David Boyer, MD: Clinically significant macular edema (CSME) is a common consequence of retinal disease, including exudative age-related macular degeneration (AMD), diabetic retinopathy (DR), retinal vein occlusion (RVO), and uveitis. Vitrectomy is used in many of these eyes for which laser has been ineffective or is contraindicated. Following a vitrectomy procedure, the treating retina specialist may choose to introduce pharmacotherapy to improve visual outcomes in the setting of persistent macular edema. Although data exist regarding how eyes with ME respond to antivascular endothelial factor (anti-VEGF) agents and corticosteroid therapy, there are fewer clinical studies that have looked at how vitrectomized eyes will respond.
What common ground do these diseases have in regard to inflammation and VEGF?
Karl Csaky, MD, PhD: What we observe angiographically in terms of fluorescein leakage is similar in many cases. Although microaneurisms are specific to diabetes, it would be correct to say that VEGF and inflammation play a role in many of these retinal vascular diseases. The breakdown of the blood retinal barriers represents a commonality in the pathogenesis for AMD, DR, and RVO.
Seenu Hariprasad, MD: There are subtle differences between AMD, DR, and RVO, but with all of these diseases, the common ground, as Dr. Csaky said, is the disruption in the blood retinal barrier. VEGF and inflammatory factors likely cause the macular edema.
Szilárd Kiss, MD: Aiello et al1 Adamis et al2 and several other groups have shown in animal models and human vitreous samples that VEGF upregulation and the inflammatory cascade play a role in all of these disease states. The contribution of inflammation, however, differs among differing diseases. In AMD, VEGF appears to play a more prominent role. In RVO and DR, the inflammatory component has a significant role in both the macular edema and the drive for neovascularization.
Dr. Boyer: Corticosteroids have been widely used with little scientific evidence to treat the inflammatory component of RVO and DR. Recently, we have had the benefit of some large scale clinical trials on these subjects. The Diabetic Retinopathy Clinical Research Network (DRCR.net) study on laser vs steroids for DR3 the SCORE studies for branch retinal vein occlusion (BRVO)4 and central retinal vein occlusion (CRVO),5 and the intravitreal dexamethasone implant (Ozurdex, Allergan, Inc.) for RVO study6,7 all looked at steroids for the management of secondary ME.
THE BASIC SCIENCE OF STEROIDS
Dr. Boyer: Steroids themselves are not interchangeable.
Many clinicians, however, use steroids without differentiating
methods of action or side effect profiles. Dr.
Cidlowski, you have performed volumes of research on
the action of corticosteroids and their effects. Can you
provide an overview of your research as it pertains to
the eye?
John Cidlowski, PhD: There have been many developments in the basic science of corticosteroids in the past several years. Based on the data that we have acquired, we now know that not all glucocorticoids are equal in their mechanism of action and their effects. If one looks at the published profiles of steroid-induced genes, for example, there are significant differences in the ability of steroids to regulate gene expression. How this translates to their use in humans remains in question, however, and further research to this end is required.
We do have an understanding of how some of the mechanisms lead to this discrimination. Dexamethasone and triamcinolone acetonide (Kenalog, Bristol-Myers Squibb), both of which are synthetic steroids, are different from one another and also different from the natural glucocorticoid. The goal is to create a steroid that will have fewer side effects and less toxicity than our currently available agents, and as we gain a better understanding of these, the goal becomes closer.
The glucocorticoid receptor (GR) is the protein that mediates the actions of all pharmaceuticals. Genetic evidence exists to show that without the GR, no response other than a drug response occurs. For example, if cells without glucocorticoid are cultured and treated with a steroid, approximately only 50 genes will be upregulated. With the receptor present, however, as many as 6,000 genes will be upregulated within 6 hours of adding the steroid. Within 12 or 24 hours, even more genes are upregulated.
GRs are ligand-dependent transcription factors that are not only necessary to achieve steroid effects in the eye, but that is also necessary for survival of man. There is not an area of medicine that is not influenced by GR— they regulate metabolism and are most well known for their role in inflammation; without them there would be no steroid response. GR is at the very top of the inflammation pathway, working in a multidirectional manner to quiet inflammation (Figure 1).8
In contrast, anti-VEGF antibodies work at the bottom of the pathway by halting the biosynthesis of the product, including proinflammatory molecules. With natural steroids, there are few side effects. In fact, we are all exposed to high concentrations of steroids on a daily basis. The side effect problem occurs with synthetic steroids.
