The objective of this lecture is to review the etiology, diagnosis, complications, and treatment of retinal vein occlusion.
Lecture Location: on-board the Orbis Flying Eye Hospital, Kingston, Jamaica
Lecturer: Dr. David G. Miller, Retina Associates of Cleveland, Ohio, USA
DR MILLER: My second talk this morning is on retinal vein occlusion. And the objective of this lecture is to review the etiology, diagnosis, complications, and treatment of retinal vein occlusions. So a few prelecture questions again. So complications of retinal vein occlusion include: Neovascular glaucoma, retinal neovascularization, macular edema, vitreous hemorrhage, or all of the above. Okay. The audience is going with all of the above. They’re very much in line with each other. A lot of concurrence. Laser treatment is definitely effective for which of the following? I think the keyword here is “definitely effective”. So okay. Pretty good concurrence there. One and three. Looks like one or two people are picking number five, all of the above. We’ll go through the lecture and see what we find out. Atypical retinal vein occlusion. So I’m talking about younger patients or bilateral vein occlusions, should be investigated medically for an underlying cause. True or false? 100% true. The audience is very convinced. Last question. So complications of intravitreal injections include… And the audience again, pretty concurrent there. So let’s start the talk and see what we can find out. So retinal vein occlusion — I’m just gonna use the abbreviation RVO to mean both branch and central — so the general retinal vein occlusion… Etiology and a central retinal vein occlusion, CRVO, there’s a thrombus or a clot at the central retinal vein near the lamina cribrosa, that’s in the optic nerve, as the vein exits the nerve. And our branch retinal vein occlusion — there’s a clot or a thrombus at an arteriovenous crossing point, so the retinal artery will cross over the vein, cause a little compression there, typically as you’re getting older, there’s more compression, and a clot can form at that crossing point. Risk factors in retinal vein occlusion include: For CRVO, hypertension, high blood pressure, open-angle glaucoma, diabetes mellitus. And branch retinal vein occlusion: Risk factors are more or less the same. High body mass index. But not diabetes mellitus. And again, the central pathology — I’m just gonna mention on the BRVO here that at that compression point, where the artery crosses the vein, it’s thought that there’s some turbulence set up, as that compression gets worse over time. That turbulent blood there is clot-inducing. Pathogenesis, a little different story here, between non-ischemic and ischemic vein occlusions. So you get a vein occlusion, whether it’s partial or complete. If it’s partial, we’ll call it venous stasis, just to get a little more specific. There is some VEGF created in that situation by the eye. And you get the development of macular edema. If you get capillary closure or complete lack of blood flow to areas of the retina, it’s more ischemic vein occlusion. You get a higher production of VEGF, and you can get macular edema certainly, but you also get neovascularization. So retinal vein occlusions can cause retinal ischemia, neovascular complications such as glaucoma, vitreous hemorrhage, and retinal traction, and macular edema. The primary prevention here is your maximal management of associated risk factors. So again, people who are under treatment for hypertension, hypercholesterolemia, diabetes are probably gonna be in much better shape than if they’re untreated. I think your risk goes up. For the diagnosis, it’s mainly done on the examination. Intraretinal hemorrhages, dilated veins. Cotton wool spots and blood — often called a blood and thunder appearance — and of course macular edema. Here’s a couple photographs of branch retinal vein occlusions. Mild on the left and moderate on the right. And you can just judge that by the amount of hemorrhage you’re seeing in each, and also the size of each. This is relatively small. It affects the macula, still. That one’s a bigger vein occlusion. And in a central retinal vein occlusion, the blood and thunder appearance would be more typically the one in the middle. Okay? Down on the bottom left is more of a perfused or mild central retinal vein occlusion. The veins are dilated and tortuous. Maybe some edema on the disc. Scattered hemorrhages. And a few cotton wool spots popping up around the arcades. And the photograph on the top right — more of a special type of CRVO. We call that like a hemiretinal vein occlusion. You could argue to put that into the branch occlusion category, but clinically, they respond to treatments and outcomes more like central retinal vein occlusions. So I tend to lump it that way. Instead of PRVO. So the prognosis on