Lecture: Diabetic Retinopathy

The objective of this lecture is to review the diagnosis and treatment of diabetic retinopathy including medical considerations, laser, intravitreal injections and Vitrectomy.

Lecture Location: on-board the Orbis Flying Eye Hospital, Kingston, Jamaica
Lecturer: Dr. David G. Miller, Retina Associates of Cleveland, Ohio, USA


DR MILLER: Well, thank you, everyone, for attending. Thanks for the introduction, and thank you to Orbis and Cybersight for inviting me to speak here today from the Flying Eye Hospital in Jamaica. As Maria mentioned, I’m Dr. David Miller from Cleveland, Ohio, Retina Associates of Cleveland. So today’s objective is to review diagnosis and treatment of diabetic retinopathy including medical considerations, laser, intravitreal injections, and vitrectomy. And we’re gonna go through a few prelecture questions. So the first one is: Panretinal photocoagulation can reduce severe visual loss by 50% in high risk proliferative diabetic retinopathy. These are people’s answers. But we’ll review it again at the end. The next question is: Complications of panretinal photocoagulation can include which of the following? Night vision loss, peripheral vision loss, decreased visual acuity, or all of the above? Okay. All of the above. They took the hint. The next question: Instituting tight glycemic control may cause a short-term progression of diabetic retinopathy? True or false? Another one of my favorite test questions, because it’s 50/50. You’ve got a good chance. Right? There we go. Very good. So… On with the lecture. So diabetic retinopathy, public health issues. Leading cause of new cases of blindness among working age people worldwide. People with diabetic retinopathy have 25 times the chance of being blind than non-diabetic people. Diabetes, magnitude of complications. As a disease, diabetes mellitus is a systemic disease affecting the microvasculature throughout the body, including the eyes, the kidneys, and the heart. And some other organ systems. But those are the main three. Diabetic retinopathy prevalence. For type Is, if you’re in your first five years after diagnosis, the chance of having diabetic retinopathy is fairly low. Less than 13%. If you’re over 10 to 15 years in duration, the chance of showing some diabetic retinopathy is as high as 90%. So the longer you’re diabetic, of course, the greater your chance of having some diabetic retinopathy. And a type II diabetic — if you’re under five years’ duration, there’s a higher chance of having diabetic retinopathy. 24% of those not taking insulin, 40% of those on insulin. And of course, with longer duration, the numbers go up again, just like in type I. Rising to 53% and 84% for non-insulin-taking type I diabetics. Type II diabetics, I’m sorry. So medical considerations: The most important tool for treating diabetes mellitus and diabetic retinopathy is prevention, and the best way to do that is to have medical management of the diabetes mellitus. The diabetes control and complications trial, which my practice was fortunate enough to participate in back in the ’80s and ’90s, proved beyond a doubt that tight control of blood sugar can prevent diabetes mellitus eye complications and slow the progression of diabetic retinopathy. So the DCC trial again showed intensive treatment of diabetes mellitus — and these numbers are pretty astounding, when you think about it, compared to our medical treatments and what we can do surgically and with laser, which is later in the lecture. But if you just look at these numbers, it’s really the most impressive results in the whole talk. You can reduce the risk of developing retinal disease by 76% with intensive blood sugar control. You can slow the progression of the disease by 54%. Reduce the risk of developing severe retinal disease by almost 50%. And reduce the need for laser by over 50%. And none of our laser treatments are that effective. Some of the anti-VEGF treatments approach that, but again, that’s ongoing treatment. So tight control has preventative value for retinopathy. Instituting control does not reverse existing disease, necessarily. And sudden tight control may cause a rapid retinopathy progression. Which is something to watch out for, in the first year or two of someone going from very lax control of blood sugars to going to tight control. You want to watch out that there won’t be a step up in the retinopathy progression. I did want to mention about the tight control. My own experience over the last 20 years, practicing in Ohio, in Cleveland, that I’ve seen much less diabetic retinopathy, and certainly the severe forms of retinopathy, have dropped as these types of