Lecture: STOP Diabetic Retinopathy

This lecture will review the spectrum of clinical findings in diabetic retinopathy, its classification and appropriate treatment strategies. The skills necessary for effective intervention in this common blinding disease will be highlighted.

At the conclusion of this presentation, the student will:
• Be familiar with the ophthalmoscopic appearance of diabetic retinopathy.
• Recognize the critical fundus lesions which require a treatment decision.
• Be familiar with a variety of available treatment strategies.

Lecturer: Dr. Rosalind Stevens


(To translate please select your language to the right of this page)

DR. ROSALIND STEVENS: Good evening. It’s wonderful to have you all here. We’re quite excited to have such a large number of participants across East Africa. My name is Rosalind Stevens, and I’m an ophthalmologist, a specialist in medical retina and vitreoretinal surgery. I did my training in San Francisco, then Johns Hopkins for medical retina, Emory University in Atlanta for surgical retina, after which my family and I went to live in Saudi Arabia for seven years, where we had a lot of exposure to the same level of diabetic retinopathy that I believe you do in your practices. So I can appreciate the challenges that you face, and I hope that you’ll be able to gain some further insight in the management of these patients during our discussion this evening. If you have any questions, don’t hesitate to chat and put them online, and we can answer them either during the course of our discussion or at the conclusion of the presentation. Again, thanks so much for coming. I’m going to be speaking this evening about diabetic retinopathy, and specifically how to stop its progression to blindness through the use of photocoagulation and some other treatment methods. I wanted to ask the group a few baseline questions, so that I would have a better idea of who all of you are. I do have a list of your names and your organizations, which is very helpful. And following the presentation, we’ll essentially ask some of the same questions, to get a sense of whether you’ve been able to absorb some of this information. So first I wanted to ask: How many of you take care of diabetic patients? And it looks as though the vast majority do. 80% do. And I think that the majority of you are involved in eyecare, so I presume that that relates not only to patients with diabetes mellitus, but those with eye disease related to that process as well. Our learning objectives for this evening’s presentation include being able to recognize diabetic retinopathy, understanding at least the most accepted classification, basic classification of diabetic retinopathy levels of disease, to be able to identify patients who need treatment, which is critical, and to increase your awareness of treatment methods, as well as some of the skills involved in performing those treatments. As background, I wanted to remind you that diabetes mellitus is becoming an epidemic across Africa, as it is globally. It’s estimated that by the year 2030, there will be a global increase of 51% of diabetes mellitus worldwide, and the largest growth will be in Africa, with an 81% increase, relative to baseline. This represents 9% of the world’s population, so this is in fact a public health epidemic. In 2013, some studies done in Sub-Saharan Africa found that roughly 15 million people had diabetes, and of those, greater than 90% were type II. Some further research in both rural Uganda and urban Kenya highlighted the difference in prevalence rates, at least as they’re defined by our screening methods. The rate of diabetic retinopathy disease and diabetes mellitus in Kenya was found to be somewhere upwards of 12%, and the rate of diabetes mellitus in rural Uganda, with or without retinopathy, 1%. So that’s probably much lower than the actual level of disease, but that’s the reported level, based on screening. The five-year mortality due to diabetes mellitus ranges from 4% to 57% in East African countries. So I wanted to ask you if systemic diabetes has been present for more than 20 years, what percentage of type II and type I diabetics will have some level of diabetic retinopathy across all types of diabetes? So the key to this question is: Based on a 20-year history of diabetes, what percentage will have retinopathy? And it’s okay to go ahead and vote. And I think the answers ranged from 40% saying 30% will, 20% of you saying 50%, 20% saying 75%, and over 90% by 20% of you. So it’s not important to think about that now. During the course of the discussion, though, I think this will become more evident. The next question: After the diagnosis of diabetes has been made, how often should a 35-year-old patient’s eyes be examined? So they come into the endocrine clinic. They’re discovered to be a diabetic. How often should their eyes be examined? And the answers range. Every 6 months, every 12 months, or 2 years is our spread. With a majority of you saying 12 months. No one said every 5 years, so that’s good. I can tell you right away that’s not the correct answer. The 4:2:1 rule for severity of non-proliferative diabetic retinopathy includes which of the following? A, macular edema, B, vitreous hemorrhage, C, retinal hemorrhages greater than 20 in number, in all four quadrants, and retinal hemorrhages greater than 40 in number in all four quadrants of the retina. So you’re voting, voting. That’s good. And the answer is: Among the group watching, 75% said retinal hemorrhages greater than 20 in number, in four quadrants. And we’ll learn the answer to that during the course of the presentation. Pan-retinal photocoagulation — this is laser treatment — is defined as the following. And by this I mean as defined in the early treatment diabetic retinopathy study and the diabetic retinopathy study, the classical definition of pan-retinal photocoagulation, using a standard argon laser. And the choices are 600 spots, 900 to 1100 laser spots, 1200 to 1600 spots, and 2000 to 2500 spots. And 60% of you said 1200 to 1600 spots, with a couple of other votes lower and higher. So, again, we’ll emphasize this during the presentation. Focal/grid laser treatment for clinically significant macular edema includes the treatment of: A, macular edema, B, neovascularization, C, microaneurysms, or D, A and C together. Voting, voting. And I think people are a bit unsure about that, so we were clear by the time we got to the end of the choices, A and C. 100%. So when we think about the epidemiology of systemic disease in diabetes, we know that advanced diabetic retinopathy is really a metric or a correlate for the level of systemic disease. We think of the eye as the window to the soul, and it’s also a window to one’s systemic health in diabetes. If you see advanced diabetic retinopathy, as evident on the slide on your right, with a lot of vascularization, scar tissue, areas of focal detachment