature so that each systolic pulse wave is able to introduce a much greater volume of blood into the choroid with a corresponding increase in ocular volume and a great pulse amplitude. That’s one piece of the puzzle.

Lamina cribrosa C-mode of healthy (A-B) and glaucomatous (C-D) eyes. SS-OCT offers enhanced penetration into the optic nerve, providing the ability to assess the lamina cribrosa microstructure.

But, there’s another downstream phenomenon. We reported a few years ago that brimonidine caused a marked increase in pulsatile flow by what must be an adrenergic effect. And what we found — these were normal subjects — was that IOP dropped to near or below what’s normally thought of as episcleral venous pressure. We took this to ARVO to report it and I was afraid that I’d be attacked for these findings, which you’re at great risk for. But, at the same meeting, a study by Jeff Kiel showed that if you cannulate the aqueous veins and measure the pressure change following brimonidine — it was really fortuitous this occurred — that within the same time course, the pressure dropped from about 8 mm Hg to 4 mm Hg in cannulated rabbits. A lot suggests that we’re working with both upstream and downstream issues.

Dr. Fellman: I know Sy has done work with the interesting part about, the average person makes 2.5 µl per minute of aqueous. But, from what I’ve understood from Sy’s work especially, is that for some people it’s like a bell-shaped curve, so it’s biologically normally distributed. Some people make 1.5 µl per minute normally and some people make 5 µl per minute. That was a shocker to me to even think about that and what effect that might have on medicines I use and maybe somehow related to outflow. And then she was teaching about precision medicine and kind of how that really all fit in.

I was thinking, can you actually measure uveoscleral outflow? No? Because, if I put a CyPass in somebody and wanted to see if it helped, can I measure that?

Dr. Moroi: We don’t have methods to measure uveoscleral outflow. It’s all calculated from the Goldmann equation after we measure all the other aqueous humor dynamic variables.

Back to distal outflow, even with the excitement of OCT angiography (OCTA), measuring aqueous outflow is still a challenge. According to Murray’s colleague Ricky Wang, PhD, at University of Washington, OCTA can image aqueous veins that contain blood cells, but current OCT instruments cannot image the deeper systems that only have aqueous humor yet. That will be a challenge.

Also, I was very intrigued to examine the distribution in aqueous flow, as was previously studied by Richard Brubaker, MD. It’s quite striking that there is a normal distribution in aqueous flow in the morning and evening, and now that circadian rhythms have gotten the interest because of the Nobel prize, hopefully, there will be more research into this. But I asked Dr. Brubaker the simple question, are those “highflow” individuals during the daytime the same “relative high-flow” individuals in evening with a circadian rhythm change? And he said, “I didn’t look at that.”

So, I just essentially replicated his study and I answered the question with yes, there is indeed concordance of aqueous flow that individuals who are high-flow in the morning are high-flow people in the evening. So, we’re taking the next steps yet to still look at that and see if there’s a genetic determinant to that because there is no question there’s concordance.

Dr. Grover: How did you measure that? How did you identify the patients that were high-flow or low-flow?

Dr. Moroi: We measured aqueous flow by standard fluorophotometry.

Dr. Johnstone: Why is it that basically flow can reduce by about 50% at night? Why does that occur?

Dr. Moroi: Still don’t know, do we? Yeah, we have no drug yet that’s as good as that endogenous biorhythm.

Dr. Johnstone: Okay. If the flow drops to half at night, but the pressure rises at night, what’s happening? Because it suggests that there is some profound change in outflow at night that must occur as well if the pressure actually rises in the face of a reduction in flow.

Dr. Moroi: Yes, so we need to do tonography in the evening to figure that out because I think most of us have assumed that outflow facility is relatively constant. And then, the same thing, of course, with uveoscleral outflow but that’s all calculated.

And then what happens at nighttime with the aqueous veins? Perhaps in a supine position with habitual IOP, outflow is reduced in a certain population so that the fluid can’t get out as well, and those patients are the nighttime so-called peakers. So, I think this issue of ambulatory IOP measurements will be interesting to look at.

