CLINICAL TRIALS IN DME
David S. Boyer, MD: In the last 3 to 4 years, there have been several clinical trials that have demonstrated efficacy of new paradigms for the treatment of diabetic macular edema (DME). I'd like to review some of these studies and discuss what has been learned thus far.
Dante J. Pieramici, MD: The Diabetic Retinopathy Clinical Research Network (DRCR.net) trial was one of the first phase 3 clinical trials to demonstrate the positive role of antivascular endothelial growth factor (anti-VEGF) for DME. Protocol I compared ranibizumab (Lucentis, Genentech) either with prompt or deferred laser to laser alone and laser plus steroids. The results of this trial showed for the first time that anti-VEGF could improve vision to a significant level, greater than that with the standard of care, which is laser photocoagulation.1 The results were striking, and they have changed how many of us are managing our patients with DME. We can now not only stabilize vision, which is what we could achieve with laser photocoagulation, but we can also expect vision improvement for many of these patients.
Pravin U. Dugel, MD: Going back to where much of this began, we have to look at Protocol B from the DRCR.net, which compared laser photocoagulation with corticosteroids.2 This trial was performed because clinicians were beginning to use corticosteroid injections (Kenalog, Bristol-Myers Squibb) more frequently to treat DME, and there was some important information that was gained. The results of Protocol B certainly taught me that laser still has a place in DME. It is unfortunate, however, that this publication led to a misunderstanding as to the role of steroids in this disease. In Protocol B there was a comparison made with steroids in patients who had visual acuity of 20/40. I would submit that most of us do not use steroids as first-line therapy for patients who have vision close to 20/40 or better. Rather, we reserve the use of steroids for patients who have worse vision. Indeed, Table 5 from Protocol B (Figure 1) shows the subdivision of visual acuity data: patients who had vision of 20/200 or worse had better outcomes with steroids than with laser photocoagulation. This illustrates that although both laser and steroids have a place in treating DME, the best choice depends on the degree to which the disease has progressed.
Szilárd Kiss, MD: It is encouraging that we are moving beyond laser photocoagulation as the only treatment option for DME. As we learn more about the underlying pathology of DME, we are gaining more data on the role of antivascular endothelial growth factor (anti-VEGF), but also the role of inflammation, which is where steroids come into play. In the DRCR.net Protocol I study, the subset of pseudophakic patients did very well with steroid treatment in comparison to anti-VEGF treatment.
As Dr. Dugel mentioned, the disease stage is a critical factor in how a patient responds to a particular treatment— there may be a significant difference in response between a patient with long-term diabetic retinopathy and one who is newly diagnosed. And then, we have to consider patients who have been receiving a combination of treatments. For example, there were patients in the FAME clinical trial for the fluocinolone acetonide insert (Iluvien, Alimera Sciences) who had received laser and then the insert; however, during the 2-3 year follow- up, they were also receiving whatever the “standard of care” was, and so the comparison arm was not to just laser alone, but it was also to anti-VEGF injections and other treatments that may have been given. Also, there were clearly patients who had macular edema for more than 3 years, and there was a significant added benefit to using the steroid in addition to laser and in addition to what we can presume was anti-VEGF during that time period.3
Dante Pieramici, MD: Going back to Protocol I from the DRCR.net, I was amazed that patients who were pseudophakic seemed to do equally well with steroids plus laser as they did with ranibizumab plus laser.
Dr. Dugel: We are learning that DME is a continually evolving disease. As time goes on, we continue to gain more insight. We will increasingly apply all available treatment modalities either alone or in combination depending on the stage of the disease.
DME CLASSIFICATION
Dr. Boyer: The old classification with which we are familiar is clinically significant macular edema (CSME).4,5 With the new treatments (steroids, anti-VEGF) does this classification still stand? Additionally, how do you utilize optical coherence tomography (OCT)? This is an important point, as certain cases of CSME require slit lamp and a fine microscopic diagnosis. How are the new modalities of imaging, such as spectral-domain OCT (SD-OCT) changing your treatment?