of the retina, and associated subretinal hemorrhaging, this type of patient has a much higher risk of heart attack, stroke, kidney disease, amputation, and death. In fact, the risk of myocardial infarction in a patient with severe proliferative disease is approximately 50% in three years. We know that diabetic retinopathy is prevalent, and that in fact it’s the most important cause of adult blindness in the United States, as well as globally, in working age adults. And that means people aged 20 to 64. So it has a huge economic cost. If the duration of the disease is greater than 20 years, 99% of insulin-dependent diabetics and 60% of non-insulin-dependent diabetics will show some level of retinopathy. The rate of blindness in the United States in those who have diabetes onset under age 30 is 86% due to diabetic retinopathy. So that’s all causes of blindness, with the vast majority being due to diabetic retinopathy. If the onset of diabetes mellitus is over than 30 years, and when looked at all causes of blindness, still a third are due to diabetic retinopathy. So why should we discuss diabetic retinopathy? As you know, it’s quite a challenge to diagnose it. There’s a tendency to delay treatment. Patients go from one clinic to the next, without a decision being made to initiate intervention. Often the treatment is inadequate. I know that in our practice in the United States, many, many patients come to the retina clinic that are undertreated. They have some level of laser treatment, perhaps, but many are not treated to the point that involution of their diabetic retinopathy occurs, and the disease is progressive. It goes through a natural life cycle from mild to moderate to severe, over a period of a few years. And it’s our job to make the diagnosis, to grade the level of disease, and to intervene with our very effective treatments we have available, prior to permanent structural damage or progression to blindness. First, the most important thing that we need to do is be able to examine the posterior segment of the eye. One critical area that we need to be expert at is examining the macula, because of the prevalence of macular edema as a cause of visual loss in diabetes, and for this, we need to be good at using our slit lamp, with a magnification lens that allows us to have good stereo vision. So we like to use a 90 diopter or a 78 diopter lens, and you see these pictured here on your left. These are lenses that you hold in front of the patient’s head, with a slit lamp in place, and you learn over a period of a few days to focus the lens 3 to 4 centimeters in front of the patient’s eye, and learn by tilting it to obtain a good, sharp image of the macula, and be able to perceive in stereo, three-dimensional context, whether there is thickening of the retinal tissue present or not. If there’s poor media in the eye, you often have to tilt the lens slightly, to allow the slit beam that you’re using to go past the opacity and focus on the retina, and a lot of these patients may have bits of vitreous hemorrhage floating around that you need to learn how to focus through. When you’re doing that, it helps to make the slit beam as small as possible. Make it the size of the pupil and as thin as possible, and this will be much like using a small pupil exam with your indirect ophthalmoscope, in order to see the posterior pole most effectively. In order to see the retinal periphery, one likes to use a 3 or 4 mirror lens, and you see those pictured on your left. The far peripheral mirrors, the tall mirrors — the taller the mirror shows the more peripheral part of the retina — and these are important lenses to learn how to use, because you use them to treat retinal tears and breaks in the retinal periphery, in addition to diabetes. The downside of these lenses is you have to twirl them for 360 degrees to move the mirror around the eye, to enable you to see the entire fundus. The other type of widely used lens is a wide angle lens, seen on your lower right. And the fisheye lens configuration generally means that you’re seeing a much broader expanse of retina, but the view is small. It’s very minified. This is useful for doing a faster, more efficient laser treatment, because larger areas of the retina can be in focus at once. This lens needs to be held very still by the user, and tilted slightly toward yourself, in order to get a good image of the retina. Lastly, indirect ophthalmoscopy is a critical element in the treatment of diabetes, because this allows us to image most of the posterior pole and the entire retinal periphery. Those who do a lot of vitrectomy surgery tend to use 28-diopter lenses, which gives a smaller image, but gives a view further out, toward the peripheral retina. Also, this 28 lens is the size that we use for doing retinopathy of prematurity screening in babies, because the view with the 28-diopter lens equals so-called zone 1 or the central optic nerve and macula in an infant. So these are excellent images to learn how to use. You will become facile with holding this lens about 5 centimeters in front of the patient’s head, and over a period of several weeks learn to tilt the lens opposite the user. So that you’re, in a sense, using the image that you see in the lens a bit like a scleral depressor, viewing a portion of the retina at a time, and then working your way 360 degrees around the eye to do so. You don’t have stereo glasses to see this, but I wanted to show you what elevated neovascularization elsewhere looks like, when it’s in focus. This fibrous tissue that you see on your left and the fine diaphanous red lines of blood vessel growth are in the plane of focus of your indirect ophthalmoscope, whereas the retina behind this elevated tissue is out of focus. And this shows you an example of the relative levels of tissue pathology, elevation, scar tissue, and traction on the retina that you’ll be exposed to, when caring for diabetics. So how will we diagnose diabetic retinopathy? The screening recommendations vary, but in general, the American Academy of Ophthalmology and the International Council of Ophthalmology agree that you should have an examination five years after the onset of diabetes, at least. Now, you have to know when the onset is. It’s frequent for patients to feel that they’ve just been diagnosed, but in fact they’ve been diabetic for 10 or 15 years. So that is dependent on what your first exam shows. If they’re normal, those patients will come back in a year. If the patient’s 30 years old or older, they’ll have an examination yearly. If they’re pregnant, they’ll have an exam at the time the pregnancy is discovered, and then certainly every 3 months during the course of the pregnancy, or sooner if retinopathy is discovered. So what would we look for, when we’re looking at diabetic retinopathy? The spectrum of diabetic retinopathy goes through a life cycle. You can see in your upper left a posterior