recommendations have been implemented across the medical community. So in a community where patients have access to the doctors and the hospitals and the medications to control their diabetes, and education gets out there, we’re seeing a definite drop in severe cases of diabetic retinopathy, in surgical-type cases. Instead of doing a surgery once or twice a week for severe diabetic retinopathy, I’m only seeing one or two a month now. And when I go out into the more rural areas of Ohio, where maybe the medical care is less intense, I still see more severe disease. So there’s no doubt that this is really the way to treat this condition. Control the diabetes, if possible. So classification of diabetic retinopathy, just to review for everybody a little bit: There’s non-proliferative diabetic retinopathy, NPDR, with or without DME, diabetic macular edema. It’s an early stage. It shows intraretinal microvascular changes. It was formerly called background diabetic retinopathy. A term that we’re gonna avoid. The other main division of diabetic retinopathy is called proliferative diabetic retinopathy, or PDR. PDR with or without diabetic macular edema can occur. It’s a more advanced stage. And what is the hallmark of PDR is the growth of extraretinal neovascularization on the retinal surface. So this fundus photograph is a nice example of non-proliferative diabetic retinopathy, showing hard exudation or lipid material that’s being deposited, microaneurysms, and blot hemorrhages. And the macula is right here. And that’s also a photograph of diabetic macular edema. Proliferative diabetic retinopathy will show neovascularization, including rubeosis, which is new blood vessels growing on the virus, of course. Fibrous proliferation, bleeding and traction, retinal detachment, and blindness. And can have all the findings of non-proliferative disease, including macular edema, hard exudation, microaneurysms, scattered hemorrhages. A photograph here, a couple of fundus photographs. One is neovascularization of the disc. The abnormal blood vessels are the lacy ones. These are the normal retinal vasculature, coming off the big vessels. The abnormal vessels are the lacy ones coming off the disc. And in the peripheral view, again, that lacy pattern of neovascularization growing off the main retinal vessels are the abnormal findings. In this photograph of the anterior segment of the eye, the iris, you can see the large amount of neovascularization over the iris. And in this photograph, another example of proliferative diabetic retinopathy, you have a preretinal hemorrhage, a hemorrhage trapped on the retinal surface, from neovascularization, where the vessels broke open and bled over the retinal surface, and now blood’s trapped between the back of the vitreous and the front of the retina. So the talk — a little bit about diabetic macular edema. On this slide, we’re gonna call it clinically significant macular edema, which is a study term. And that’s basically diabetic macular edema that can occur with non-proliferative diabetic retinopathy, and again, also with proliferative diabetic retinopathy. Macular edema, diabetic macular edema, is retinal thickening involving or threatening the macular center. And in the rest of this talk, I’m just gonna call that diabetic macular edema, instead of focusing on the more cumbersome term “clinically significant macular edema”, which, again, was what was used in the trials. We participated in those trials too. Early treatment of diabetic retinopathy study, ETDRS. And that clearly defined what was considered treatable macular edema. But in practice, diabetic macular edema that’s a threat or causing vision loss you’ll treat, and whether we have to stick to the rigid rules of CSME is a bit of a debate. So the treatment of DME includes medical treatment. Again, I’m always gonna come back and emphasize: Control the diabetes. ETDRS, which is the early treatment of diabetic retinopathy study, showed the benefit of laser for diabetic macular edema. The DRCR network, diabetic retinopathy clinical research network, has run multiple trials, showing the benefits of laser, triamcinolone injections, ranibizumab and aflibercept injections, and also bevacizumab injections, which were also shown or looked at in the BOLT study, laser and the bevacizumab, which is Avastin. So this is Avastin, this is Lucentis, and this is Eylea, to go with the trade names. So to define clinically significant macular edema, it’s retinal edema in the trials that were anywhere within 500 microns of the macular center. Hard exudates at or within 500 microns of the center if associated with retinal thickening. Zone of thickening greater than 1 disc