Steroids have a diffusion mechanism that allows them to access to every cell in the body. This receptor protein is in the cytoplasm of cells and moves into the nucleus, upregulating gene expression by binding to DNA or other transcription factors that regulate functions. There are new data showing that GR is not a single protein but a family of transcription factors. Molecular biologists today know that there are 25,000 genes, which make 100,000 proteins in the body's cells and mechanisms for reusing or using one messenger RNA to make a large number of different proteins. GR acts not as one protein but as a family of proteins.
There are eight GR subtypes, and all are produced from a single messenger RNA: GRα-A, GRα-B, GRα-C1, GRα-C2, GRα-C3, GRα-D1, GRα-D2, and GRα- D3. What is interesting is that the prevalent receptor in the eye is not at the top of the model as it is in other locations of the body, but rather the C- and D-forms of the receptor.9
Dr. Csaky: So in the retina within an endothelial cell, a Mueller cell, a photoreceptor cell, and a bipolar cell, it is conceivable and probable that each one of those cells has a different repertoire of these receptors?
Dr. Cidlowski: Yes. And the repertoire dictates the activity of the receptor.
Dr. Csaky: The assumption is that as the repertoire changes, the gene expression profile changes.
Dr. Kiss: Are the steroid side effect profiles different depending on the levels of activity on the glucocorticoid receptor?
Dr. Cidlowski: Without question, yes.
Dr. Hariprasad: Does the expression within cells or receptors change based on the term of exposure to a steroid?
Dr. Cidlowski: Yes. Steroid response can change with chronic steroid use. Often there is a desensitization mechanism that exists in the periphery, but not particularly often in the eye, which surprises me.
Dr. Boyer: How does concentration, for example 1 mg triamcinolone acetonide vs 4 mg triamcinolone acetonide, affect response?
Dr. Cidlowski: There is a dose response, but steroidsclassically have biphasic dose responses—as the dose isincreased the response initially increases, but if the doseis increased even more, the response goes down.
There are different promoters in different cells. Thereare five promoters for the GR gene. I do not think that itis currently understood which promoters are used in cellswithin the eye. The promoters drive the expression levelof the receptor to different levels. In a lymphocyte, theexpression level is 2,000 to 3,000 receptors per cell. A livercell has 25,000 receptors per cell. Every cell is completelydifferent.
Dr. Csaky: Does the number of receptors per cellaffect susceptibility to steroids? One of the surprisingthings about the intravitreal dexamethasone implant isthat in humans it appears to have a 4-month clinicalduration, but in animals and in-vitro release it lasts outto 6 months. I was taught that steroids were similar tohormones in that they act in femtogram amountsbecause of the high affinities, which makes it all themore surprising to me that we are losing the clinicalefficacy at 4 months.
Dr. Cidlowski: There is a mechanism in some cells—Ido not think that it has been studied in the eye—wherethe steroid actually turns off the expression of thereceptor. Long-term steroids can lead to desensitizationand resistance and although there are a couple ofmechanisms for resistance, response is generally relatedto concentration. When we perform microarray analysis,we put x amount of receptors (eg, two or three xreceptors) into a cell and create an artificial environment.The more receptors you put into a cell the moreresponse you get. As I had stated earlier, the C and Disoforms are the most prevalent in the trabecular meshworkcells. The D isoforms are distinctly different fromthe others and do not induce apoptosis.
GRβ is an alternatively spliced isoform of hGR primarytranscript. GRβ is tissue- and cell-specific in its expressionof mRNA and proteins and does not bind agonists nor isit transcriptionally active on GRE-containing receptors. Itis located in the nucleus of the cell, rather than in thecytoplasm. What is surprising is that GRβ regulates genesand also blocks the ability of the GR to function andbind with steroids. GRβ has been studied in many diseasestates and is associated with glucocorticoid-resistantasthma, fatal asthma, leukemia, ulcerative colitis, nasalpolyposis, rheumatoid arthritis, as well as in eye diseases.