pole that has scattered focal hemorrhages, tiny red hemorrhages, which are within the tissue of the retina, and many yellow-white appearing areas of lipid or cholesterol deposition throughout the middle layers of the retina. These are areas where the normal retinal vasculature has lost its tight junction configuration, and because it’s no longer watertight, the large lipid or cholesterol molecules can leach out through the retinal vasculature into the surrounding retinal tissue, and become visible. So when we see this pattern on the left, which is one of typical non-proliferative diabetic retinopathy, that patient is on our screening schedule right away. Now, without any treatment, we can anticipate that about 20% of patients will progress from non-proliferative to proliferative disease, which means the growth of new blood vessels on the surface of the optic nerve and along the vascular arcades. In the middle photograph, you see that most of the diabetic disease is in the posterior pole, whereas the equator, moving anteriorly, is not so much involved. There have certainly been studies showing that there is significant vascular non-perfusion in the retinal periphery, but it’s quite unusual to see far peripheral anterior neovascularization. There have been studies in Japan showing about 5% of the Japanese population presents with far anterior peripheral neovascularization, but in your patient population, I would be concerned more about the presence of sickle cell disease, if peripheral neovascularization is seen. Now, without treatment, the patient with proliferative disease may progress. It will undergo a process of involution or drying out of the neovascularization. A lot of the neovascularization becomes fibrous and forms scar tissue. This scar tissue can contract and pull the retina away from the back wall of the eye, and as you see in your lower right, this white fibrous tissue with the interstices, the middle part of the retina, having many red blood vessels within it — this patient has a partial traction macular detachment, with severe vision loss, due to the natural involution of untreated diabetic retinopathy. So we know that retinovascular disease is associated with hyperglycemia in diabetic retinopathy. What happens initially is the hyperglycemia causes the small pericytes, which are cells on the outside of the blood vessels themselves, to flake away or fall off, much like dandruff. And this leads to a loss of integrity of the retinal vasculature, loss of barrier function, and hence leakage into the retina, in the form of either clear serum, as in macular edema, or a lipid itself, as I noted. When we classify diabetic retinopathy, there are two general categories of retinopathy. The earlier one, so-called non-proliferative diabetic retinopathy, without proliferation, and the later stage, so-called proliferative diabetic retinopathy, or with proliferation of new blood vessels. So an example of non-proliferative disease is seen on your lower left, where we see again lipid leaking into the interstices of the retinal tissue, but no evidence of proliferation of new blood vessels. In the middle picture, you see evidence of new blood vessel growth on the optic nerve and inferior to the optic nerve, associated with focal areas of tractional hemorrhage. This is typical proliferative disease. In between the two, there’s a category called severe non-proliferative diabetic retinopathy, and this implies that the eye demonstrates intraretinal hemorrhages greater than 20 in number, in all four quadrants. Hence the 4 numeral. 2 represents the presence of venous beading. And you you can see in the background photograph here some irregularity in the size of the vein. There are focal areas of widening of the vein and also focal areas of thinning of the vein, much like jewelry or beading. So this is another sign of progression to severe non-proliferative disease. And finally, a 1 represents IRMA, intraretinal microvascular abnormalities, which are areas of reorganization or change in configuration of the retinal capillary bed, but not yet at the stage of neovascularization. And the little orange arrow points to an area of IRMA. That’s the squiggly S-shaped blood vessel that you see. And this is another sign of progression to severe disease. Lastly, clinically significant macular edema means retinal thickening observable at the slit lamp. And we’ll go over that in more detail. So non-proliferative disease shows visible microaneurysms, the little tiny red dots seen throughout this posterior pole, and broader, larger areas of intraretinal hemorrhage. The hard exudates is another name for lipid deposition. The hemorrhages in the retina can be of many shades. The so-called dot hemorrhages tend to be within the retinal tissue. The blot hemorrhages are more anterior, superficial in the retina. The flame hemorrhages follow the nerve fiber layer and hence look like flames. Macular edema or retinal thickening seen at the slit lamp can be characterized as focal, small areas of edema where the rest of the retina is fairly normal. Or diffuse, where the entire posterior pole all looks edematous. Or there may in fact be cystoid medium, CME, and this is represented by a petaloid pattern of leakage within the foveal avascular zone, and we’ll see that momentarily. So during the course of progression of non-proliferative diabetic retinopathy, the capillaries undergo shunt formation, and they change or remodel. Some areas lose their capillary perfusion function, and you can see in the photo on your right, in the lower right hand corner, the areas of grey loss of any kind of vascular tissue. So within the green circles I’m drawing here, you can see typical examples of capillary non-perfusion. And this is important to recognize, because we know that these are the primary areas that promote the generation of what’s called vascular endothelial growth factor, or VEGF. And this is the chemical fertilizer, so to speak, that induces the growth of new blood vessels when an eye becomes ischemic enough. This capillary dropout may also occur in the macula, and at this point in time we don’t have a standard treatment for this loss of perfusion, should it occur in the central macula. It typically leads to severe vision loss which is untreatable. We may also see cotton-wool spots, and these cotton-wool spots look like the ends of Q-Tips. And you can see on your lower left screen the tiny fuzzy white areas that look just like a Q-Tip. These are areas of non-perfusion, which has caused loss of energy or blood flow in the area underneath the nerve fiber layer, which then leads to a stasis or slowing of axoplasmic flow, or flow of chemicals up and down the nerve fiber layer. When one sees cotton-wool spots like this, take this as a sign of increasing activity of the retinopathy. Many years ago, this used to be termed preproliferative diabetic retinopathy. So if one sees a