area, if located within one disc diameter of the macular center. Treatable retinal lesions. Discrete points of hyperfluorescence as focal leaks, greater than 500 microns from the macular center. How do you judge that when you’re looking in the eye? Right? We don’t have a ruler in there. You can generally think that the disc diameter is about 1,000 to 1,500 micrometers, so half a disc diameter away from the fovea is probably a safe range to start your laser. At least for the first time. You can always try to tighten that up. Because you’re probably at 700 microns there. But the initial focal laser — I would recommend going about half a disc diameter from the foveal center all the way around, and no shots closer than that to the fovea. Areas of diffuse leakage can also be treated in a grid pattern. So here we have a fundus photo, showing diabetic macular edema. We call that a circinate ring of exudate, and down here the ring of exudate — some scattered exudates and hemorrhages up temporally. Here’s the optic nerve again. The center of the macula is right here. So you’re certainly within 500 microns, with a large area of associated edema. And we’re gonna show the corresponding angiogram, showing where this leaking is coming from. Again, remember, the circinate exudates — the ring of exudates are down here, around this cluster of leaking points, and there are more exudates up there, around those clusters of leaking points. And so laser therapy was then done, focal laser. You can see the whiteness of the laser burns here. Some blanching. And some whiteness up there. Superotemporal. And we wait about six months to a year, and you can see the resolution of the hard exudates. That typically takes more like a year. A little bit of exudate remaining here, but most of it’s gone. You see some atrophy from laser scarring here. Ideally we would not have laser that intense. You see some of the laser scars also superotemporal. There’s still some background diabetic retinopathy here, non-proliferative retinopathy, but not threatening the center, and there’s a nice response. You can probably wait. Here’s a photograph of someone following a macular laser for macular edema. And the center of the macula is right here, associated with some exudates and edema centrally, and a very tight pattern of focal laser was placed. That’s a little close for comfort in my book. You’re probably within 500 microns of the foveal center. Well within it. And I would reserve that type of treatment nowadays for someone who’s failing laser at least twice before, or you don’t have access to anti-VEGF injections or other treatments. You generally try — not the laser quite that close to the fovea in that intense or surrounding pattern, nowadays. But if laser is all you have, anti-VEGF injections are not an option, because of cost, or patient hesitancy to have a needle stuck in their eye, you can get more aggressive with the laser. So the results of the macular edema trial and laser showed that you can decrease the risk of moderate visual loss. Actually, that’s severe visual loss, which is three lines. That’s a typo. If you’re changing vision by three lines, we consider that severe visual loss in my book. You can decrease the risk of that. You increase the chance of visual gain, and you can reduce macular thickening. By the numbers, you reduce the three-year risk of losing greater than three lines by 50%. The control group that was not lasered — 24% lost 3%, and the laser treated group, lost 12%. 12% lost that many lines. The thing that’s interesting on this data is that both the laser group and the control group are still losing vision. Some people are still losing vision. 24% in the control group and 12% in the laser group. So you still have a lot of people not doing well, even with laser. Which I just mentioned. And only 17% of eyes improved 15 letters. And at 24%, almost 1 in 4, still had persistent edema at the fovea in three years. So laser was effective, but not great. It was just better than nothing. Other problems with laser treatment for diabetic macular edema includes laser scars and resultant scotomata, which increases with time, some loss of color vision, likelihood of vision improve is low. Which brings us to other treatments for diabetic macular edema. And the advent of anti-VEGF treatments, antivascular endothelial growth factor treatments, which is injections in the eye of Lucentis, Eylea, and Avastin, have proven to be very useful. And the reason that is — is because in diabetic retinopathy, VEGF levels are increased in the retina and in the vitreous. VEGF is a protein produced in the eye that causes leakage of normal vasculature. Which then leads