When it comes to treating diabetic macular edema, this is arguably both the best of times and the worst of times. “We’ve never had more treatment options,” notes Allen C. Ho, MD, FACS, director of retina research at Wills Eye Hospital and professor of ophthalmology at Thomas Jefferson University School of Medicine in Philadelphia. “At the same time, we’re witnessing more patients with the problem than ever, because as much as 10 percent of the U.S. population may have diabetes mellitus. Over time, particularly if your blood sugar remains uncontrolled, you’ll develop problems such as diabetic retinopathy, and more specifically, diabetic macular edema. That’s the number one mechanism of vision loss in patients with diabetic retinopathy.”
The positive side of the equation—the availability of multiple treatment options, including anti-vascular endothelial growth factor drugs, steroids and laser—raises some important questions regarding which of the alternatives are most effective and how they might best be used.
Working With Anti-VEGF
Currently, drugs that treat an overabundance of VEGF inside the eye are the most popular approach to managing DME. However, their use requires a number of decisions involving which drug to choose, how to use it, and whether to combine it with other treatment options such as steroids or laser.
Steve Charles, MD, FACS, FIC, founder of the Charles Retina Institute in Memphis, Tenn., is particularly concerned with dispelling one popular myth regarding the choice between ranibizumab and bevacizumab. “Doctors often have the impression that the two anti-VEGF drugs are equivalent because of the results of the Comparison of AMD Treatments Trial, which showed that the two drugs were almost equivalent when treating macular degeneration,” he says. “What they may not realize is that eyes with DME have 5,000 times as much VEGF as eyes with macular degeneration. This is also true for central retinal vein occlusion, as was shown by Lloyd Paul Aiello, MD, in one of the most important retina articles ever published, back in 1994.1
|Anti-VEGF drugs are often the first-choice treatment for DME. This eye received ranibizumab monthly (baseline, left; 24 months, right). ETDRS letters improved from 64 to 71; Snellen acuity improved from 20/50 to 20/40. (Images courtesy of Genentech.)
“At the same time,” he continues, “the molecular affinity of ranibizumab for VEGF is 25 times higher than that of bevacizumab. With 5,000 times as much VEGF to manage in DME, this increased molecular affinity makes a huge difference. So when surgeons tell me they use steroids because they tried bevacizumab and were not impressed by the results, I tell them they tried the wrong drug.” Dr. Charles adds that aflibercept also appears to be 25 times as effective as bevacizumab, although it isn’t approved for treating DME at this juncture.
Dr. Ho agrees that when it comes to DME, the two anti-VEGF options are not equal. “In my clinical experience, Lucentis is superior to Avastin for diabetic macular edema,” he says. “So, that’s my first choice for treating a patient, as long as the patient has no issues with insurance coverage or copays. I think it’s a better drug. But which one I use becomes a reimbursement issue, because most patients can’t afford Lucentis at $1,800 to $2,000 per injection, per eye.”
David S. Boyer, MD, in private practice with the Retina-Vitreous Associates Medical Group in Los Angeles and part of the clinical faculty at the Doheny Eye Institute, USC School of Medicine, says he has also found ranibizumab to be a little more efficacious, but agrees that insurance coverage is an issue. “A lot of these patients have commercial insurance,” he says. “Until we can sort out the in-surance issues, I’d rather go with bevacizumab just to make sure the patients don’t have to pay out of pocket. You don’t want people who have private insurance with large deductibles to end up in a bad situation. However, if the eye has a large amount of leakage that doesn’t respond to treatment, I might add a steroid or change from bevacizumab to ranibizumab. Hopefully, aflibercept will also be an option in the future.”
Despite the general agreement that ranibizumab is more effective, Dr. Boyer notes that the jury is still out on whether this is true in a clinical setting. “Our feeling is that it dries the retina better, and patients do a little better,” he says. “The big question we have is, in a year or two, will that make a difference in vision? That’s why they’re doing the Protocol T trial at DRCR.net, working with the National Eye Institute. Protocol T is comparing treatment of DME using bevacizumab, ranibizumab and aflibercept. After the initial treatment, subjects are treated as needed, whenever fluid is present. We’ll see how all three drugs do. Recruitment is going well and will probably be done sooner than expected, because everybody really wants to know the answer to this question.”