patient with cotton-wool spots, there is a higher suspicion that the retinopathy may be progressive. Non-proliferative diabetic retinopathy may lead to vision loss, even though there’s no proliferative disease. And the vision loss is caused by swelling or thickening of the macula. Macular edema. Macular exudate, lipid, in the fovea — in fact, if the lipid that you see here encroaches on the central fovea, it will often form a chronic star pattern, and we call this not only cystoid macular edema, but a star pattern of lipid maculopathy. And this has an extremely poor prognosis. There may be capillary non-perfusion in the foveal area, and focal ischemia of the central macula as well, and all lead to vision loss. Here’s an example of a focal area of macular edema. You can see an area of lipid formation, and there are some central areas of microaneurysm formation. So you assume that the lipid is coming primarily from the central microaneurysms. And this is an example of a person that could be treated with focal laser treatment directly to the microaneurysms within the lipid ring, which is sometimes called a circinate ring. Sorry, just having a little trouble. I think my main issue is going to the next slide. Thank you. This is another example of the fellow eye of the same patient, with a focal area of circinate lipid maculopathy. You can see the early formation of a star lipid within the fovea. So this is a high risk eye for visual loss. The source of the microaneurysmal leakage is within the circinate ring, superiorly, and again, temporal to the inferior temporal arcade, and nasal to the lipid ring. And those would be areas for potential focal laser treatment. Late angiogram of a similar patient shows accumulation of dye within the interstices of the retina, typical for this pebbly pattern of cystoid macular edema. And this is a patient that we would, as a first choice, treat with anti-VEGF therapy. And we’ll go into that in a bit. Here’s another example of a patient with multiple areas of microaneurysm formation. The angiogram shows a few small areas of non-perfusion, where the retina looks diffusely grey, and there’s some loss of integrity of the margin of the foveal avascular zone as well. This is an example of a 51-year-old male with excellent acuity in his left eye who’s been insulin dependent for 4 months, he says. Now, we know he’s been diabetic for at least 8 to 10 years, because of the level of his diabetic retinopathy. One might describe this as multifocal macular microaneurysms with scattered dot hemorrhages. There are some scattered areas of lipid maculopathy. Only one tiny bit is within the central 1 millimeter or so of the macula. The rest of the lipid is exterior. And there are some light cotton-wool spots seen along the superior temporal arcade. Wide angle view of the same retina shows a similar pattern in the nasal fundus, but this disease is limited primarily to the posterior pole. On fluorescein angiography, the integrity of the foveal avascular zone is limited. You can see that instead of being perfectly round, there’s sort of a light grey pattern. And that tells us that there is some non-perfusion of the fovea. The temporal fields and superior temporal fields of this image show that the retina has a granular, bald-looking appearance. So it’s almost completely non-perfused. And this patient is forming a lot of vascular endothelial growth factor and is at very high risk of progression to proliferative disease. Wide angle photos of the same image show that this non-perfusion exists essentially 360 degrees around the posterior pole. Severe non-proliferative diabetic retinopathy, you will recall, meets the 4:2:1 definition. When we see progression to more extensive intraretinal hemorrhages, venous beading and looping in two quadrants, and IRMA, intraretinal microvascular abnormalities, this is a higher risk patient. Here’s an example of a patient who’s starting to develop more than 20 microaneurysms — 20 focal areas of hemorrhage, more than 20 focal areas of hemorrhage. And this particular patient almost has the appearance of an early vein occlusion or venous stasis, because of the striated pattern of the hemorrhage in their temporal optic nerve field. In this example, one can see another pattern of scattered lipid in the inferior temporal macula. And many tiny intraretinal hemorrhages, superior temporally and superior nasally. When we do an angiogram, however, on this patient, although this was not visible clinically, one sees an area of bright white hyperfluorescence in the upper fundus, and we call this lightbulb sign. And this is a critical identifier for progression to early proliferative disease. If you see this angiogram on a monitor screen from across the room, you would be able to identify this lightbulb sign. And at this stage, you’d know the patient is proliferative and may require, at the very least, laser treatment. We know that if the patient has severe non-proliferative disease, and we call severe one of the 4:2:1 criteria — if they have one of those, they have a 15% chance of becoming proliferative in the next year. If they have two of these, two of the criteria, for example, if they had four quadrants of intraretinal hemorrhage and two quadrants of venous beading, they have a 45% — almost 50% — chance of becoming proliferative in one year. And you can see the patient here has profound pronounced venous beading in these two veins, and there is some early IRMA here, which, if you did an angiogram, would likely show progression to proliferative disease. Proliferative diabetic retinopathy is classified by several types of neovascularization. When we talk about neovascularization, we divide it into essentially two types. Neovascularization occurring over the optic nerve, on the optic disc, is called NVD. Neovascularization of the disc. NVE is neovascularization elsewhere. And that implies the vascularization is greater than one disc’s diameter from the optic nerve. So in the lower photo, you see some black and white imagery showing flat neovascularization elsewhere. The size of this neovascularization is important in clinical trials, and if this neovascularization occupies more than roughly one third of the surface of the optic nerve, we call that significant, and if it’s more than one disc diameter in size in the peripheral retina, that’s significant. But the key thing is to recognize and identify the neovascularization. When the NVD and/or NVE is present, proliferative disease will often manifest hemorrhage from this blood vessel growth. This may be preretinal, and that means on the surface of the retina, but behind the vitreous. And this typically takes a boat-shaped or flat pattern, because the vitreous holds the hemorrhage in place, or sequesters it. Or there may be vitreous hemorrhage. If the bleeding occurs out into the body of the vitreous itself. When the proliferation begins to regress or stop