Dr. Huang: There’s an old paper, I don’t remember who did it or the numbers, but the message was that adrenalectomized animals or people had less IOP change at night, implicating catecholamines and steroids.

Dr. Fellman: Do you know what’s interesting to me? It’s really hard to successfully lower somebody’s pressure without a tube, or a trabeculectomy, to those really low numbers that some of the older people here may know about because all we had was trabeculectomies. And you read about the work that some of the classic physiologists have done where they talk about the outflow system [and how it] is rigged for hypertension because we all know what hypotony does to your vision. The outflow system is rigged for hypertension and when you start looking at Murray’s work especially with all that histology and the electron micrographs and the fact that the deep outflow collector channels are parallel to Schlemm’s canal and they don’t really go out straight like all the books show, and then you realize that oh, wow, that’s so easy to just collapse. No wonder when I opened the canal, it didn’t work. And you look at that and you go, wow, and you say to yourself, wow, the deck is stacked against us for this kind of MIGS.

And then you say okay, but, somehow everybody’s pressure varies, like, four points throughout the day, so there’s this homeostatic mechanism in place that’s so good that it prevents people from getting glaucoma in a rigged system. And if you think about it that way and you see patients whose family members had glaucoma and you follow their children for 30 years, they look normal, Then suddenly they come in and their pressure has doubled. And you go, what has happened? Your gonioscopy exam didn’t change. Nothing changed except the IOP. I couldn’t see anything different on gonioscopy, yet the IOP doubled.

So, what in that homeostatic mechanism went wrong? Now, they talk about mechanosensing and the elasticity and the extracellular matrix, and then you read one study and it says well, we looked at the genes that express RNA related outflow and there are over 200 genes and I just closed the book and said this is overwhelming to me. But clearly, we need to better study this problem.

Dr. Grover: What do you think, Murray? Do you think it’s the change in pulsatile flow?

Dr. Johnstone: I’ve been sort of emphasizing the concept of pulsatile flow and that is I think one manifestation of the machinery that’s driving the system and tells you how the system functions in general. But it may well be a system where it can either act as a conduit, just fluid flows through it, or in a pulsatile fashion and that’s also true of the lymphatics. They basically have either a conduit or a pulsatile component depending on the pressure gradient surrounding them. So, even though I think pulsatility provides a lot of information, some of it may be that the conduit phenomenon is important too. There’s evidence that as we age the tissues stiffen and there is collagen everywhere, and the collagen in the meshwork is almost like that of tendons. The idea is that as we age we get this progressive stiffening of the tissues and that they no longer can undergo changes in configuration in response to pressure gradient changes so they can’t maintain that homeostatic set point in such a normal range. That’s one conceptual framework at least. But other alternatives, your thoughts on that?

Dr. Huang: The focus on the distal outflow pathways, where we see much of the pulsatility, may be very important here. Aiming for the angle and opening things up is really great but doesn’t necessarily get at the underlying disease and at what changed. I mean, what is causing the stiffening or scarring phenomenon? Remember, the P of POAG means that we don’t know the exact cause. However, the most reproduced finding in the aqueous has been elevated TGB-beta which can cause scarring at any level. In that sense, just opening the outflow pathways and the TM doesn’t actually get at the underlying cause. In fact, if we then start opening everything up, the question becomes whether while there is definitely a short-term IOP benefit for patients with early disease to change their overall disease trajectory, are we just opening things up more for all this bad humor and TGF-beta to mess up the distal areas as well? You may still ultimately need the bigger surgery.

Dr. Grover: Your comments about stiffening of tissue and structures over time made me think about another phenomenon; corneal hysteresis. Corneal hysteresis, as you all know, is essentially the shock absorbing ability of the eye. There is very good evidence that eyes that are very good shock-absorbers are less likely to progress and develop glaucoma, whereas eyes that are bad shock absorbers are much lower in corneal hysteresis and are more likely to develop disease.

Dr. Moroi: So, have you measured hysteresis and correlated the values in the context of surgical outcome?

Dr. Grover: We routinely use it in our practice.