What has been discovered is that over-expression ofGRβ inhibits glucocorticoid induction of glaucoma-associatedgenes such as myocilin, fibronectin, and alters trabecularmeshwork cell phagocytosis. A deficiency of GRβis associated with glaucoma in normal individuals.10,11There is also a polymorphism in this gene that has beenstudied extensively12 and in almost 8,000 patients in onestudy.13
It is estimated that approximately 1.3% of the populationhas this polymorphism, which is responsible for athreefold increase in risk for rheumatoid arthritis and atwofold increase in risk for coronary artery disease, thelatter being a more significant risk factor than elevatedcholesterol.14
To test the effect of dexamethasone on the mobilityof YFP (yellow fluorescent protein)-hGR, we performeda FRAP (fluorescence recovery after photobleaching)analysis. We put YFP on a control GRα in the absenceof hormone, most of the YFP is in the cytoplasm. Whenyou apply dexamethasone, however, the YFP moves tothe nucleus. We then photobleached the YFP andobserved the fluorescence recovery in real time. Wefound that steroids have a significant influence overhow fast the receptor moves around in a cell (Figure 2).Further, the mobility of hGRα is highly ligand-dependentand it varies depending on what steroid is applied— between synthetic and natural steroids, hGRα mobilityis higher in the presence of natural steroids (Figure 3).When comparing the synthetic steroids, the mobilityof hGRα is similar between triamcinolone and dexamethasone,but lower in the presence of triamcinoloneacetonide.
Dr. Boyer: Where does fluocinolone acetonide fit in?
Dr. Cidlowski: From what I know of its structure, Iwould place it in the range of triamcinolone acetonide.Fluoride is used in synthetic steroids to stabilize them; ifone were to remove the fluoride, the mobility would goback up to the level of a natural glucocorticoid.
Dr. Csaky: Are there relative potency differencesbetween steroids?
Dr. Cidlowski: Yes. The differences in apoptosisoccur primarily at low concentrations of steroid and it isusually a reflection of affinity of these ligands forglucocorticoid.15
Dr. Csaky: On lymphocytic apoptosis scaling, whichsteroid is the most potent?
Dr. Cidlowski: In our hands, dexamethasone is themost potent—three- to fivefold more potent than triamcinoloneacetonide.
Dr. Boyer: Which is more toxic: triamcinolone acetonideor dexamethasone?
Dr. Cidlowski: It is hard to say which is more toxicbecause these steroids have a different clearance andthere is little information on toxicology in the literature.Additionally, it would depend on the model system,species, cell type, and tissue being studied. We do knowthat all of the corticosteroids have an affinity for theinflammatory pathways and that all inhibit the biosynthesisof VEGF, particularly in the eye. Triamcinolone acetonideand dexamethasone, however, are unique;although similar in appearance, they actually cause differentconformations in the nuclear receptor that mediatestheir actions.
Dr. Kiss: The side effects seen in the triamcinolone acetonidevs laser4,5 and the intravitreal dexamethasoneimplant trials6,7 were different, and some occurred asearly as 6 months. The longer-term data from 2 or moreyears will provide much more information on the specificside effects of the two steroids.
Dr. Boyer: For sustained-release corticosteroiddevices, such as the Ozurdex, the Iluvien (fluocinoloneacetonide, Alimera Sciences, Inc.) and Retisert (fluocinoloneacetonide, Bausch + Lomb), does the location ofthe device in the eye change the profile of which genesare upregulated to increase intraocular pressure (IOP)and cause glaucoma? For example, the incidence ofglaucoma is far different with Iluvien, which is placedmore posterior than the Retisert, though both containingsimilar amounts of the same steroid.
Dr. Cidlowski: Basic science tells us that almost everycell in the body will respond differently to steroids, so bychanging the placement of the pellet, different cells andtheir functions are influenced.
Dr. Kiss: Many of the side effects that occur withsteroids seem to happen in the front of the eye, so itmakes sense that if the steroid is kept posterior, some ofthe more dramatic side effects may be avoided.
Dr. Csaky: We have performed research on drug distributionthat clearly showed that when drug elution occursat the pars plana, the flow moves in an anterior direction.When placed more posteriorly, the convection currents will “catch” the drug, taking it further back into the eye.16
Dr. Boyer: Is this altered in a vitrectomized eye?