becoming so active, which happens spontaneously, this may pull the retina off or cause tractional retinal detachment, another component of proliferative disease. And lastly, if there is a lot of vascular endothelial growth factor, this may cause growth of new blood vessels not only on the retina, but also on the surface of the retina, the iris itself, within the iris, on the surface of the iris, and eventually in the trabecular meshwork or angle. If the angle becomes occluded, then neovascular glaucoma is a possibility. High risk characteristics, which we just reviewed, increase the two-year risk of severe visual loss without treatment by 25% to 35%. So in the patient you see here, who has subvitreous, subhyaloid hemorrhage, which is sequestered, versus vitreous hemorrhage on the right, both of these are characteristics which increase the chance of severe visual loss. In the patient on your left, vitrectomy would be required to physically remove this hemorrhage. The patient’s vision is in the range of counting fingers, and this type of hemorrhage takes a long, long time to resolve spontaneously. However, the hemorrhage on your right — we would typically wait a month or so before doing vitrectomy, because at least 70% of the time, these so-called minor vitreous hemorrhages will clear spontaneously, to enable laser therapy. This is an example of a patient who appears to have only mild scattered microaneurysms, although there are some suspicious areas of new blood vessel growth on the surface of the optic nerve. Fluorescein angiography highlights this hyperfluorescence, which occurs in the transit phase, or approximately one minute into the fluorescein angiogram study, and this is pathognomonic, absolutely typical, for neovascularization of the disc, thus moving this patient into the category of proliferative diabetic retinopathy. Here’s the fellow eye of the same patient, and we can see clinically. We don’t need the fluorescein angiogram to see these fine, thin, hair-like blood vessels growing on the surface of the optic nerve, and essentially within one disc’s diameter of the optic nerve. The fluorescein angiogram shows acute hyperfluorescence early if the transit phase of the angiogram, and this, again, highlights the lightbulb sign or the leakage of fluorescein through the new blood vessel tissue. The neovascular blood vessels are not competent. They’re not watertight. They’ve lost their barrier function. And therefore the fluorescein easily leaches its way through the blood vessels and shows us the presence and the location of the new blood vessel tissue. Later in the same study, we can see the fluorescein dye pouring out of the neovascularization of the disc, and one can also see the pattern of cystoid macular edema, or we know it to be clinically significant, because it’s cystoid in the presence of the neovascularization. Here’s another patient with neovascularization elsewhere in the inferior temporal quadrant. You can see fibrous tissue associated with subhyaloid hemorrhage. It’s evident that there’s a tremendous amount of capillary non-perfusion in the temporal retina, where the white blood vessels are totally closed and have become sclerotic, and you can see that the vessels in the superior temporal periphery are quite thin and attenuated, also implying ischemia. There’s a patch of neovascularization elsewhere, at the 11:00 position in the midperipheral retina, which looks like a fine lacy pattern. This is a more temporal view of the same fundus, highlighting the white sclerotic vessels, and this patient has a massive amount of vascular endothelial growth factor being produced. Fluorescein angiogram shows many, many tiny foci of hyperfluorescence, many little lightbulb signs, showing diffuse neovascularization elsewhere, as well as neovascularization of the disc. And here is the superior temporal periphery of the same patient. The dark grey granular-appearing areas are areas of massive capillary non-perfusion, in contradistinction to the fine areas of fairly well-perfused tissue in the inferior nasal portion of this image. We can see in this eye multiple intraretinal dot hemorrhages in the temporal periphery, and there’s an area that’s suspicious for neovascularization elsewhere at the 2:00 meridian in the retinal periphery. Fluorescein angiography highlights the hyperfluorescence of the same area we’ve just been looking at, and one can see string-like areas of hyperfluorescence on the optic nerve, representing active neovascularization of the disc. This is another patient with associated subhyaloid or subvitreous hemorrhage, associated with neovascularization elsewhere, along the inferotemporal arcade, neovascularization of the disc overlying the optic nerve. This is a red-free photograph, which highlights the new blood vessel growth. It’s a little bit more evident visually. And this is a late recirculation phase of the angiogram, showing multiple areas of relative hyperfluorescence, consistent with active neovascularization of the disc and elsewhere. Finally, the last stage of proliferative disease. The fibrous tissue forms when the blood vessels involute or dry up. They often leave behind scar tissue, which you can see the whitish tissue along the superior temporal arcade. It has a curved shape, and that implies tension or traction on the retina. And this is a severe end stage patient with the same involutional fibrous tissue, causing macular traction and detachment of the retina. This is another angiogram. If you look at the central macula, you see an early petaloid pattern, consistent with macular edema. And this is associated vitreous hemorrhage. The grey material that obscures the fundus detail is in the mid-vitreous, and therefore you’re unable to see the retinal detail. So this is one of the other sequelae of proliferative diabetic retinopathy. The late recirculation phase shows more pronounced CME, cystoid macular edema, and lastly, the image in the upper left and lower left show early fine neovascularization growing on the surface of the retina, and also the pinkish appearance, superior-temporally, in the image, showing vascular tissue growing in the trabecular meshwork. So how do we manage diabetic retinopathy? First of all, there’s a general follow-up timeline. If a person is screened and has no retinopathy, but you know they’re diabetic, they still need to be re-seen every year. If they have early, mild, non-proliferative disease, a few scattered microaneurysms, but no macular edema, they should come back yearly, or if you’re concerned that they have a few more risk factors, they could come in 6 months. But just microaneurysms with just a little bit of dot hemorrhaging — 6 months to a year is fine. If they have developed macular edema, this slide has since been updated, through the use of vascular endothelial growth factor injections, which I’ll touch on in a minute. If they have macular edema, they’re going