Dr. Dugel: We are still learning how to classify macular edema. Currently the first step in choosing my treatment algorithm is determining whether the DME has a mechanical etiology, and if it does, I tend to use a surgical approach. It not, I will treat with a medical approach. If I see that the disease is purely focal, laser photocoagulation still has an important role; however, I usually see patients when the disease has progressed to diffuse or ischemic DME and has become more complex. This is where combination treatments are most effective.
How should we follow treatment efficacy? OCT is a good indicator of VEGF, but as we know, DME is more than a VEGF-driven disease. In some cases, there is a significant disconnect between what we see on OCT and what we see in functional testing.6-8 The FAME studies have recently shown this, as have other previous studies evaluating therapy for DME. The bottom line is that a thinner retina is not always better because neuronal loss is a factor in DME; OCT, although helpful, is not the sole indicator of progress; and, finally, DME is more than a VEGF driven disease—it is multifactorial.
Dr. Kiss: When patients come in with newly diagnosed diabetes or for one of their early eye exams, I take both OCT and fluorescein angiography (FA) into consideration. If I see even just 1 or 2 cysts on OCT or leakage or peripheral nonperfusion on FA, I explain to the patient that the cascade of DME has begun if it is not yet CSME. I then will talk to the patient about systemic control and how we will treat the DME.
As our diagnostic technology improves, it allows us to treat patients earlier. If there is obvious vitreomacular traction or another component that may be amenable to surgical intervention, then we will go down that path. In general, however, I am loosening my criteria in terms of how much edema is present to initiate treatment. In terms of what my first-line of treatment is, if there is a circinate exudate coming from a focal point, I will use laser, particularly if it is not directly in the macula. However, I tend to use combination therapy for the majority of my patients.
I think that some of us forget to also tell our patients that no matter what we do, if they do not work to control their blood pressure and blood glucose levels, the DME will need to be treated far more frequently.
Dr. Pieramici: I still find the classification system of CSME helpful. My initial thoughts when I see a patient for the first time include: Is there CSME and is it involving the center of the fovea? If CSME is present but it is not involving the foveal center, then I will use laser. When the edema involves the foveal center, I will tend to use agents like anti-VEGF or steroids, which more directly address the edema.
OCT and FA can be very complimentary to making the diagnosis and deciding on how you are going to treat these patients, OCT determining foveal involvement and helping us quantify the efficacy of treatment. FA, however, in my opinion, is more important for determining the stage of the disease and for identifying ischemia.
Dr. Boyer: Do you use widefield FA, and if so, do you alter your treatment decisions based on this mode of imaging?
Dr. Kiss: I use widefield FA almost exclusively. The evidence toward the significance of the peripheral retina is emerging, so more data are required. However, it makes sense that viewing the retina at 200° offers an advantage. For example, we may not see anything on examination in diabetics who have hemoglobin A1c levels of 9 or 10, but then all of the sudden, standard of FA will reveal large areas of ischemia and nonperfusion that clearly must have been visible in the periphery at an earlier stage.
We have looked at widefield FA compared with standard fields and how ischemia is related to ME.9 There appears to be a loose association between the amount of ischemia and, although not necessarily the amount, but the presence of edema. So if the periphery plays a role, then how do we use this in clinical practice? Targeted panretinal laser photocoagulation (PRP) has been tried in the past, but I do not think that these attempts were looking far enough out into the retina. Lasering the entire retina, even the perfused retina, may not be necessary if the patient has controlled his or her diabetes and hypertension.
Dr. Boyer: We also have indirect laser that can go further out and treat some areas in the peripheral retina, increasing access. Does anyone else use widefield FA?
Dr. Pieramici: We use widefield FA to a limited extent. In my opinion, you can achieve a similar evaluation of the patient just using standard-field FA with nice sweeps into the periphery. FA is a helpful staging tool because there are patients with very severe nonproliferative or early PDR that might be overlooked with ophthalmoscopic exam alone. Knowing the patient is going to need PRP might also push us to address non-center involved CSME or non-CSME prior to initiating the PRP.
THE ROLE OF INFLAMMATION IN DME
Dr. Boyer: Dr. Dugel, you indicated that depending on the stage of diabetic change, there may be different factors in play, including an early role of inflammation. How does inflammation play a role in this condition?