Dr. Csaky: How convection currents react dependslargely on the eluding profile of the drug. Diffusion characteristicsare altered in the vitrectomized eye,17 whichmay allow the drug to penetrate more posteriorly.However, it is important to place the drug in a more posteriorlocation if the goal is to have the drug reach theback of the eye.
Dr. Boyer: In animal models, it appears that drug kineticschange when exposed to a fully formed vitreous ascompared to a liquefied vitreous or a reduced vitreous,as with a vitrectomy.18-21 The drug clearance speeds upwith less vitreous and the long-lasting effects if drug in afully formed vitreous are lost.
Dr. Csaky: This is correct. The vitreous forms a diffusioncoefficient that allows the drug to remain. When vitreousis removed, the drug will diffuse freely through theaqueous environment, which has a more robust clearancemechanism. This is particularly true in the case of anintravitreal injection in a vitrectomized eye. The turnoverof the volume of drug is rapid because injections aregiven mostly anteriorly. Placement and vitrectomy statusare critical aspects when considering the duration andclearance of drugs.
TREATMENT APPROACHES FOR DME AND RVO
Dr. Boyer: What is your treatment approach for apatient with DME?
Dr. Hariprasad: If the patient has focal disease and discreteleaking microaneurisms on fluorescein angiography(FA), I will use focal laser. For a patient with diffuse DME,however, I have found that the outcomes are not as goodusing focal grid laser as a first-line treatment and myapproach becomes more complicated. I will typicallyorder an optical coherence tomography (OCT) scanbecause the pattern seen on the OCT helps me make treatment decisions. For example, I may consider vitrectomywith membrane peeling earlier if a DME patient hasvitreomacular traction or distinct taut or diffusemechanical traction. If a patient has subfoveal seriousretinal detachment in addition to intraretinal leakage,he or she may do well with steroids. Overall, for diffuseDME, I am a fan of combination therapy (ie, steroid orAnti-VEGF therapy at baseline and focal laser treatment2 to 3 weeks later). Of note, more recently I have beenusing widefield FA to assess peripheral nonperfusion inpatients with recalcitrant DME. I believe that the VEGFdrive in these more difficult-to-treat cases may be comingfrom the ischemic periphery. With new generationwidefield imaging, it is amazing how many people wepick up that have peripheral nonperfusion that wewould not have seen on clinical exam alone.
Dr. Boyer: Do you reserve vitrectomy for patients whohave vitreomacular traction or an epiretinal membrane?
Dr. Hariprasad: I reserve vitrectomy with membranepeeling for two clinical scenarios: 1) DME recalcitrant toall in office treatments and 2) patients with distinctmechanical traction. Obviously, before considering vitrectomy,we try in-office treatments and also make sure thatmacular ischemia is not responsible for the vision loss.
Dr. Kiss: My approach is similar. I will counsel thepatient to see if he or she may be having difficulty controllingtheir diabetes, hypertension, and/or lipids and ifso, I address this first. I use widefield angiography(Optomap fa, Optos, Dunfermline, UK) for all of mypatients with DME, because areas of nonperfusion orischemia in the periphery may be contributing to the MEat the posterior pole. Because there is emerging evidencethat there are different types of ME, I prepare toapproach each case differently. For a patient with diffuseME, I may laser the peripheral ischemic area, even if thereis no obvious neovascularization. Depending on the lensstatus and the presence of glaucoma, I will use either ananti-VEGF agent or a steroid to flatten the area prior tousing laser.
For patients at the other end of the extreme—thosewith more focal CSME and a periphery that does notappear to be ischemic on widefield angiography-my firstline of therapy remains focal laser.
I am more likely to perform a vitrectomy for patientswith either neovascularization or DME that is resistant tolaser or intravitreal injections.
Dr. Csaky: Every time I think that I have figured outwhich patients will respond to steroids and who will respond to anti-VEGF agents, it turns out that I have itwrong. When we talk about inflammation cascades andthe role of VEGF in CSME, particularly in DR and RVO,we are finding that these processes are far more complexin terms of how a patient responds to a steroid, anti-VEGF agent, or a combination.
It is also important to consider that patients mayrespond differently depending on the steroid agent beingused. Individualization of therapy is necessary.