retinal vein occlusion is highly dependent on the anatomy and the geography inside the eye, of what’s involved. As you can tell by some of those photographs, how severe is the occlusion? How much blood? And also the location. Central macula or not central macula. And the history: Patients often present with acute loss of vision. Maybe some risk factors. Hypertension being the most common. On the physical exam, we like to get the visual acuity, get the tension, of course. Make sure there’s not glaucoma going on, or neovascular glaucoma, with a highly elevated pressure. And then we want to look at the slit lamp to make sure there’s no vessels growing on the iris, especially in the cases of the central retinal vein occlusion or the hemiretinal vein occlusion. Enough VEGF can be produced and induce rubeosis and neovascular glaucoma. Testing that can be useful is both fluorescein angiography and optical coherence tomography. Fluorescein angiography or FA is useful in determining the degree of ischemia, lack of blood flow, and the OCT is very useful to monitor the macular edema, especially when you’re under treatment with anti-VEGF injections or laser. So here’s some of the diagnostic testing examples. A little bit of a busy slide. But just wanted to give you the idea of what it can show. Here’s a vein occlusion, shot in sequence. You know the transit phase, the arteriovenous filling phase and the late phase. And this is a pretty severe occlusion. The retinal veins are not filling. In this top right panel. And even a little later, they’re still not filling. And in the very late photos, the veins do eventually fill, but you have all this dropout, areas of no capillary perfusion. So that’s a very ischemic CRVO. The slide below it is very similar. Lack of flow through all these vessels that are missing, and the lack of capillary flow also. The top right is showing leaking in the late phase, on the angiogram. All the brightness there in the middle is the fluorescein dye leaking out in macular edema. And if you look on the slide, you probably have something like 1,000 microns of edema there in the central macula, whereas normal should be 250. And a little closer view of a macular OCT in a vein occlusion, showing that you can get fluid collecting in the retinal layers, and also subretinal fluid in severe occlusions. So retinal vein occlusion signs: Intraretinal hemorrhage, cotton wool spots, macular edema, subretinal fluid, collateral vessels on the disc, or elsewhere, if it’s chronic. Iris and retinal neovascularization. Dilated and tortuous veins. And ghost vessels. Common symptoms: Central or peripheral loss of vision. The differential diagnosis is somewhat limited, so it’s often just made by a history and physical exam alone. But we do have a few that we can consider. For other workup. Older patients with cardio risk factors, probably no laboratory tests are needed. If you know you’ve got an older patient with signs of hypertensive retinopathy, AV nicking, they’re on antihypertensives, we can assume it’s just a standard vein occlusion and move on. In atypical cases, so younger patients, bilateral or recurrent vein occlusions, I think it’s worth considering, certainly, complete blood count, fasting serum glucose, protein electrophoresis, homocysteine, serum viscosity, and thrombophilia screening, including factor V Leiden mutation, protein C or S deficiency, antithrombin III deficiency, antiphospholipid antibodies. A few other diagnoses that can mimic the appearance of a retinal vein occlusion include ocular ischemic syndrome, diabetic retinopathy, human immunodeficiency virus or HIV retinopathy, hypertensive retinopathy itself can mimic vein occlusions, and any retinopathy related to blood dyscrasias, like severe anemia or leukemia. So here’s a few photographs of not vein occlusions, and we’ll see if you can kind of take a guess — it does match up with this slide, actually. So it’s the first four. Ocular ischemic, diabetic, HIV, and hypertensive. And each one of these is one of those diagnoses. And I’ll try and point out a few differences. The first slide on the top left is ocular ischemic syndrome. The reason that’s not a vein occlusion is, one, the retinal veins are nice and straight. They’re not dilated and tortuous. Right? There’s no impediment to venous blood flow. It’s kind of like if you have a plugged up garden hose, the hose will kind of swell and be snake-like. The same thing happens in the veins. So if the blood is flowing, the veins stay straight. There’s no disc edema, which you typically see in a central retinal vein occlusion. At least a little bit of disc edema. There’s no cotton wool spots. And what’s really remarkable in an ocular ischemic finding is all the hemorrhages tend to be peripheral, probably in the midperiphery, and farther out, and they’re really large