Dr. Dugel: There is little doubt that inflammation plays an important role in DME. I think we are starting to understand this. Remember that decades ago we did not refer to diabetic retinopathy as such, rather, it was termed diabetic retinitis.
Steroids have a powerful effect on DME because of the inflammatory factors involved in the disease. I remember hearing the late Judah Folkman, MD, saying that there are no drugs he knew of that changed a cell as profoundly and in so many ways as a steroid. It is not just the type of steroid that we choose, but the dose of the steroid as well as the stage of disease in which we use the steroid that determines its efficacy. The same steroid may have different effects depending on its dose and the pharmacokinetics of how it is delivered.
We are at the tip of the iceberg in our knowledge concerning steroids and the role of inflammation in DME. A bolus of a drug may work very differently than drug released in a sustained zero order or near zero order kinetics distribution (Figure 2). The data from clinical trials, such as FAME3 and other studies, clearly show that there is a significant difference in effect when you manipulate how a drug is released.
Dr. Boyer: Not all steroids are equal. For example, there are common genes that are upregulated by fluocinolone acetonide, dexamethasone, and triamcinolone acetonide, but there are unique genes that each of these activate. Additionally, the side-effects profiles for these 3 steroids are different. Our experience with intravitreal injections of triamcinolone acetonide is different than that with sustained-release fluocinolone acetonide and sustained-release dexamethasone (Ozurdex, Allergan, Inc.).
Dr. Kiss: I agree. Just as with the anti-VEGF agents where we would agree that the molecular makeup of pegaptanib sodium (Macugen, Eyetech) is not the same as ranibizumab, which is in turn not the same as aflibercept (Eylea, Regeneron), there are different molecules that make up each different steroid. With steroids, this difference is even greater than anti- VEGF because they have a stronger and more broad effect.
In regard to the treatment paradigm, patients with diabetes and DME tend to be younger than our patients with age-related macular degeneration (AMD) and have more chronic disease than those with retinal vein occlusion (RVO), so DME may require a sustained-delivery drug release more because patients will be on a drug regimen for a longer period of time. Injecting a patient aged 50 years frequently with anti-VEGF or steroids is not an optimal regimen for the long term. DME is where sustained-release drug formulations will play a significant role.
I think that we need to gain a better understanding of how these steroids affect genes at a micro-level in the retina. If we can find the exact amount of steroid that is effective over a period of time with minimal side effects, we will have an even better treatment for some of our patients.
Dr. Pieramici: Steroids as a treatment make perfect sense in the presence of inflammation. Injecting intravitreal triamcinolone has been the approach that many of us have taken because it is convenient and seems to last longer than injecting a bolus of dexamethasone. However, we have essentially been treating a disease that is more akin to a marathon with an approach suited to a 100-yard dash. With an intravitreal injection of triamcinolone, we treat and the ME comes back, so we treat again, and it comes back, and so forth. A lower level sustained release of a steroid will most likely have a better long-term effect for maintaining reduction in ME and stabilizing vision in patients.
Dr. Kiss: We always talk about the VEGF pathway and the inflammatory pathway, when in fact, VEGF is part of the inflammatory pathway. It is not that one is more important than the other.
Dr. Boyer: These pathways do overlap. If you look carefully at what activates VEGF, it is the inflammatory pathway, which then activates hypoxic inducible factors and tissue necrosis factor (TNF).
I think that we will be using combination therapies more frequently in the future, including pharmacotherapy with laser in order to address residual edema and to improve visual acuity.
THE NEED FOR A SUSTAINABLE TREATMENT MODEL
Dr. Boyer: A key point that was made earlier is that DME is a chronic disease that affects the working population, and having patients come in for a monthly injection of anti-VEGF, for example, is not a sustainable solution. There is a definite need for longer-delivery treatment options.
Dr. Dugel: Sustainability is definitely an issue. It is very difficult for patients to take off time once a month to come in for an injection. Additionally, we know that physiologically there are hundreds, if not thousands of factors that contribute to DME, and if we inhibit just 1 of these (VEGF), we are missing the mark. Both combination therapy and sustained-release drug delivery make sense in this setting.