Dr. Kiss: I agree with Dr. Csaky about the need toindividualize therapy. If a patient has distinct focal leakage,I think all of us would say we use focal laser, whichworks well and resolves the leakage in two to three sessionsin many cases. With diffuse DME, however, theretruly is an unmet need. We all have different algorithmsand that is for two reasons: first, we lack data and second,there is an enormous amount of variability withthis disease in terms of OCT and angiography findings,mechanical traction, and systemic control of bloodglucose and blood pressure. When peripheral nonperfusionis thrown into the mix it becomes even moredifficult.
Dr. Cidlowski: Although there is still a great dealthat we do not understand regarding steroids, it isimportant to keep in mind that when a patient is notresponding to a steroid, it may reflect the polymorphismsin the glucocorticoid receptor. There is a singlestudy published on 56 eyes that showed no effects ofpolymorphisms on steroid response, but this is unconvincingbecause of the small sample size.22 In fact, inother diseases, profound effects in steroid responseof polymorphisms have been shown, as I discussedearlier.13,14
STEROIDS IN THE SETTING OF VITRECTROMY
Dr. Boyer: Dr. Hariprasad, do you pretreat yourpatients with anything prior to performing a vitrectomyfor DME?
Dr. Hariprasad: I do not pretreat, but at the end of acase I inject these patients with 0.10 cc of 4.0 mg triamcinolone triamcinoloneacetonide or preservative-free triamcinoloneacetonide (Triesence, Alcon Laboratories, Inc.) into thevitreous cavity to stabilize the blood retinal barrier andto help with CSME. I am impressed with how quietthese eyes are the first postoperative day comparedwith eyes not injected with steroid immediately after avitrectomy. The disadvantage is that by 1 week postoperative,the steroid effect is very minimal. Clearly there isa need for a steroid with longer durability in both vitrectomizedand nonvitrectomized eyes.
Dr. Kiss: Whether I pretreat depends on why thepatient was taken to the OR. If there is neovascularizationof the retina, I will inject an anti-VEGF agent such asbevacizumab (Avastin, Genentech). If it is more of a diffusemacular edema, I tend to inject triamcinolone acetonide.I do not base this on solid clinical evidence, ratheron my own experience and the small amount of scientificevidence suggesting that these two may have slightly differingunderlying driving mechanisms. I agree that theduration of action of a drug in the vitrectomized eye isshorter, but I continue to do this because it gives me anextra tool to settle the patient's disease, even if for just alittle while.
Dr. Csaky: I try to avoid vitrectomy in patients withCSME, because once the eye has been vitrectomized, aclinician has one good shot at getting the macula flat. Ifthe patient comes back in 6 months with CSME, anymore injections will have little durability and impact.Because of this, I have been conservative regarding vitrectomyfor diabetic eyes.
Dr. Boyer: Often, patients do not require long-termsteroid therapy; rather, they need steroids for 3 to 4months to get them through a period of increasedinflammation. We recently reported data from thephase 2 dexamethasone intravitreal implant study. Inthis 26-week, open-label study of 56 patients who hadundergone vitrectomy at least 3 months prior, a statisticallysignificant improvement was found in both centralretinal thickness and visual acuity throughout thestudy duration.23
These results are very encouraging. The patientsin our study required intravitreal injections of triamcinoloneacetonide or bevacizumab on a weekly orbiweekly basis to get a response and some hadreceived multiple treatments with more than oneagent. It was only after receiving the intravitrealdexamethasone implant that most of these patientswho had undergone vitrectomy experienced a lastingresponse.
Dr. Csaky: Given these data on the intravitrealdexamethasone implant, which is release driven ratherthan clearance driven, I would be more likely to performvitrectomy for my patients with diabetes.Although as I stated previously, I am hesitant to performvitrectomy for patients with diabetes because ofthe problematic CSME that often occurs postoperatively,vitrectomy has distinct advantages for cases thatrequire anatomic correction. Having a medical therapythat has durability in a vitrectomized eye is an enormousbenefit.