Dr. Moroi: How about looking at that in the context of the outcome?

Dr. Grover: We are going to look at that. We wanted to look at some of our MIGS-based studies and then focus on two components. One is, does hysteresis change after MIGS, and two, does just general hysteresis predict a response to MIGS? The confounding factor I think we’re going to have a hard time with is changes in corneal hysteresis when a patient’s eye pressure changes. When a patient comes in with a pressure of 25 mm Hg, the reliability of their hysteresis may not be the best and they’re going to have a much lower hysteresis. Then, when you lower the eye pressure, the corneal hysteresis improves. The question is: would the hysteresis change whether you altered the eye pressure with medications or with a paracentesis. What alters hysteresis itself? Is it the pressure-lowering that changes the hysteresis? Is it the medication? Is it a specific medication? Or is it the surgery? I don’t know a way to tease that out.

Ron is one of the greatest thinkers in glaucoma. When I first started practicing with him, he brought me in and he showed me a couple patients. One had a trabeculectomy. He checked eye pressure with a pneumotonometer with the patient sitting up and laying back. He did the same maneuver with a patient after a tube shunt. These patients behaved differently! The trabeculectomy patient’s eye pressure did not raise as much when laid back as compared to the tube patient.

I think that trabeculectomies have a greater capacity to handle that variation than tubes. It’s interesting to see whether we can tease this observation out with corneal hysteresis. Does corneal hysteresis change differently when you lower the pressure with a trabeculectomy versus a tube?

Dr. Moroi: I’d be interested to see that data.

Dr. Fellman: I agree.

Dr. Moroi: Because that’s the only clinical tool we have to give some global measurement. I’m not sure what it’s measuring, but --

Dr. Grover: It’s measuring something.

Dr. Fellman: When you’ve seen a patient for a long time who does develop disease and you keep looking gonioscopically and you don’t see anything you wonder what really is happening. When you started talking, Murray, it made me think of a new term that we need. Imagine that.

We need better alphabet soup in glaucoma. For example, we are all familiar with PAS, but what term do we use when the anterior wall of the canal is stuck to the back wall with adhesions? That situation might alter your MIGS, but we can’t see it yet, but when we use OCT, we need to have nomenclature to advance the outflow field.

Dr. Johnstone: Well, I think there’s a lot of clinical evidence based on the work by Schirmer, Suson and others. They showed that when you reflux blood in the Schlemm’s canal in normals, the blood just moves in and out of the canal very rapidly. Blood entry and exit progressively slows in ocular hypertensives. When you get into significant glaucoma, you get this patchy filling of the canal and in advanced glaucoma no filling. It suggests this phenomenon of blood movement slowing is probably an earlier issue based on clinical observations.

Dr. Fellman: We just don’t have any nomenclature for the specifics of what is happening in the canal. There’s PAS and posterior synechiae, then there’s canalicular synechiae or whatever the word will be to help people understand the disease better, at some point.

Dr. Moroi: I talk to the residents and fellows and say, “Notice the difference in flow pattern,” and make the comparison to fluorescein angiography that they learned on retina rotation about how the aqueous veins have tight junctions, the outflow is quick and it goes into probably the inferior or superior vortex veins and outflow. But these deep intrascleral vessels, I haven’t found where they go to and they leak. Are these terminal vessels within sclera conducting aqueous?

Dr. Johnstone: This is, you’re looking primarily at fluorescein?

Dr. Moroi: Yes

Dr. Huang: Yeah, leakage is a problem with fluorescein. It’s just like IVFA for the retina. After the late phase, there is leakage of some fluorescein from the retinal vessels. It’s for that reason that I also use ICG for outflow angiography

Recall that retinal folk use ICG because it is protein-bound and maintains a better intraluminal retention. This helps imaging of choroid vessels that is very hard with fluorescein. For aqueous outflow, with ICG, the image can thus appear tighter or sharper, and I think that difference is just a representation of the molecular characteristics of fluorescein.