to be coming monthly. And that’s because many of them may be receiving intraocular injections of anti-VEGF growth factor. So this has been revised. If they have macular edema that’s being treated with a laser therapy, it will depend on the location of the macular edema what the follow-up schedule is. It may be every 3 to 6 months, depending on location. That’s a more complex diagnostic issue. And the same is true for clinically significant macular edema, meaning visible at the slit lamp, and threatening central vision or in fact involving central vision. And we’ll touch on that in a moment as well. If they have the 4:2:1 rule, severe or very severe non-proliferative disease, they need to come at least every 3 months, and many of us recommend doing a prophylactic panretinal photocoagulation, once you make the diagnosis of 4:2:1, or severe non-proliferative disease. So those patients would be treated and then return in 2 to 3 months for follow-up. And certainly patients with proliferative disease would be treated with laser therapy, panretinal photocoagulation, they will be treated with either intravitreal anti-VEGF injections or a focal to grid laser treatment, which I’ll show in just a minute. If they have hemorrhaging, it depends on the location of the hemorrhage. Vitreous hemorrhage that’s non-clearing typically receives a vitrectomy after 1 to 3 months, rather than waiting a year for the hemorrhage to clear, and certainly retinal detachment is treated emergently or urgently, depending on the macular status, with vitreous surgery. Which technique is best? Many, many clinical trials have been performed to help determine when we should treat. The most important thing to know is that laser therapy or panretinal photocoagulation was developed 30 to 40 years ago, and at that time, many, many laser spots were given in an effort to treat the ischemic areas of the retina and reduce proliferation of new tissue. The number of treatment spots underwent a reevaluation over the course of several decades. The point I wanted to make is that the early treatment diabetic retinopathy study tried to treat diabetes with half the typically recommended number of laser spots. The typical number of laser spots recommended in the diabetic retinopathy study was in the range of 1200 to 1600 spots of 500 micron spot size, with a standard argon laser. Now, the early treatment diabetic retinopathy study tried to use 600 spots to reduce the amount of laser treatment, and that did not work. That is why as a baseline, it’s recommended to put in at least 1200 spots. Although many more may be needed. The diabetic retinopathy vitrectomy study showed that if a vitrectomy was required for treating vitreous hemorrhage, it’s better to do that in the first month or 2 or 3 after the hemorrhage to allow treatment of the back of the eye, rather than observation for a year, which would allow a progression of proliferative disease and perhaps detachment, without the physician’s being aware of it. And finally, the diabetic control and complications trial and many other trials show that controlling blood sugar is better for your retinopathy. And we of course have been telling our patients that for a long time. There is a huge economic boon to treating diabetic retinopathy, which saves many, many years of sight for patients. Treating diabetes efficiently reduces the number who progress to severe renal disease, or lower extremity amputation. And of course, extends their lifespan. Let’s touch on, now, treatment options. Laser treatment for diabetic retinopathy. We use laser to treat proliferative retinopathy, and also as one of our treatment options for diabetic macular edema. We know that based on these clinical trials, there’s quite a significant reduction in the rate of moderate visual loss in laser-treated eyes, and that vision can be stabilized in at least 60% of eyes. Clinically significant macular edema is retinal thickening that’s visible at the slit lamp. Another term for this is center-threatening or non-center-threatening macular edema. These criteria that I’m going to show you involve the number of microns in distance from the center of the fovea, where we see either a cloudy swelling, which you see here as a cloud, or lipid deposition. If you see cloudy swelling within 1 disc diameter of the center of the fovea, that is called center-involving macular edema. If it’s outside 1 disc diameter, then it’s non-center-threatening. So you don’t need to be totally picky about the number of microns we’re talking about here, whether it’s 500 microns or 1,000 microns. I think you can recognize that if you see swelling or lipid under the center of the fovea, that’s important and requires treatment. If you see cloudy swelling or lipid maculopathy further away, then you’ll determine — would focal laser treatment be helpful? Would anti-VEGF be helpful? Depends on the level of retinopathy. When we do focal or grid laser — and I should tell you, I’m emphasizing this, because many of you practice in a setting where you do not have anti-VEGF treatment available — in the US, much of the focal and grid treatment has been replaced by anti-VEGF therapy for macular edema. But I still think it’s important for you to be exposed to this form of treatment, until you have your anti-VEGF therapy more available to you. When we do focal laser treatment, we do a combination of focal or grid laser. Focal means treatment with a green or blue-green laser. 500 to 1,000-micron spot size, directly to the microaneurysms. The idea is to touch the microaneurysm with the laser. It helps induce fibrosis in the aneurysm. But if you miss the aneurysm, you don’t need to retreat it 3, 4, 5, 6 times. You are looking for a very light treatment spot behind the aneurysm. Because there’s some thought that stimulating the retinal pigment epithelium with the laser beam itself may induce reabsorption of some fluid. So you treat the aneurysm, but you don’t overtreat. Duration is 0.1 second. And you want to very lightly whiten the microaneurysm. There are areas of diffuse edema, where you don’t see microaneurysms. But a small swampy area. And that area you put a little grid pattern of treatment. Green or blue-green laser, 50 to 100-micron spot size, again, a 10th of a second duration. I usually put the smaller spots, the 50-micron spots, closer to the foveal avascular zone, and the larger spots as you move toward the temporal arcades. There is one treatment paradigm for diffuse edema, where the entire macula, except the papillomacular bundle and the foveal avascular zone is treated. This is a concentric grid treatment, and this has to some extent fallen out of favor, because the large number of laser spots many years later may lead to atrophy. But I wanted you to see this pattern, which is sort of a last-ditch effort, if you have nothing else available in diffuse edema. And see this image? The small black spots on the temporal macula are the areas of laser grid treatment after they’ve healed. And I just wanted to go ahead to panretinal photocoagulation. This is performed for the high risk characteristics we mentioned. An argon green, red, or diode laser may be used, and there are certainly pattern lasers that some people have access to at this point. The number of treatment spots with a standard argon laser is 1200 to 1600 as a baseline. More may be needed, depending on the patient. It depends on which laser lens you’re using. 