COMBINATION THERAPY
Dr. Boyer: Barry Kuppermann, MD, presented interestingdata on the effects of the intravitreal dexamethasoneimplant as an adjunctive therapy with ranibizumab(Lucentis, Genentech) that demonstrated anenhanced effect using the two agents in combination.24
With the plethora of new data regarding anti-VEGF, anti-VEGF in combination with laser25 and VEGF trap for DME,26one might come to the conclusion that these impressiveresults are the best that can be achieved. Dr. Kupperman'sdata, however, shows that this may not be the case and thatyes, we can do better by combining a sustained releasesteroid to address the inflammation as well as the VEGF.
Are you treating vein occlusions differently now thatwe have all this new information from the five major clinicaltrials that has been released?
Dr. Csaky: The bar for RVO has been raised. I am of thegeneration that considered a visual acuity outcome of20/200 to be acceptable in some cases, but we now havetherapies that can drive patients back to 20/30 or 20/40vision and we are more aggressive in our approach withboth BRVO and CRVO. For most cases of CRVO, I willtreat with an anti-VEGF agent, but if I do not see aresponse, I will go quickly to some form of steroid tocover the inflammatory response. If I can flatten out theedema in BRVO using anti-VEGF or a combination ofanti-VEGF and steroid, I am still a firm believer in focallaser to keep the eye dry.
Dr. Boyer: If you start with an anti-VEGF agent, howmany injections will you give the patient, and if you continuethe anti-VEGF, when do you introduce the steroidin combination?
Dr. Csaky: My approach will depend on the baselinevisual acuity and the anatomical details. If the patienthas very poor vision, such as 20/200 with ME, I willstart with bevacizumab. If the edema responds thanI will continue injections. If, however, the patient's vision is 20/70 and I do not see any change after oneor two injections of bevacizumab, I will use either alow-dose steroid or the intravitreal dexamethasoneimplant.
Dr. Hariprasad: In RVO more than any disease, combinationtherapy may be the key to success because ofthe complexity of vein occlusions. There is tremendousvariability in how patients present, often making treatmentdifficult.
Currently, one of my preferred combinations is theintravitreal dexamethasone implant combined with laserand occasionally anti-VEGF injection. The side effect profileof dexamethasone is appealing, in contrast with theother steroids that we typically use in the eye, and thesustained-release delivery system increases its durability.I then supplement the implant effect with laser or anti-VEGF injection 2 to 3 weeks after the initial dexamethasoneimplant injection.
Dr. Kiss: The data from the intravitreal dexamethasonetrial, as well as BRAVO and CRUISE,6,27,28 haveinfluenced me to treat BRVO earlier for two reasons.First, I find it objectionable to force all my patients towait 3 months for possible visual improvement. Second,we have data from the intravitreal dexamethasoneimplant trial that demonstrated that some patientswho were treated earlier gained more vision than otherswho were under observation.6
I agree with Dr. Csaky's approach; I flatten the retinawith an injection and then apply laser if needed. I had apatient with BRVO who was receiving bevacizumab injectionsapproximately every 5 to 6 weeks. She would makeher appointments whenever her vision started to regress.Not surprisingly, after the thirteenth injection she saidshe was tired of the repeated visits. I ended up applyinglaser after the next injection had flattened the retina andthis worked well.
Dr. Boyer: The burden of multiple injections wears onboth patients and doctors, but there are patients whotolerate multiple injections well, so it is important tolook at each case individually.
Dr. Hariprasad brought up an important point. Twentypercent of patients with CRVO and BRVO who receivedone injection with the intravitreal dexamethasoneimplant had relatively flat maculas and 20/20 vision thatlasted 1 year. Additionally, there were patients in BRAVOand CRUISE who had dramatic, lasting improvementwith one injection, demonstrating that we can achieve“grand slams” with our newly available treatments.Combination therapy, however, should be considered for many of our patients because the success seems to varypatient to patient.
Dr. Csaky: There is clearly a synergy with anti-VEGFagents and steroids. The concept of flattening the retinawith an anti-VEGF injection and then keeping it flat forlong periods of time with the sustained-release intravitrealdexamethasone is a far better strategy than usinganti-VEGF alone and treating on an as-needed basis,which results in a yo-yo effect for the patient. Waitingfor disease to recur before we treat is not good for ourpatients.