Now, I want to mention here that we are talking about living and intact eyes. In the laboratory, when we are using enucleated eyes, all tracers eventually leak all over. Because in post-mortem eyes, the continuity of the outflow pathways to the rest of the body is severed so that all tracers eventually accumulate on the surface and actually everywhere. In this case, if one waits too long, initially segmental patterns become homogeneous and are not segmental anymore.

Dr. Johnstone: I think the devil here is in the diffusibility of the fluorescein. It diffuses through so quickly and so easily that it’s really hard to use it as a tool.

Dr. Moroi: My anecdotal evidence is that when I do not see that, the patients have poor outcomes.

Dr. Huang: That’s cool and new. Because if you don’t see the leakage, in particular in cases with poor outcomes, it might tell you something about the pathways, their integrity, and hyperscarring, so the outflow pathways are too tight in a way reminiscent of TGF-beta pathway activity.

Dr. Moroi: I definitely notice a difference, and I’m sure you both do also if you use the Eye Science catheter, but the Schlemm’s canal is very different in the anatomy and access and diversion to these collector channels. I’ve had some patients whose canals are so stenotic that I’m probably doing an inadvertent trabeculotomy on some.

Dr. Grover: When we do a GATT surgery (either with a catheter or a suture), some patients have a very distensible trabecular meshwork where it is easy to cannulate Schlemm’s canal. In other patients, the canal is very tight and “sclerotic.” It’s almost like you have these herniations (or whatever it is); it seems like the TM is stuck to the back wall of Schlemm’s canal and you’re prying them open. But we’re seeing all these interesting intraoperative observations. I’ve had patients who are in their 60s who have a tight sclerotic canal and really stiff TM — these are the exceptions but the ones you remember. I’ve also had a patient in her mid-to-late 80s who had the opposite; I called it a young eye because she just had the most distensible trabecular meshwork ever.

Dr. Fellman: Yes, it would be interesting to correlate that with hysteresis. But even with Trabectome, you go in and sometimes it just glides. It’s like going through butter just like you said. And in other cases, it’s like going through concrete. You literally have to come out and you think well, maybe I was lodged into the back wall. And you go, no, I’ve done several hundred of these. I don’t think I’m doing that. And you reposition and you still get the same thing, so it’s like cardboard in there. What’s going on there? Why is the canal and TM so stiff? There’s a definite difference to the texture of the tissue that you learn to appreciate the more surgery you perform. (See sidebar, on en face OCT)

Dr. Grover: I want to get back to something that Sy was talking about with the behaviors intraoperatively with your fluorescein canalograms. One thing we’ve seen after a GATT, is we’ll artificially elevate the pressure and we’ll see just a slight wave, and then we’ll let the pressure come down and I’ll artificially elevate it again. And then something changes during that whole process where boom, it just goes. You see a 360 degree wave. And the thinking that when we (Ron and I) have had, is that we see this maneuver as a lavage of the collector system, whether you’re popping open segments that have never seen aqueous in a long time or unclogging the collector channels. Or one time, I was operating on a lady with a dysgenic angle and Ron was in the room. We saw the AC deepen dramatically and a diffuse wave right afterwards. It made us think of the effect of miotics on the scleral spur and whether we were mimicking this in the operating room. What do you think is happening?

Dr. Johnstone: Well, I could certainly envision the possibility that there’s appositional canal closure just like with the angle. It’s probably appositional phenomena going on for a long period of time before synechial problems occur and you may well be just popping the system open and getting it to function. And I have a question: Do you drop the pressure below episcleral venous pressure to allow blood to come back into the canal? Because you may be actually disrupting adhesions in the process or at least areas of apposition and allowing the thing to move again.

Dr. Grover: I purposely lower the eye pressure. But then the other thing we do is try to intraoperatively evaluate the outflow capacity. There are some patients where we’ll pump up the eye pressure and we won’t see a wave at all. There are other patients where we pump up the eye pressure and we’ll see a diffuse wave that will stay blanched. And then there are other patients where we pump it up, we see the wave, and then we see the blood reflux back into the episcleral vasculature immediately. Ron has a great video of it coming back in a pulsatile way. We’re wondering whether this observation can give you insight into outflow capacity.