200-micron burns or 500-micron. If you’re using a Goldmann 3-mirror lens, if you’re using most of the Panfundus fisheye-type lenses, the 200-micron burn is correct. And these spots will be placed 1/2 burn width apart, for 0.1 to 0.2 seconds. This treatment may require two or more sessions. Depends on the patient cooperation. You should not treat the optic nerve. You should not treat the macula. And you should avoid the papillomacular bundle. And this is a light to medium intensity grey burn, not a heavy white burn. We know that proliferative diabetic retinopathy, when treated, has a huge reduction in the rate of severe visual loss. Greater than 50%. And we know that we want to do prompt laser. Once the diagnosis is made, we wish to treat within a week or two, if we can at all schedule that treatment. Here’s an example of treatment. The macula is spared. The first treatment is generally started nasally, where you can adjust your spot size, and then extend it inferiorly, marking the temporal macula, so that when the second treatment is done, the macular area will be well circumscribed and visible to the treater. Here’s the second treatment in the same patient. You can see the superior fundus has been treated. And this is about 10 days post-treatment. All the laser spots have now turned a brownish color. This is directly after treatment in the fellow left eye. The nasal fundus can be treated first, to adjust the power. And then one can demarcate the macula, so that one does not wander into the macula in the second treatment, inadvertently. Here’s an example of a 37-year-old male truck driver with excellent vision who tells us he’s been diabetic for a year. We know that’s not the case, because he has diffuse neovascularization along the superior and inferior arcades. We looked at this gentleman earlier. If we’re going to use focal/grid treatment, we want to do that before doing our panretinal photocoagulation. And that’s because panretinal photocoagulation will cause macular edema in many patients, and we want to pretreat them, if we can. How do we do the treatment? We identify the eye, we mark the eye, so that we’re sure that we’re treating the correct eye, and confirm the consent. I usually dilate both pupils, because the patient often complains about the fellow eye, and you want to be able to look at that quickly. You turn the laser on, make sure it’s working, select your treatment settings, and then you decide on anesthesia. Most people can tolerate this with topical anesthetic drops. Some few people will need a peribulbar or even a retrobulbar injection. And in our clinical practice, I would say that’s about 10% of people. We talked about lens magnification already. This is a placement of the 3-mirror lens on the eye. And you can see that the lens has to be used with Goniosol, or a liquid within the lens. You want to be sure to get all the bubbles out of the liquid before you place the lens on the eye. Have the patient look up, look straight ahead, and then the lens, while it’s used, is going to be rotated for 360 degrees, to visualize the entire midperiphery of the retina. We mentioned the light and microscope being parallel, and often needing a slight tilt to the lens, to visualize the fundus, for the best view. Don’t put your foot on the pedal ’til you have everything set, and then at the last minute, turn the laser on. So you don’t inadvertently treat the patient. It’s better when you’re lasering to try to group your treatment spots. Otherwise you get kind of bored and you’re wandering all around the retina. I would do 16, 20, 30 spots, and then move to a different location, to be able to maximize your efficiency. If you have a patient that really does not involute or regress after full treatment, a few patients will require what’s called a pinhole technique. And this keyhole or pinhole involves sparing the papillomacular bundle, and just the very central macula, but treating the rest of the macula, in order to ablate further areas of tissue. Treatment complications — the number one complication is inadequate or incomplete laser treatment. So aim for those 1200 spots, at a minimum. And it’s also bad to treat the fovea. Corneal abrasions occasionally occur, from the contact lens. Don’t worry about treating a retinal blood vessel, because if you close it with the laser, it opens right up again. If you just hit it once or twice with a laser, it’s pretty tough. Rarely if your treatment spot is too intense, you can puncture Bruch’s membrane, and you’ll notice that you’ll see a little black spot on the retina, with a little associated blood clot next to it. If that happens, press with your contact lens, and I typically overtreat the area with laser, because it may be a source of choroidal neovascularization later. It’s rare, rare, rare to have secondary angle closure glaucoma. But maybe once every 2 or 3 years, that may happen, due to swelling of the choroid, after the laser treatment. And potentially, one could develop a retinal detachment, if the patient’s retina has a lot of fibrous tissue on the surface. That tissue is contractile, and may contract after laser. So we’re going to do our questions again. If systemic diabetes has been present for more than 20 years, what percentage of insulin dependent diabetics will have some level of diabetic retinopathy? The choices are 30%, 50%, 75%, or over 90%. And the answer is: Over 90%. If you’ve been diabetic 20 years, you can expect to have retinopathy, and we anticipate at least 99% of insulin dependent diabetics will have some level of retinopathy. After the diagnosis has been made, how often should a patient’s eyes be examined? This is after the diagnosis of systemic diabetes has been made. How often should they be examined? Every 6 months, every 12 months, every 2 years, every 5 years? And the answer is every 12 months. And that’s correct. The 4:2:1 rule for severity of non-proliferative diabetic retinopathy includes A, macular edema, B, vitreous hemorrhage, C, retinal hemorrhages greater than 20 in 4 quadrants, and D, retinal hemorrhages greater than 40 in 4 quadrants. And the answer is — wonderful! Excellent! 100%. Greater than 20 in 4 quadrants. Wonderful, thank you. Panretinal photocoagulation is defined as the following. A, 600 spots. B, 900 to 1100 spots, C, 1200 to 1600 spots, D, 2000 to 2500 spots. 1200 to 1600 spots is the correct answer. Excellent. 