Dr. Boyer: I agree. As we saw in the CRUISE and BRAVOdata, continuing with constant injections of ranibizumabprovides a very good visual result. All of us, however, areinterested in reducing the number of treatments that wemust administer to our patients. Additionally, we areworking toward refining our follow-up. Waiting for diseaseto recur will ultimately result in additional damage.Having the mechanism of sustained release will offer anadvantage for both patients and physicians.
Dr. Csaky: I think it is also important to point outthese differences between the trials, particularly betweenthe intravitreal dexamethasone implant trial and theBRAVO and CRUISE trials. In BRAVO and CRUISE,the average duration of vein occlusions was less than3 months and in the dexamethasone trial, the durationon average was longer. If, however, you look at the subsetof patients with shorter duration vein occlusions, thesepatients' results were remarkably similar to the patientsin BRAVO and CRUISE. Although the numbers weresmaller for this subset in the dexamethasone trial, it suggeststhat if patients are treated earlier with the steroidimplant, the results will be better.
I think that a key take-home message is not to wait totreat. Additionally, it must be recognized that these trialsare not completely comparable, because the populationsare not the same.
Dr. Boyer: The intravitreal dexamethasone implantdata is compelling in its own right, but when viewedcarefully, one notes that patients received the secondinjection at the end of 3 to 4 months, so we are unawareof how good that data could be if reinjection occurredearlier. In clinical practice, when I start to see the patientregain fluid after they have responded well to the intravitrealdexamethasone implant, I am going to reinject. If Iintervene when my patients begin to reaccumulate fluidon OCT, I think I have a better chance at maintaining oreven improving their vision.
Dr. Kiss: If you look at the “180-day follow-up” for thedexamethasone implant it actually ranged from 136 to210 days. We know from the pharmacokinetic studiesthat the drug remains in the vitreous cavity for approximately180 days. Comparing a patient who has had theimplant for 136 days vs 210 days seems unfair.
Dr. Boyer: Clinically, I am finding that the maximumduration I am getting from the dexamethasone implantis 4 months. I have many patients who I am beginning toobserve after injection and have found that, although apatient's vision might be good, there is often a disconnectwith the anatomical findings on OCT. I do not necessarilywait for vision to decrease because if the OCTshows fluid building up, the vision, without a doubt, willsubsequently decrease.
Dr. Csaky: One of the more exciting aspects of the12-month dexamethasone data was that there was noincrease in IOP when patients were reinjected with theimplant. SCORE, on the other hand, demonstrated thatIOP increased more with each reinjection of triamcinoloneacetonide.
The other consideration involves the hypothesis thatanti-VEGF agents may prevent the disease from maturation.David M. Brown, MD, has collected interestingdata from his ongoing 3-year study, RAVE (RubeosisAnti-VEGF Trial for Preproliferative CRVO). In RAVE,patients were treated with an anterior chamber tap andthen monthly ranibizumab injections for 9 months andcould be reinjected after 3 months observation on anas-needed basis for 3 years. The visual acuity andcentral subfield thickness improved in many/most ofthe patients in the study over 9 months. When theinjections were discontinued and the monitoringphase began, the edema in approximately one-thirdof patients came roaring back, which is similar to ourexperience in AMD. Additionally, more than 50% ofpatients developed neovascularization over the 3-yearperiod and one-third of these had posterior neovascularization(Unpublished data).
Anti-VEGF agents do not change the gene expressionprofile of the retina, rather, they gently “sop up” the VEGF.If we apply two or three steroid injections or chronic suppressionover 1 year, the VEGF may burn itself out.
Dr. Cidlowski: I agree. Anti-VEGF agents treat thesymptom, but not the disease in RVO. A combination ofanti-VEGF and steroid is extremely powerful, becauseVEGF is one of the genes repressed by glucocorticoids.Glucocorticoids, however, cannot shut off the expressionof any gene completely. There will always be some amount of leakage or some presence of VEGF, so a combinationis most likely the best solution.
Dr. Boyer: Do you still use triamcinolone acetonide invitrectomy?
Dr. Hariprasad: I have performed vitrectomies forcases of failing laser, triamcinolone acetonide, or bevacizumab.The next line of treatment is vitrectomy with orwithout internal limited membrane peeling and peripherallaser. At the end of the case, I would inject triamcinoloneacetonide. The benefit to triamcinolone acetonide,however, is short-lived in these eyes because theclearance is quick. There truly has been an unmet need invitrectomized eyes for a sustained-release steroid thatcan be injected at the end of surgery.