dot-blot hemorrhages. It’s not the flame-shaped hemorrhages in the nerve fiber layer. So it looks somewhat like vein occlusions, but when you really study it, it looks quite different. The next slide in the top right is a fundus photograph of diabetic retinopathy. So some overlap with vein occlusions, right? We see the veins are a little dilated and tortuous. So diabetic retinopathy, as it progresses from non-proliferative towards severe non-proliferative to proliferative diabetic retinopathy, the veins will become somewhat engorged and tortuous, in many of those cases. You see the scattered hemorrhages. Maybe a few cotton wool spots. But not really any disc edema, and not the diffuse global appearance of the blood scattered across the fundus. And of course, you have the underlying history of the diabetes. And in this case, you even have some neovascularization growing here, from proliferative diabetic retinopathy. The bottom slide here is HIV retinopathy. What looks familiar there to a vein occlusion are all the cotton wool spots. Right? Scattered about, kind of like a retinal vein occlusion, a central retinal vein occlusion. But what you’re missing again is the veins are nice and straight, they’re not dilated and tortuous, there’s no disc edema, and you’re missing the hemorrhages. Right? It’s more or less a pretty clean fundus, other than these cotton wool spots all over. That’s HIV retinopathy. The last photograph is hypertensive retinopathy. From, like, not just grade A or B, but more of a malignant hypertensive retinopathy. Where the blood pressure is extremely high. The patient has headaches. Both eyes are affected. Bilateral. And what you see in a hypertensive retinopathy that looks like a vein occlusion — you’ve got the cotton wool spots, and you have some hemorrhages. And the veins may be a little dilated there. A little tortuous or dilated. But what you’re missing is, again, all the blood. You can get disc edema in these cases. In this case, there’s not. But you’re missing the appearance of the blood. And this tends to be more of a posterior pole finding, in hypertensive retinopathy. Malignant hypertensive retinopathy. You’ll find more findings in the posterior pole. Whereas vein occlusions go all the way out to the periphery. The peripheral retina. So retinal vein occlusion. There’s really no treatment available to reverse retinal vein occlusions. Over the years, different treatments have ended up in the wastebasket of medical advancements. Things such as optic nerve fenestration. Also arteriovenous decompression surgery, vitrectomy techniques for both, where we take a blade and try and separate the artery from the vein, to restore blood flow. Or we’d incise the optic nerve with an MVR blade, to make a fenestration, trying to open up the optic nerve collar, to release that clot. These things were going on 10 to 20 years ago, and they come up every once and again, but today there’s been no trial to prove the benefit of those treatments. Iris or retinal neovascularization or macular edema may be managed with anti-VEGF injections or steroid injections. And again, managed but not cured. And macular edema and vein occlusion may be managed with laser intervention. Medical therapy. Anti-VEGF injections can induce regression of iris neovascularization and macular edema. The SCORE study demonstrated the benefit of triamcinolone or steroid for macular edema. But did not demonstrate the benefit in BRVO, or branch vein occlusions, versus focal laser. The reason the trial was done for CRVO versus sham was because there is no laser treatment thought to be effective for CRVO. That’s been studied. And for BRVO, there was — the standard of care is focal laser, so we compared the steroids to focal laser. So the mainstay of therapy currently is anti-VEGF therapy for macular edema with either CRVO or BRVO. Both ranibizumab in the BRAVO and CRUISE trials, which is Lucentis, and aflibercept, which is Eylea, in the GALILEO/COPERNICUS and the VIBRANT trials, showed both agents to be efficacious, compared to laser. And the results show significant gains in vision and decrease in retinal edema. Both drugs are recommended every month for six months, and then as needed. And bevacizumab has also been shown to be useful — that’s the off-label Avastin medication — similar to the other anti-VEGFs. There are also the steroid medications, steroid implants, such as Ozurdex, which is a slowly dissolving dexamethasone implant injected into the eye. It can also treat macular edema. Typically reserved as second line therapy for eyes with edema that may not be responding to the anti-VEGFs, or for some reason you’re trying to get a longer treatment interval. That medication can last longer, quite often, than an anti-VEGF injection. So in the SCORE2 study, Eylea versus Avastin there, aflibercept and