83% correct. Wonderful. And that was with a standard laser. It’s true that more spots may be required, and certainly more with the pattern laser as well. Focal/grid laser for clinically significant macular edema includes treatment of macular edema, neovascularization, microaneurysms, A and C. Excellent. A and C is correct. Both microaneurysms and macular edema. I’m just quickly showing a couple of case studies. These are areas of neovascularization. Subhyaloid hemorrhage. Subhyaloid hemorrhage with neovascularization. These are more pattern recognition images. Here’s a patient with new floaters. He has vitreous hemorrhage. And here’s a patient with a grayish appearance of fibrous tissue, which on angiography shows active lightbulb signs, neovascularization, multiple lightbulb signs, neovascularization, proliferative disease. Here’s a person with excellent vision, 20/20 in his right eye. He checks his sugar once in a while. But on angiography, we see early lightbulb sign, showing he has proliferative disease. You may see many different patterns of photocoagulation, depending on the doctor who does it. Here’s a patient with very closely spaced, excellent laser treatment, but 10 years later, they have a lot of pigmentation. This patient had 20/20 visual acuity with a fibrous ring surrounding his macula. He called, saying — I’ve gone blind last night. Why can’t I see? And he’s already had complete panretinal photocoagulation. The answer is he detached his central macula. If you look at the macula, you can see these little kind of squiggly folds in the macular tissue. This is called surface of the brain. And this is a typical pattern of retinal detachment. So we know that this fibrous tissue pulled his retina away from the back of the eye, dropping his vision to hand motion. And this is an OCT, showing a focal macular retinal detachment. So he had surgery and regained his vision. A few of you will have access to these pattern lasers, which allow you to place 9 or 16 spots at a time. The important thing about this is, number one, it’s more efficient, but number two, you need to place many more spots. In the range of 2400 spots, to get an equivalent panretinal photocoagulation. So these are some of the highlights of skills training in ocular photocoagulation. I mentioned the use of anti-VEGF drugs, and I’m sure you’re aware that there are 3 drugs at least available now, which are injected into the vitreous cavity, in an effort to treat diabetic macular edema or choroidal neovascularization or vein occlusion, primarily. From the point of view of diabetes mellitus, we see these patients now monthly, if they have macular edema, and if they have macular edema that’s center-threatening, we inject them with anti-VEGF medication. And that would be with Avastin or bevacizumab, Lucentis or ranibizumab, or aflibercept, which is called Eylea. Our first line of therapy is typically Avastin or Lucentis. And we keep doing that monthly, for a year. If there is edema, we give an injection, come back in a month, injection, come back in a month. If there’s no edema, we skip that month and say come back in a month. The typical patient needs 8 to 10 injections in the first year, 3 to 4 injections in the second year, and 1 or 2 injections in the third year. So this is the primary role of drug therapy for diabetic macular edema, which has essentially replaced much of our focal and grid laser treatment. Many, many clinical trials support this. Lastly, the occasional patient may benefit from intravitreal steroid injections as well. I do have a couple of questions here, that came in. One of them was: If a patient needs both panretinal photocoagulation and anti-VEGF treatment, which do you do first? I think the practical nature of this is you may not see the patient very many times. So I would do the anti-VEGF treatment and the first half of the PRP the first time I see the patient. I would do the laser early in the day, and then I would do the anti-VEGF following the first laser, on the same day. And that’s just because you’re pressing on the eye with a contact lens to do the laser. And in theory, I suppose one could potentially open up the little tiny 30-gauge hole you used to inject the anti-VEGF. So it likely doesn’t make any difference, but my personal preference would be laser and then injection, or if they’re staying in town, I would have them come back the next day for the injection. So there’s not a hard and fast rule about that. Of course, the downside of the panretinal photocoagulation is it may cause macular edema, and by giving the anti-VEGF, you’re sort of pretreating for that eventuality as well. And the other question was: For someone with focal exudates but good vision, what is your advice? You can have terrible exudates. Many circinate rings surrounding the fovea. And have perfect 6/6 or 20/20 visual acuity. So if there’s a treatable lesion, I still treat it, regardless of what the vision is. If there was only minor non-vision-threatening lipid, and I discuss this with the patient, including the very tiny risk of endophthalmitis from an anti-VEGF treatment, one could make the argument for observing a mildly involved patient. But a generic person at risk for loss of vision or who’s had central loss, even if the vision is good — you certainly can have some central macular edema with 20/20 visual acuity, as you see on your OCTs. I would still err toward the side of treatment. And by that, I mean treatment with anti-VEGF therapy. And the last question was: Why are people — why do people go blind in diabetic retinopathy? The major causes of blindness are diabetic macular edema or retinal detachment due to untreated proliferative disease or longstanding vitreous hemorrhage due to untreated proliferative disease. And lastly, I have a capillary non-perfusion under the fovea. As far as we know, we can’t treat. I will say about a month or two ago, at our big American Academy of Ophthalmology meeting, there was some new data presented, showing that anti-VEGF injections can actually reverse or reperfuse or improve the perfusion of the retina. And that’s a very new finding. So that would be thrilling, if it turns out to be true, that we can treat non-perfusion with the anti-VEGF drugs. So I think we’re gonna come to a conclusion. If there are any other questions, either now or… Later.

Download Slides


December 1, 2016

Last Updated: October 31, 2022

3 thoughts on “Lecture: STOP Diabetic Retinopathy”

    • Hello Armia Rifaat,

      We’re so sorry that you’ve had trouble downloading the video. I’ve just tested the link in a few different browsers and it appears to be working. Perhaps it was an issue with internet connection? Here’s the link to the video in Vimeo, which gives other options for downloading: Stop Diabetic Retinopathy Vimeo. Please let us know if you still have trouble downloading the video.

      All best,
      The Cybersight Team


Leave a Comment