I have injected the dexamethasone implant into vitrectomizedeyes where the initial indication was macularedema secondary to RVO. In these difficult to treateyes, vitrectomy was unsuccessful in resolving the macularedema (keep in mind that these eyes failed all inofficetreatments prior to vitrectomy). I have had somesuccesses with the dexamethasone implant in theseclinical scenarios. Often, the macula is dry after insertingthe dexamethasone implant; however, in patientswho have long standing recalcitrant edema, the visiongain is minimal due to photoreceptor loss. The lesson isto intervene sooner before photoreceptor loss limitsvision gains.
Dr. Csaky: We must be cognizant of what the patientpays out of pocket. The socioeconomic and insurancestatus of my patients drives part of my decision makingprocess. When using triamcinolone acetonide, I almostnever use 4 mg, rather I use 1 mg to reduce the sideeffects. However, when a patient has poor insurancecoverage, I do not use triamcinolone acetonide becauseof cost.
Dr. Kiss: The incidence of cataract and IOP increase issignificant with triamcinolone acetonide, as shown in theSCORE trial and the DRCR.net DME trial. The side effectprofile of sustained-release dexamethasone, however,appears to be significantly better than that of triamcinoloneacetonide. In vein occlusions for which I may havepreviously injected triamcinolone acetonide to try andflatten out the retina, I now prefer to use the dexamethasoneimplant instead.
SUMMARY
Dr. Boyer: What are some important take-home messagesthat the faculty can share with our readers?
Dr. Cidlowski: There is a lack of basic understandingof the biochemistry of the GR in the eye, in terms oftypes of cells and their functions and more researchis required. When patients do not respond to a medication,this lack of response may be related to polymorphismsin the protein that are necessary to mediatethe actions of the glucocorticoids.
All steroids are not created equally. It is important todispel the notion that different classes of steroids willhave similar effects—they bind to the receptor in a differentmanner and cause the receptor to regulate differentsets of genes.
Dr. Kiss: Results from several trials suggests that earliertreatment of macular edema in RVO is more likely toresult in better visual outcome. In DME, the concept ofcombination therapy is emerging, but we are learningthrough our clinical experience and recent trials thatusing laser, anti-VEGF, and steroids in combination mayprovide for better visual outcomes over the long term.
For vitrectomized eyes, sustained-release dexamethasoneappears to be as durable as it is in nonvitrectomizedeye, in contrast to triamcinolone acetonide andanti-VEGF agents, which appear to be clear much morequickly.
Dr. Hariprasad: As retina specialists, we need tokeep an open mind to implementing new therapies sowe can maximize outcomes in our patients and minimizeside effects. It will be our job to figure out howto put these new monotherapies together to make themost of their synergies. Different stages of disease mayrequire different therapies and regimens. With greaterknowledge and experience we will be able to individualizetreatment, which will hopefully yield the bestoutcomes.
Dr. Csaky: The ultimate goal in RVO and DME is tokeep the retina thin without rebound effects and toreach a point where we can maintain a flat retina withoutcontinuous retreatments.
My patients with diabetes typically go through a phasethat I have termed the “2-year roller coaster ride.” Duringthis period of time, patients' edema and visual acuity isextremely unstable. If I can get these patients throughthis phase and stabilize them with lasting treatment, theyoften reach a point where the edema burns itself out—this is not always driven by laser treatments.
Dr. Boyer: Even with the new treatments that wehave in the form of anti-VEGF agents and steroids, wecontinue to only be able to postpone these diseases—we are not modifying the disease. For RVO, we may notneed longer duration (eg, 18 months) sustained deliverybecause the life of vein occlusions is not that long, butfor a chronic condition like diabetes, this could beadvantageous.
The bottom line is that we should not attempt tomake interstudy comparisons. The study data must beviewed as different and specific to the patient populationsthat were being treated. If studies are not designedto be head-to-head, any conclusions drawn comparingdata from different studies can be very wrong. I do agreethat combination therapies may provide better alternativesand overall outcomes for our patients.
_1773249222.png?auto=compress,format&w=75)