bevacizumab, for macular edema, due to CRVO, it showed that bevacizumab or Avastin was non-inferior to Eylea, with respect to vision and OCT thickness, but after six months, good responders were divided into two different — the good responders were divided into a monthly and treat and extend regimen, using the originally assigned drugs. If you were not responding at six months, you could be switched, and I think that slide’s coming up. There was no difference between monthly and treat and extend for each drug. Of course, there were fewer injections for treat and extend protocol. If you’re a poor responder at six months, you could be switched. So the eyes that received Eylea, aflibercept, after a poor response on Avastin, did do very well. That’s an interesting part of this study. So if Avastin wasn’t working, you could switch to Eylea and get a nice response. The reverse wasn’t true. If you were already on Eylea, and switched to a steroid, there were very few eyes that even had a poor response to Eylea. So the take-home point there is that the eyes that don’t respond to Avastin still have an opportunity to respond to Eylea, if you have that medication available. And just to show where these agents come into play: These therapies, if there’s a vein occlusion with venous stasis, creating VEGF, your anti-VEGF agents can block that right there, preventing the macular edema. If these are not effective, the steroids can come in, block the VEGF some, and also treat inflammatory components of the macular edema. If you have capillary closure, there’s a higher VEGF production. And the anti-VEGF agents are very effective at blocking the VEGF and preventing the neovascularization. Looking at those two agents, steroids versus anti-VEGF, steroids have a bigger effect, being antiinflammatory, than anti-VEGF. Inflammation does increase in areas of ischemia. For vascular permeability, that’s stabilizing the vessel wall, so they’re not leaking, creating edema. Both steroids and anti-VEGF do have antivascular permeability effects. And blocking VEGF is more effective with anti-VEGFs than steroids. So you get a stronger neovascularization regression effect with anti-VEGF over steroids. Just a quick photograph of giving an intravitreal injection. That’s me with a few less gray hairs, a few years back. And this patient was kind enough to let us take the photograph here, of putting the needle through the pars plana. Lid speculum in the eye is my preference. Some lidocaine drops on the eye. A minute or two before this. And then some Betadine drops on the eye immediately before the injection, and typically in the inferotemporal quadrant down here. The nice thing about the lid speculum is that it keeps the lid margins away from the needle. Because you don’t want the edge of the needle to brush the lashes or the lid margin. You just want the needle tip going through where you put the Betadine on the eye, and that’s probably your biggest way to prevent infection, endophthalmitis. So complications associated with intravitreal injections include infection, bleeding, retinal detachment, cataract, and both infectious and sterile uveitis. However, those are relatively rare. Triamcinolone acetonide can cause steroid-induced glaucoma. So that’s one problem with the steroids, compared to the anti-VEGFs. A little photograph of the triamcinolone injected into the eye. It’s a powder. This was just injected a few minutes before the photograph. Topical anesthesia, just like the anti-VEGFs. Question about having the patient remain upright for 24 hours. It helps the steroid kind of sink to the bottom, into the vitreous space. It’s not essential, but maybe the vision clears a little more quickly, if they’re upright after the injection for a little bit. The Ozurdex injection. The device is there on the right. Kind of like the size of a pen or a pencil. The needle there at the tip. And the little plunger button, which injects the actual dexamethasone implant, which is seen here in the photograph at the bottom. And you can see this is a vein occlusion, by the way. Again, the dilated retinal vessels, the nerve fiber layer hemorrhages, and some disc edema. So looking at triamcinolone versus the Ozurdex for the steroid treatments and vein occlusions, a couple things jump off the page. One is: For the eye pressure effect, you get more of a rise in intraocular pressure, steroid-induced glaucoma, with the triamcinolone powder than you do with the Ozurdex. 