Steroid Pros and Cons
In addition to the anti-VEGF options, a number of steroids are sometimes used to address DME. “We have several products available off-label for the treatment of diabetic macular edema, including triamcinolone, available from Alcon as Triesence,” says Dr. Ho. “That’s not approved for diabetic macular edema, but it works. I like that product because it’s not compounded; it’s made specifically for the eye, and we know its benefits and risks. There’s also the Allergan dexamethasone implant, Ozurdex, which can work for diabetic macular edema, although it’s approved because of its results with retinal vein occlusion. Our other tools include laser treatment, and of course the general medical guidelines that address diabetes by optimizing blood pressure and blood sugar levels.
“Beyond that,” he adds, “many retina specialists, including myself, were surprised that the FDA did not approve the Alimera fluocinolone implant that’s approved in Europe for the treatment of DME. I believe the lack of that option, particularly for pseudophakic eyes, is detrimental to our patients.”
However, not all surgeons see steroids as an acceptable treatment option. Dr. Charles says he may be the only vitreoretinal surgeon in the world who never uses steroids to treat DME. “This is true for two reasons,” he says. “First, although steroids have positive effects—they do make the edema regress—nobody has found a way to decouple them from their negative side effects, which are steroid glaucoma and cataract formation. Those side effects are built into the mechanism of the drug. Second, I believe the reasoning behind choosing steroids over anti-VEGF treatment is based on faulty assumptions.”
Regarding the first point, Dr. Charles notes that people often underestimate the likelihood that side effects will occur in any given patient. “Steroids provide immediate gratification,” he says. “You inject them and the edema goes away and the patient’s happy. But if the patient is phakic, the odds of creating a cataract are about 90 percent if you use triamcinolone, and a significant percentage if you use Ozurdex.
|A patient treated with Kenalog before treatment (left) and at five weeks (right). Corticosteroids can reduce edema, but frequently produce serious side effects such as glaucoma and cataract. (Image courtesy David S. Boyer, MD.)
“Then, there’s steroid glaucoma,” he continues. “There’s an idea left over from the 1970s that only 6 percent of the population are steroid responders. That’s simply false. That’s based on the notion of using topical 1% prednisolone drops—Pred Forte and the like. With intravitreal steroids, the incidence of steroid glaucoma is much higher, and it’s dose-related. For example, with fluocinolone steroid injections, the incidence of glaucoma is greater than 90 percent, and about 35 percent of patients receiving this drug need glaucoma filtering procedures. That’s completely unacceptable. It might be less so if glaucoma filtering procedures had an enormous success rate and no complications, but that’s not the case; they have a significant failure rate and often lead to hypotony, choroidals, infected blebs and other problems. This is an operation you want your patient to avoid at all costs.
“The other issue with steroid glaucoma is that we were taught, based on the use of 1% topical prenisolone acetate, that the side effects are reversible; stop the drops and the problem goes away,” he says. “But that’s not true with intravitreal steroids. After a few months the drug will lose its effect, but steroid glaucoma may become permanent. So if you’re relying on steroids to manage the DME, you’re going to have to inject again and again, and in time the patient will develop a cataract and glaucoma. That’s why I never use steroids.”
Steroids vs. Anti-VEGF
Dr. Charles also sees problems with the reasoning that leads many surgeons to choose steroids over anti-VEGF treatment. “I’m well aware that some surgeons say they’ve never had good results with anti-VEGF compounds,” he explains. “Inevitably, when I ask which one they tried, it was bevacizumab. Usually, the rationale for choosing bevacizumab is that the surgeon doesn’t want to burden Medicare with enormous fees, but the fact is that bevacizumab simply isn’t very effective when treating DME, while ranibizumab is.”
Dr. Charles adds that insufficient frequency of injection is also a common reason for inadequate results. “There is no six-week rule or eight-week rule, or every-three-months rule,” he says. “Patients have different levels of VEGF, and when you inject the drug into the vitreous cavity it dissipates at different rates. Some surgeons inject every six weeks, and when the patient complains that vision deteriorates after three weeks the surgeon concludes that anti-VEGF drugs don’t work as well as steroids. In fact, the doctor just needs to shorten the interval for that patient. Some patients produce more VEGF than others, and they need more frequent injections. Surgeons treating macular degeneration encounter the same phenomenon all the time.