to edema. If there is enough VEGF present, it also causes the growth of neovascularization. So if we can block VEGF with another protein, which is an anti-VEGF injection, we can stop the pathogenesis. So let me get the lines put in here. I’ll just put them all in. So looking at this graph, pretty simple to understand. This is your mean change in vision, and going up is good. So you’re gonna gain more letters going up. These are the number of weeks done in the trial. And this trial was comparing sham injection and laser treatment, Lucentis or ranibizumab and laser, ranibizumab and no laser, and triamcinolone plus laser. And the groups that did well, the top two lines, the ones who gained the most vision at the primary endpoint and even beyond the primary endpoint, are the patients who got the ranibizumab injections with or without the laser. And the group that did the worst was the triamcinolone with laser or the laser alone. Pretty equivalent. So intravitreal ranibizumab summary shows that with prompt or deferred focal grid laser, it had superior visual acuity in OCT outcomes versus laser alone. Over 50% of eyes had substantial improvement, greater than 10 letters, while 30% gained greater than 12 letters. Substantial visual loss, defined as greater than 10 letters’ loss, was uncommon, and results were similar with or without laser, as long as you were getting the ranibizumab. You did much better. So what do we summarize to this point? We know that the ETDRS study 25 years ago showed focal laser is effective for DME. Not great, but effective. The DRCR network showed in 2010 that ranibizumab with or without laser is superior to laser alone. Through two years of follow-up. And that with ranibizumab, significantly more eyes gained substantial visual recovery more quickly. Other anti-VEGF results. The BOLT study looked at bevacizumab or Avastin. May be as effective as ranibizumab. And other studies include the RISE and RIDE study, which were for Lucentis, showing the effectiveness of that injection for diabetic macular edema, and the VISTA and VIVID studies, showing the benefit of Eylea or aflibercept, better than laser, for that study, and then the DRCR.net Protocol T, comparing multiple drugs and Avastin, showing that it is not inferior to those agents, overall. So the treatment — to switch gears here — the treatment of proliferative diabetic retinopathy is a similar scenario, where anti-VEGF injections and laser can prevent visual loss from proliferative diabetic retinopathy. Panretinal photocoagulation is the type of laser performed for proliferative diabetic retinopathy, as opposed to focal laser for diabetic macular edema. A panretinal laser is more broad in scope, and in the peripheral retina. So looking at the studies done on this in the 1980s, the diabetic retinopathy studies showed that panretinal photocoagulation or PRP can reduce the risk of severe vision loss by about 50% in high risk PDR. High risk PDR was defined in the study of certain parameters where neovascularization of the disc is greater than standard photo 10A, about 1/3 of disc area. Neovascularization with preretinal or vitreous hemorrhage was also considered high risk. And neovascularization elsewhere with preretinal or vitreous hemorrhage. So laser surgery was most beneficial in patients with NVD or NVE with vitreous hemorrhage. And laser surgery for retinopathy is most effective before visual loss occurs. Vision loss occurs from bleeding, most typically in these patients, from vitreous hemorrhage. Very tough to place laser after the hemorrhage occurs. Better off to catch them with the beginning of neovascularization and putting the laser in at that point. So how is panretinal photocoagulation applied? It’s an outpatient procedure, 1,000 to 2,000 burns in the retinal periphery, placed over 1 to 3 sessions. And here’s a photograph of recent placed panretinal photocoagulation, or just placed. These white marks are the burns, very intense burns, in the retinal periphery. These are a cluster of burns that are confluent over an area of neovascularization. I typically don’t do this myself. I don’t try to put confluent burns over neovascularization. I tend to just put the pattern in the periphery, and probably less intense than this in general. That’s almost like endo laser burns from surgery rather than burns placed in the clinic. Here’s a little histopathological slide of a laser burn. So here’s the retinal layers. Right? RPE here. And over here is a laser scar, where the retina is completely obliterated. Scarred down to the Bruchs membrane in the RPE. And here’s a fundus photograph of pretty aggressive panretinal photocoagulation, already placed. These