Dr. Pieramici: There may be a downside to taking a VEGF inhibition approach. There certainly is some physiologic response or pathophysiologic response in patients with diabetes that causes the body to upregulate VEGF. It is possible that VEGF may be a neuroprotective agent in such an instance and if we are applying pan-VEGF suppression, although we are reducing neovascularization and helping the retina look better, we may causing harm in the long term. Although this is just a theory, it is logical to say that in a disease with so many contributing factors, we cannot just take 1 target factor and completely eliminate the pathology. Instead, we should try to manipulate a number of factors to achieve the best response.
Dr. Dugel: This is a valid point. There have been experimental studies in rodent models that show that chronic VEGF suppression causes changes in the choroidal vasculature.10 This may not be as important in patients with AMD because they tend to be older, but it is important in younger patients who we will be treating for longer periods of time. At this point, we simply do not know how important this is, but it is something to keep in mind.
CASE EXAMPLES
Dr. Boyer: Let's consider some cases and how we would approach them.
A man aged 45 years presents with 20/40 vision, an A1c hemoglobin level of 9, and hypertension. The patient is relatively asymptomatic because his vision in the fellow eye is 20/20. How would you approach this patient?
Dr. Kiss: I would examine the patient with widefield FA and OCT to see how much damage exists. I would then talk to the patient about systemic control and explain that that, although he may not realize the effects, damage is occurring in the affected eye, the unaffected eye, along with his heart, kidneys, and brain. I routinely tell my patients that once they have lost visual acuity, it is harder to regain it vs maintaining vision.
I usually start treatment with an anti-VEGF agent, if there is not enough focal leakage to treat with laser. Once the retina is flattened, I may choose to use laser. Often I find that the threat of an injection or laser is enough to motivate patients to improve their systemic control and so when they return in 3-6 months, their blood sugar and hypertension is much better, changing the cycle of DME. If, however, nothing has improved, I open a dialogue with the patient about long-term management of the DME, with repeated intravitreal injections and laser. If there are peripheral areas of ischemia or nonperfusion, they may be contributing to the diffusion of the edema, or I may detect something on OCT that can be addressed with surgery.
In sum, my approach with this patient would be to get more information, educate the patient, and then apply a step-wise treatment regimen.
Dr. Pieramici: This is a good approach. This patient's main problem is his medical control and showing him the damage it has caused on FA and OCT may be valuable, not only from a clinical standpoint, but also from an educational standpoint. Coordinating care with the patient's primary care physician is important to bring him under control.
I would also treat this patient with an anti-VEGF agent to address the edema. What I add later depends on his response and how often I will need to treat.
Dr. Boyer: Let's say a patient presents with a history of diabetes for the past 12-14 years. He is in his late 60s, and his visual acuity is 20/200 with massive amounts of edema. The patient is generally noncompliant and does not follow up with his internist as often as required, and so he has no idea what his A1c levels are.
How would you approach this patient?
Dr. Dugel: I see this type of patient frequently. I follow steps similar to what Drs. Kiss and Pieramici described in terms of patient counseling. For treatment, the first thing that I want to do is to ensure there is no mechanical component that I could address with surgery. It is more than likely with this type of patient that I will find many areas of ischemia on FA. If there is a history of noncompliance, I believe it is unrealistic to think that this patient will come in for injections every month for several years.
I would counsel this patient toward a sustained-delivery steroid to lower the frequency with anti-VEGF injections that would be needed. I may consider laser also. In addition to treatment, I would want to work with their diabetologist to ensure better compliance and systemic control.
Dr. Boyer: The mechanical component to DME has been brought up a few times in this discussion. I have been pleasantly surprised that in some cases, mild epiretinal membranes (ERMs) that I considered to be a contributing factor returned to a normal contour after treatment with anti-VEGF or steroid. The edema may recur more frequently due to the ERM, but the patients' vision improves with injections.