“Another problem with injection frequency is that practical realities often cause the interval to become lengthened,” he notes. “For example, the doctor may ask the scheduler to give a patient a one-month appointment. How many days does that turn out to be? In many offices, it ends up being six or seven weeks, because at the one-month mark the doctor’s out of the office going to a meeting or the patient is busy. The next thing you know, the one-month appointment happens eight weeks later, and the patient has had good vision for four weeks and four weeks of not-so-good vision. Some doctors then conclude that the anti-VEGF drug isn’t working and switch to steroids instead. The reality is that ranibizumab works for almost every patient if you get the interval right.”
In fact, Dr. Charles feels strongly that ranibizumab should be the first-line treatment. “This isn’t rescue therapy, where first you do grid laser and then you see the patient in 91 days, the global period for reimbursement, and if that doesn’t work you use steroids, and if that causes steroid glaucoma, then you see the patient in 91 days,” he says. “Then you finally try bevacizumab and the edema goes away but the patient’s vision doesn’t improve enough. Big mistake. You should have started with ranibizumab.
“Does the data support this approach?” he adds. “Yes. The DRCR.net study published two years ago did a head-to-head comparison of triamcinolone, laser and ranibizumab. Ranibizumab won big-time. Given that fact, it’s amazing to me that some of the steroid options are still so popular.”
Still in the Running
Despite the downsides of treating with steroids, many surgeons still see them as a valuable second-line option. Dr. Ho says he generally only resorts to steroids if anti-VEGF injections and laser are not working. “I tend not to use steroids primarily because of their side effects: elevating the pressure in the eye and causing cataracts,” he says. “However, there’s definitely a role for steroids. Some patients don’t respond to anti-VEGF, but respond very well to steroids.”
Dr. Boyer says that in his experience steroids still have a place in the treatment of DME. “I think steroids are a second-line drug, not a primary drug, mainly because of the side-effect profile,” he says. “Nothing dries as effectively as anti-VEGF, but if there’s a lot of fluid and vision is down, steroids do a very good job. I’ve had pseudophakic patients in whom I used a steroid and then continued anti-VEGF afterwards, periodically giving the steroid. They’ve done very well. As long as patients can get followed for the glaucoma aspect of it, I think it’s a reasonable way to treat.
“Some studies have shown that patients with poor vision and a lot of fluid do better with steroids,” he continues. “And if you look at patients who are aphakic, where cataract formation is not an issue, steroids can be very comparable to anti-VEGF as far as drying and visual outcomes. In addition, one of the steroid implants that isn’t yet FDA approved has shown significant benefit in patients with longstanding macular edema. An implant could turn out to be a valuable tool to add to other treatments in appropriate patients, hopefully reducing the number of treatments required and improving vision.
“Of course, you have to pick your patients,” he adds. “You’re not going to use a steroid in a young patient who is phakic, because he’ll develop a cataract. But if you pick your patients properly, I think there’s a place for steroids. They’re just not a primary treatment.”
Treating With Laser
Laser was one of the earliest DME management options, and although it’s largely been eclipsed by the drug-based alternatives, most surgeons still see it as a valuable part of their armamentarium. For example, Dr. Charles says he still uses laser to treat DME, but only certain types of laser, and only under certain conditions. “Treating microaneurysms using focal laser with short-duration pulses, like those produced by the OptiMedica Pascal laser, is still an appropriate therapy,” he says. “Of course, you can’t use the laser if the microaneurysms are parafoveal.”
Dr. Boyer agrees that a key factor when deciding on treatment is whether the edema is center-involving. “If the patient has clinically significant macular edema that’s not center-involving, we may decide to do laser treatment,” he explains. “If you have a lesion outside the fovea, laser is usually a very effective treatment. If we find center involvement, and the leaks are near the fovea, we’re going to start with an anti-VEGF drug. The problem has been that in all of the studies up to now that have center-involving macular edema, the addition of laser has never been shown to be that helpful. I rarely use it in center-involving because the anti-VEGF drugs do such a good job. But I think if the edema is non-center-involving there’s a place for it.”