are the laser burns in the periphery. The white areas and the dark ones are all laser burns. The optic nerve — these burns are very close, up to the optic nerve. I think today very few retina specialists place burns that intense or that closely linked or that close to the nerve. Now, depending on your population and your community, that’s still fine. This works. Maybe you have to get more aggressive, because you don’t have the advantage of the injections to help you. With the advent of using anti-VEGFs, we definitely have backed off, I think, in the intensity of panretinal photocoagulation. A little before and after photograph of proliferative diabetic retinopathy. Illustrating the regression of neovascularization on the disc. Again, the neovascularization is the lacy pattern of vessels around the disc. The normal retinal vasculature are the big straight lines. And after the panretinal photocoagulation is placed, you can see the laser scars in the periphery there. You can see the lacy pattern of vessels have largely regressed. Just a little residual there, right over the disc. There are some side effects to panretinal photocoagulation that one should be aware of. Some loss of night vision. Some loss of side vision. And occasionally a slight decrease in central vision or color vision. Another stage of proliferative diabetic retinopathy is vitreous hemorrhage. The symptoms here is: A shower of dots or floaters. Vision may decrease to the light perception-only range. This is a type of vitreous hemorrhage. This one is actually trapped against the retinal surface. Because there’s not yet been a vitreous separation. So this blood is stuck between layers of the vitreous and the retinal surface. Neovascularization ruptured in that space, filled that potential space with blood, covering the macula. That patient would typically be seeing in the count-fingers range, since the central macula is covered. So treatments for non-clearing vitreous hemorrhage: If it clears, you can go on with panretinal photocoagulation. If it clears over a few weeks or months. If it’s not clearing, vitrectomy surgery becomes your option. It’s to remove the vitreous hemorrhage, to treat or prevent retinal detachment, and also place the panretinal photocoagulation to treat the underlying proliferative diabetic retinopathy. So before and after slides of vitrectomy surgery. Vitreous hemorrhage here. Through the vitrectomy, the hemorrhage is gone. And then the end stage of proliferative diabetic retinopathy results in diabetic retinal detachment, fibrous tissue from new vessel growth contracts and pulls on the retina, detaching it. It can also tear the retina, so you have a combined rhegmatogenous and tractional retinal detachment. It results in permanent vision loss if uncorrected, or if the macula is involved. And it’s treated with vitrectomy techniques. And here’s a photograph of a tractional diabetic retinal detachment. Large amounts of neovascularization. These large vessels down here are all neovascularization. Fibrous tissue. The optic nerve is somewhere buried under here, under the fibrous tissue. The macula is being pulled this way. The retina is detached. So the basic schematic of vitrectomy surgery is to put instruments in through the eye at the pars plana level. Use a vitrector to cut away the vitreous gel. Pull away the neovascular fibrous tissue, and get an outcome like on the right. So here’s the pre-op photograph. Tractional retinal detachment. Fibrous tissue. Optic nerve. Vitreous hemorrhage, of course. And then the postop photograph, with the panretinal photocoagulation in place. The macula is spared there in the center, for a nice couple, a nice visual improvement. Yes? >> Do you know the change in visual acuity for that patient? DR MILLER: You know, I don’t know it offhand, but just by looking at the photographs, you’re probably in a 20/400 vision here. The macula is not really in the detachment, covered by blood. This out here, this detached retina, is outside the macula, which is fortunate. So just by taking that hemorrhage off the macula, you’re coming back to 20/30 or 20/25. It can be very, very good. If this was left untreated, this detachment that’s outside the macula here is gonna progress right through the macula. You end up with a very poor outcome. Hand motions. So to cover a little bit about intravitreal injections, again, for anti-VEGF, sometimes steroids, in proliferative diabetic retinopathy, there’s bevacizumab, ranibizumab, aflibercept, and triamcinolone, and Ozurdex, which is dexamethasone. All those can be used for diabetic macular edema. And proliferative diabetic retinopathy is more often just the anti-VEGFs. But