Dr. Dugel: The problem with anti-VEGF agents for DME is that we are addressing a long-term, frequently recurring disease. The short-term data that we have is important, but in DME, even 2-year data is short term because this is a disease that endures for decades. There is no doubt that monotherapy with anti-VEGF agents has proved successful in the short-term, but it is the long-term sustainability that makes me think that a combination approach will be needed.
Dr. Kiss: With my patients who have diffuse edema, I am not always getting as a robust response from anti- VEGF agents that I can get from a steroid injection. In some cases, I find that a bolus injection of steroid will flatten massive amounts of edema much faster than repeated injections of anti-VEGF.
Dr. Pieramici: The noncompliant patient with large amounts of diffuse edema is, unfortunately, fairly typical. Like Dr. Dugel, I am starting to treat these patients with more than anti-VEGF agents because unlike with AMD, these patients have several comorbidities and to add on a burden of monthly injections can be overwhelming for everyone involved. I will definitely use combination therapy, whether with anti-VEGF and laser, anti-VEGF and steroids, or steroids and laser.
We are going to have clinical evidence suggesting that all of these are beneficial, but how we put them together will be on a case-by-case basis. It is easy to forget that when we look at results from clinical trials, we are generally looking at an idealized average patient. With our individual patients, we will have to tailor the care. This is an interesting time be a clinician because instead of working with a cookbook approach, we will have to practice medicine using the studies as guidelines but customizing the care to the individual.
Dr. Boyer: I find that many times I will start treatment with an anti-VEGF agent, and if I do not see the response I am seeking, which is relatively good flattening, I will add a steroid. Certainly, if we have a long-term steroid available, this will be better for both the patient and the doctor.
TAKE-HOME POINTS
Dr. Boyer: What are some important take-home messages for those who are participating in this continuing medical education activity?
Dr. Dugel: I believe it is important to closely review the recent studies that have evaluated anti-VEGF, laser, and steroids for DME. For example, in Protocol B, it is interesting to look at the patients who were selected and who benefitted vs those who did not from steroids and laser (Figure 1). Although this represented just a few patients, there is a signal that I think is important as to where steroids may fit in: in the study, it is a marker for the importance of inflammation.
Additionally, the FAME data subanalysis showed a significant difference in how patients responded based on the duration of the diabetes. FAME was actually 2 studies, A and B, conducted in very different regions and demographics, but the results were remarkably consistent. It is also important to note that in both A and B studies, there was a complete OCT/visual acuity disconnect. There is a good amount of emerging evidence that will help us to see DME more as a staged disease, with each stage requiring a different combination of treatments.
Dr. Kiss: As we are understanding more about diabetes, we are learning that a multifactorial approach will be the way to treat. The 2 patients that Dr. Boyer described illustrate this. DME is a different disease process at each stage, and each patient is different also, so we have to individualize our treatment to account for lens status, systemic status, and even to the age of the patient, because the current treatment paradigm of monthly intravitreal injections for the next 40 years in the life of a patient is unrealistic. If you consider other areas of medicine, such as cancer, there is not 1 drug that works for all patients; the same is true in DME.
Dr. Pieramici: It is a wonderful time for treating patients with DME. We used to have only 2 tools, laser and surgery, which was frustrating for us and for our patients because we would treat and treat again, and the edema would persist with the vision rarely improving. We now have many more things we can pull off the shelf with others in development. We also have a better understanding of diabetes itself and the importance of good medical care.
The clinical trial data are helpful to us in that they help guide treatment, but as has been said before, we will have to take a customized approach.
Dr. Boyer: In the future, I think we will have better ways to identify who is going to respond to the various drugs. The direction where we are headed with research is to look at genetics and attempt to identify those who may respond better to anti-VEGF or possibly better to steroids, or to find out how much ischemia is necessary to create the VEGF.
Early in the course of diabetes, hyperglycemia is present, and the inflammatory component starts the VEGF cycle. Often we see patients present in the middle part of the VEGF cycle when it is full blown and where anti- VEGF drugs can help. It is the chronic phase of DME where sustained-release steroids will have the biggest role in conjunction with debulking VEGF. Combination therapy may help us keep patients stable for long periods of time and achieve better visual results.
_1773249222.png?auto=compress,format&w=75)