However, the way the laser is used makes a difference. “I’ve never used, nor do I recommend, grid-style laser,” says Dr. Charles. “I’ve never understood the rationale for grid laser—ablating areas of the retina that are not involved, hoping somehow to affect the macular edema. Some people say that it’s similar to selective laser trabeculoplasty, upregulating cytokines, which somehow reduces the edema. There are some studies showing that it does have an effect, but it still seems like a questionable approach. Others use micropulse laser and claim that it works.
|An eye previously treated with long-duration laser (nasal) and short-duration laser (temporal) at different times. Conventional laser burns with a duration of 100 to 1,220 ms (close-up view, center) can cause heat diffusion, inducing collateral cell death (sometimes called “RPE creep”). This can cause a 100-µm spot to grow to 300 µm in two to three years. In contrast, PASCAL-type laser burns (close-up view, right) are 2 to 30 ms in duration; these eliminate heat diffusion and lesion growth post-treatment, remaining smaller and more regular in shape. (Images courtesy Mark S. Blumenkranz, MD.)
“One of the main problems with these approaches is that if you don’t use Pascal-style laser, thermal diffusion can cause the damage from the laser spots to grow to triple the size two years after you treat the patient,” he says. “There can be sub-lethal damage that ultimately kills cells surrounding what you thought was a 50-micron spot. Once those cells die, it becomes a 150-micron spot. Because of this phenomenon, surgeons who treat fairly close to the fovea can leave the patient with visual field loss. In our practice we’ve seen many grid laser treatments done by other surgeons where the patients’ current chief complaint was simply, ‘I can’t see,’ even though their vision was 20/25 or 20/20. The reason they couldn’t see was that the visual field surrounding their foveal visual field had been ablated. For all of these reasons I never use grid laser. I think it’s a bad idea.”
Nevertheless, Dr. Boyer believes that in some cases there may be an argument for using panretinal photocoagulation. “When you see large areas of capillary dropout, at least in diabetics, there may be a place for panretinal photocoagulation, targeted just to the area of nonperfusion,” he says. “That’s something I think should be investigated to determine whether it really is helpful.”
Dr. Boyer adds that he’s been using the Navilas laser (OD-OS, Teltow, Germany), an advanced, computer-guided system. “I’m extremely impressed with its accuracy,” he says. “When the accuracy is this high it may be possible to get a better result with less energy and less damage than we can get with conventional lasers. The problem is, there aren’t a lot of these lasers out there. For that reason, we don’t yet know when to use it and when it will give us better results.”
Ultimately, Dr. Boyer admits that knowing when to use the laser is an imperfect science at present. “We don’t have a paradigm that tells us when to apply it,” he says. “We all think we know, and we try to do the best for our patients. If a patient has a lot of leaks outside the fovea I apply it, but in the long run, we don’t really know if it makes a difference. So I’m using it judiciously.”
In terms of treatment protocol, Dr. Ho says he usually starts with injections of Avastin or Lucentis. “We see how the patient responds,” he explains. “We do that monthly for a while to dry out the macula, sometimes in combination with laser treatments. Corticosteroids are my next option; surgery with peeling of the internal limiting membrane is my last option. Often, our primary therapy will shift to less-frequent-than-monthly injections, with laser used secondarily. Sometimes when the macular edema is discrete and outside the center of the macula I’ll just use laser therapy and not even resort to the injection.
“The secret becomes, how do you figure out the right cadence for the injections?” he continues. “Let’s say a patient responds to a series of three or four injections of Lucentis and is stable. We’ll gradually extend the interval between the injections because patients often don’t need them on a monthly basis. Some will only need an injection every six, eight or 10 weeks. It’s the treat-and-extend approach.”
Dr. Ho notes that surgeons don’t have to be quite as aggressive about treating diabetic macular edema as they tend to be with wet macular degeneration. “The macula is more tolerant in diabetic macular edema, so I’m not as aggressive in terms of looking at an OCT scan with macular edema and reflexively injecting,” he says. “When the problem is due to diabetic retinopathy you don’t have to dry every little cyst in the macula. I’m more aggressive with wet macular degeneration, based on how the OCT looks. However, I do treat DME if I find significant edema in both the OCT and the clinical exam, including areas of leakage in the macula.”