all of these agents have been shown to induce regression of neovascular tissue in proliferative diabetic retinopathy. Recurrent injections could be required, and I think the most effective place to use these anti-VEGFs really, for proliferative disease, is really just for presurgical cases. So to treat someone with proliferative disease, with anti-VEGF injections, that can be done, but multiple injections may be required over quite a long time, and there’s always a risk of recurrence. Laser has a benefit of being a permanent change in the eye. Where you don’t need as much patient compliance and follow-up. And my thoughts on that have always been: You know, the patients who get into trouble with their diabetes and develop proliferative diabetic retinopathy are the same ones not going to the doctor and being treated. And to think they’re gonna come show up for their eye shot every other month and be monitored for years on end for possibly more eye shots, I think, is a little risky. So I think there is some advantage in still using panretinal photocoagulation over anti-VEGF in PDR. Just because of the patient population. You have a really compliant patient who’s a changed person and going to the doctor all the time and just didn’t know they had diabetes, then you can make the argument: Anti-VEGF injections on an ongoing basis, monitoring on an ongoing basis, can be just as effective, and studies have shown that. But as far as using these injections before surgery, a single shot of any of them, Avastin or Eylea, anywhere from two or three days prior to vitrectomy up to two weeks, can make the surgery much simpler to perform, much less bleeding. I think it’s been a real game changer in the last ten years, for vitrectomy surgery in diabetic retinopathy. And I believe I have a video here. Yeah, we do. So just a vitrectomy in a patient with PDR who got an Avastin injection, and the thing that’s interesting here: Optic nerve here, and you have all this regressed neovascular tissue over the optic nerve, coming over the macula, causing tractional attachments, and we can strip that off with the vitrector, with virtually no bleeding underneath. Right? And so this was impossible to do this easily ten years ago. And so just a much simpler case, more predictable, quicker case, better postoperative visual recovery sooner. Because you don’t have the bleeding afterwards. And the panretinal photocoagulation is already in this eye. Certainly that would have helped too, to induce some of that regression. But just the extra Avastin shot before surgery — simple to do, and certainly makes the case much more effective. You’ll see here how the vitrector almost can just grab — a little grainy on the video, sorry — and there’s a little bleeding here on the optic nerve, where we separate the neovascular tissue, but we can take that fibrous tissue right up, here, off the superior arcade also. It’s almost gonna jump into the vitrector, as it kind of snaps all the attachments, being very brittle, at that point, from the regression. So conclusions on diabetic retinopathy: That early treatment may prevent blindness. That’s both for diabetic macular edema and non-proliferative disease, and also for proliferative diabetic retinopathy with or without diabetic macular edema. Improved screening can ensure early treatment. And visual loss is a late symptom of diabetic retinopathy. I would add one more point, in that good diabetic control, blood sugar control, can prevent all the complications in the eye. So I think we’re gonna revisit the questions here again. So panretinal photocoagulation can reduce severe visual loss by 50% in high risk proliferative diabetic retinopathy? True or false? And the answer is true. Very good. Complications of panretinal photocoagulation can include which of the following? Night vision loss, peripheral vision loss, decreased visual acuity, or all of the above? Correct. All of the above. And instituting tight glycemic control may cause a short-term progression of diabetic retinopathy? True or false? So… You know, the DCCT trial showed the benefits of tight glycemic control. But in the first year or two, you have to watch out a little bit. Because the answer on this is true. There can be an acceleration of the diabetic retinopathy, seen especially in the first year. So I think that concludes the lecture on diabetic retinopathy, and we tried to cover the objectives, the treatments of diabetic retinopathy, medical considerations, laser, intravitreal injections, and vitrectomy surgery. Thank you very much for your participation on the questions and your attention.

April 11, 2019

Last Updated: October 31, 2022

Leave a Comment