Dr. Johnstone: Let me just pose this. If you see these phenomena occurring, and particularly if you don’t see them — in other words, if the response doesn’t seem to be happening properly at all, should you be perhaps rethinking what you decide to do with that patient? That may increase your success rate of, in terms of your procedures if you identify the folks who have a distal outflow system that’s working and treat them vs ones that don’t, as a functioning distal system, and modify your procedure to deal with that. And I’m not sure what the algorithm should be to do that, but I’d like to hear your thoughts.

Dr. Grover: That’s what we published on and Ron can speak to that. But you have touched upon the dilemma. Only when we get smarter about the behavior of the outflow capacity in the operating room, can we make intraoperative decisions. The study that Ron was the lead author on and that we published in Ophthalmology showed that only one-third of patients who did not have a wave in the operating room needed further glaucoma surgery. And that’s what we’ve been stuck on because you can flip that around and say two-thirds of patients with no wave in the operating room didn’t need subsequent surgery. Have you ever done this in a patient who was relatively high risk that you knew at that moment that was the last time you were going to take them in the operating room?

Dr. Fellman: We just had a case that we did together of a gentleman who had a tube and a prior failed trabeculectomy and a tube; we had to take his lens out. So, we thought okay, we really think his distal system is probably pretty bad. Let’s go with a CyPass, let’s say, or we can tap into uveal scleral outflow if there’s no wave at all. That study hasn’t been done in a prospective way, but that’s kind of what you’re thinking, and that makes a lot of sense.

And then the question is, these 21-mm eyes, their actual absorption area for uveal scleral outflow is, I think, probably a lot less than these bigger eyes. So, I don’t understand how that fits into it either yet.

But one last thing, and then we’ll let Sy talk, is that work, when Davinder said, “Wow, did you see that chamber deepen like that?” Then all of a sudden, we saw that wave. I said, “I have this article you’ve got to read by Mike Van Buskirk,” who would depress the lens and it would significantly increase outflow. And he surmised that he probably would pull the scleral spur down, opening up the whole outflow system. I said, “Wow, that’s I think what we’re doing.” I sent Davinder this article and he wrote me back, “This is awesome.” It’s good to know the old literature and pass it on.

Dr. Moroi: Well, I was just going to say trying to integrate. Everybody is contributing great work. But integrating with the biomechanics, the thickness of the sclera and just getting somebody to get into scleral biology because that’s still such a stepchild area. Dr. Jody Summers has done a lot with basic chemistry looking at sclera, the GAGs (glycosaminoglycans) and the collagen documenting how it decreases with aging. But I do believe with the scleral biology, all of us feel how the sclera differs from patient to patient, especially in young sclera, dealing with congenital glaucoma, the scleral is very thin and mushy with usually great outcomes with GATT procedures or trabeculotomy. And then you have the TM that’s like cardboard. I’ve tried to finger touch to see what it feels like in some sclera, but if we had a sensitive shock absorber-mometer to let us know the tissue properties. Another topic to integrate into IOP regulation is vascular biology and how we can reintroduce aqueous veins and rejuvenate the sclera to “youthful” conditions for IOP control.

Dr. Johnstone: One thing that I think we maybe could pay more attention to is that in cardiovascular physiology, there’s a great deal of evidence to show that if you don’t have flow through an area, basically eliminating any shear stress and wall stress, that you basically have involutional changes in the vasculature. And many years ago, Dvorak Theobald looked at glaucoma eyes and found that in the distal outflow pathways there were major involutional changes that were occurring. She showed very clear herniations of trabecular tissue in the collector channel entrances and then the distal channels beyond them had undergone involution. So, it may, that’s another reason, of course, for thinking about getting at the problem earlier. But also, with your techniques, trying to figure out which of the folks have gotten into this involutional phenomenon where you’re not going to get much benefit from your outflow procedures.

Well, this has been great. I think we can just have some wine and sit here for another five hours talking.

I am so excited about this session, bringing these leaders and thinkers together for interaction and synergy, and hopefully, future collaboration. Thank you. OM

Roundtable Contributors