Although he doesn’t use steroids, Dr. Charles has no problem combining laser with anti-VEGF treatment. “I say, if it’s bad neovascularization, knock it down with ranibizumab and keep it down with laser,” he says. “I often combine them. And the same argument applies with DME.”
Dr. Boyer agrees that combining options may be advantageous in some patients. “This is not a case of one or the other,” he says. “In my experience, combination therapies can work very well. I tend to use combination therapy if I think it’s really a bad case, but it’s a paradigm that’s still being fine-tuned at this point.
“The bottom line is, every patient is different,” he adds. “The blood pressure is different, the hemoglobin is different, and every patient has a different level of capillary dropout. So every patient may respond to treatment differently.”
Helpful Clinical Strategies
The following additional strategies can also help maximize the effectiveness of DME treatment:
• Don’t skip the fluorescein angiogram when there’s a disconnect between visual acuity and clinical exam and OCT information.
“A fluorescein angiogram still plays a role in evaluation of diabetic macular edema, particularly when you treat and the edema goes away, but vision is not better,” says Dr. Ho. “That may be because there’s macular ischemia. Even without adequate macular capillary perfusion the macula can become edematous due to a breakdown of the outer blood retinal barrier at the level of the retinal pigment epithelial tight junctions. The fluorescein angiogram can help you determine that.” Dr. Ho notes that he would try anti-VEGF and steroids to treat the macular edema in the setting of macular ischemia, but would avoid using the laser.
• Be persistent with your treatment.
“One good thing about DME is that it’s inner-retinal, not outer-retinal, so if the patient does miss a treatment and leakage increases, usually they can get back the vision they’ve lost,” says Dr. Boyer. “Nevertheless, you have to be persistent. Whether you do treat-and-extend or some other protocol, you can’t treat three times and give up. And the patients have to realize they’re going to receive a number of injections so they can get the best possible results.”
• Remember that steroids have fewer downsides when the patient is pseudophakic.
“You can resort to steroids more quickly when the patient is pseudophakic,” Dr. Ho points out. “A pseudophakic patient obviously can’t develop a cataract.”
• Get the patient involved in treatment.
“The patient has to take part in the treatment, by coming in, and by controlling the disease,” says Dr. Boyer. “I try to emphasize to patients that they can control their blood sugar and blood pressure and lipids, and doing so will give them a much better chance of stopping or reducing the progression of their macular edema.”
Dr. Boyer says he uses images of the retina to help motivate patients. “I’ll do photographs and fluorescein and show patients what’s going on in their eye,” he says. “I explain that if that edema doesn’t settle they’re going to lose vision and not be able to function. That’s usually a big wake-up call. A lot of noncompliant patients become very compliant after that.”
Getting Treatment to the Patient
Of course, none of the current treatments can prevent or reverse the disease itself, but that simply increases the urgency of getting treatment to the individuals who need it.
Dr. Ho notes that when diabetic patients come in for an exam, it often looks like a bomb exploded inside their eye—yet they claim they haven’t been having any problems and didn’t think they needed to be seen. “This is not acceptable,” he says. “We need to get to these patients sooner. They need to know that their eyes should be examined regularly and completely, including a dilated retinal exam.
“Every day I come into the clinic and see patients who are legally blind from advanced diabetes who haven’t received any care,” he adds. “It may be their own fault, but we’ve got to do a better job of educating them. They need complete eye exams, whether or not they’re complaining of vision problems.”
Dr. Ho is a consultant for, or on the scientific advisory boards of, Alcon, Allergan, Janssen / J&J, Genentech, Merck, ONL, Ophthotech, Regeneron and Thrombogenics. Dr. Charles is a consultant for Topcon Medical Lasers.
Dr. Boyer is a consultant for Alcon, Allergan, Genentech, Regeneron and Novartis.
1. Aiello LP, Avery RL, Arrigg PG, Keyt BA, Jampel HD, Shah ST, Pasquale LR, Thieme H, Iwamoto MA, Park JE, et al. Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders. N Engl J Med 1994;331:22:1480-7.