16% to 25% versus 4%, in particular. And here’s a series of photographs looking at anti-VEGF treatment. Central retinal vein occlusion on the left, dilated, encouraged veins, scattered hemorrhages, macular edema, a little optic nerve swelling. The date is not on that slide. Here’s the OCT with intraretinal edema and fluid in the spaces. I would — just as an example, I’d say to get from this photograph to that photograph — you know, 6 to 12 months. So it’s not a one-month improvement. The hemorrhages are resolving. The veins are less dilated and tortuous. And the OCT is essentially normalized. So you can go from a 20/400 to 20/30-type outcome. Here’s a series of photographs of the same eye, with a resolving retinal vein occlusion. From April of 2006, 20/200, many, many injections, over almost two years, resolution of the hemorrhages, all the cotton wool spots, and a nice outcome. 20/25. So ischemic retinal vein occlusion can cause iris or retinal neovascularization, in addition to macular edema. So you need to observe that periodically, with undilated iris and angle exams, looking for this neovascularization. Especially after the onset of the vein occlusion. That’s your highest risk time. The first 90 days. That risk decreases through 6 months. Fluorescein angiogram can be useful to get you an idea of how serious that risk is. Another problem with vein occlusions is if you’re treating with anti-VEGFs, and a central vein occlusion, and you dry out the macular edema and the hemorrhages kind of clear, but the vision’s not good, you’re 20/400, so you kind of abandon treatment. Because it’s not gonna be worthwhile. The vision is so poor. Those eyes are the highest risk ones to still develop neovascularization. Once you withdraw the anti-VEGF, you’re opening yourself up to the VEGF production in the eye that you never treated with laser, with panretinal photocoagulation. So if you’re gonna treat with anti-VEGFs for six months or a year, and then withdraw in an ischemic eye, you want to watch that very closely. In fact, you can even make the argument just to place the panretinal photocoagulation because you know that your risk of rubeosis and neovascular glaucoma is very high. So here’s a photograph of iris neovascularization and a CRVO. So surgery — we’re talking about laser surgery here. And the central retinal vein occlusion study supported the use of panretinal photocoagulation for iris neovascularization. It did not support the use of grid photocoagulation for macular edema, so even though the macular edema would resolve with a grid laser in CRVO. We did not get a visual benefit. The branch vein occlusion supported the use of grid photocoagulation for edema, and also showed the benefit for panretinal photocoagulation for retinal neovascularization. So here’s some photos of resolving CRVOs. Same eye in the top. Nasal field, temporal field. Lots of hemorrhage. And over time, normalization of the retina and panretinal photocoagulation scars in the periphery. On the angiogram, even after normalization, there’s still some staining of the disc and the vessels. And you can see the capillary derangement there on the small vessels in the macula. Another image of an angiogram close up, after a CRVO is resolved somewhat. Typically this will occur over one to two years. You see the laser scars in the periphery. The veins are still dilated and tortuous. There are some collateral vessels on the disc that are helping drain out that blood now, and there’s a little bit of leaking on the angiogram near the macula in the late frames. So in summary, retinal vein occlusion has a highly variable course, often determined by the severity. Treatment with laser and intravitreal injection are allowing for better outcomes than ever before. And prompt diagnosis and commencing of treatment are essential for the best outcomes. And I believe the questions will be revisited here. So review of the questions from the beginning of the lecture. Complications of retinal vein occlusion include: Neovascularization of the retina, glaucoma, macular edema, vitreous hemorrhage, and all of the above. I think the audience had all of the above. Correct. Laser treatment is definitely effective for which of the following? Macular edema for BRVO, macular edema for CRVO, rubeosis, 1 and 3, or all of the above? And correct. Number four. It’s 1 and 3. Though you can get resolution of macular edema in CRVO, there’s really not a visual benefit. In a situation where you don’t have anti-VEGFs or you’re using anti-VEGFs and want to try to get off of the Avastin injections, the question of whether there’s a benefit to laser now is kind of reopened a little bit. Maybe it could be used to lessen the injection burden, or to get off the injections. I don’t think anyone has that answer. Atypical retinal vein occlusions should be investigated medically for an underlying cause? True or false? And that’s true. Looking for clotting disorders, underlying blood dyscrasias. And complications of intravitreal injections include cataract, sterile uveitis, infection uveitis or endophthalmitis, retinal detachment, and all of the above. And correct. All of the above is the correct choice there. So very good.
April 11, 2019