Evaluating and Integrating New Evidence In The Treatment Of Diabetic Macular Edema
Susan B. Bressler, MD, is the Chair of this program. She is the Julia G. Levy, PhD Professor of Ophthalmology at the Wilmer Eye Institute, Johns Hopkins University School of Medicine, where she also became only the second woman promoted to Professor in Ophthalmology there. She attended The Johns Hopkins University School of Medicine where she received the Paul Ehrlich Award for Clinical Investigation at Hopkins’ Young Investigators’ Day in 1982 for her first project in ophthalmology on age-related macular degeneration. After completing an internship in internal medicine at Johns Hopkins Hospital in 1983, Dr. Bressler pursued an ophthalmology residency at Harvard Medical School’s Massachusetts Eye and Ear Infirmary. Following a fellowship in medical retina at Wilmer and additional retinal surgery training at Massachusetts Eye and Ear Infirmary, she returned to Baltimore in 1988 to join the faculty at the Wilmer Eye Institute.
Dr. Bressler has a large national and international referral practice at the Wilmer Institute. In addition, she has published 185 peer-reviewed articles and 53 book chapters. Her editorial board positions have included American Journal of Ophthalmology, Survey of Ophthalmology, Retina, and the Wilmer Retina Update. Her main research interest has been collaborative efforts in clinical trials as Principal Investigator of an image reading center that has served as a central unit for many clinical trials and epidemiologic investigations, Vice-Chair of the Diabetic Retinopathy Clinical Research Network (DRCR.net), and Principal Investigator of a participating clinical center in several major clinical trials. Most studies have specific emphasis on the treatment of both non-neovascular and neovascular age-related macular degeneration and all aspects of diabetic retinopathy. Dr. Bressler has received awards that include the Rosenthal Award from the Macula Society, the Olga Keith Weiss Scholar Award from Research to Prevent Blindness, a Senior Achievement Award from the American Academy of Ophthalmology, and the Gertrude D. Pyron Award for Lifetime Achievement of Outstanding Research and a Senior Honor Award from the American Society of Retina Specialists (ASRS).
Sophie J. Bakri, MD, is Associate Professor of Ophthalmology at the Mayo Clinic in Rochester, Minnesota. Her clinical interests are in diseases and surgery of the retina and vitreous, in particular age-related macular degeneration, diabetic retinopathy, and repair of complex retinal detachments. She conducts both clinical and translational research in the pathogenesis and treatment of retinal diseases. Dr. Bakri graduated from the Queen’s Medical Center, University of Nottingham Medical School, England, where she earned a degree in Medicine and Surgery, and a Bachelor of Medical Sciences. She completed her residency at the Albany Medical College in New York, where she was elected chief resident, as well as a 2-year vitreoretinal fellowship at the Cleveland Clinic Foundation in Cleveland.
Dr. Bakri has authored over 80 peer-reviewed papers and 15 book chapters on retinal diseases. She is a Principal Investigator on numerous multicenter clinical trials on novel drugs for retinal disease. She is the editorin- chief of the book Mayo Clinic on Vision and Eye Health, and is on the editorial board of the American Journal of Ophthalmology, Retina, Seminars in Ophthalmology and Clinical and Surgical Ophthalmology. She is a member of the American Society of Retina Specialists and the Macula Society, and has received an Achievement Award from the American Academy of Ophthalmology. Dr. Bakri is actively involved in teaching residents and fellows, and is the Director of the Vitreoretinal Surgical Fellowship at Mayo Clinic.
Scott M. Friedman, MD, is in private practice in central Florida, and is Assistant Clinical Professor in the Department of Ophthalmology at the University of South Florida Medical School in Tampa. He completed his undergraduate, medical school and residency training at the University of Missouri- Columbia. His residency training was in ophthalmology and he became board certified in 1988. He then completed a fellowship in retina and vitreous at Northwestern University Medical School Department of Ophthalmology; and was Assistant Professor and Chief of the Retina and Vitreous Section in the Department of Ophthalmology at the University of Florida School of Medicine in Gainesville, after which he went into private practice.
Dr. Friedman has published over 70 peer-reviewed manuscripts, including in journals such as Ophthalmology, American Journal of Ophthalmology, Archives of Ophthalmology, and Retina. He has been Principal Investigator in several Diabetic Retinopathy Clinical Research Network (DRCR.net) trials, including the evaluation of vitrectomy for DME, and the trial comparing intravitreal triamcinolone acetonide and laser photocoagulation for DME. He serves on the Executive Committee of DRCR.net, and is a Vice-Chair. He was the recipient of the American Academy of Ophthalmology Achievement Award, as well as the AAO Secretariat Award.
Michael S. Ip, MD, is Associate Professor, Department of Ophthalmology and Visual Sciences, and Co-Director of the Fundus Photograph Reading Center at the University of Wisconsin Medical School of Madison. He serves as Director of the Retina Service at the William S. Middleton Memorial Veterans Hospital in Madison. He also has an active clinical practice as a vitreoretinal surgeon, concentrating on the surgical management of retinal detachment and other conditions, as well as the treatment of age-related macular degeneration, diabetic retinopathy and others. Dr. Ip received his medical degree from New York University and completed an internship at Lenox Hill Hospital. He served his residency at the University of Pittsburgh School of Medicine, followed by a fellowship in vitreoretinal surgery at New England Eye Center, Tufts University.
Dr. Ip’s research focuses on the design and conduct of clinical trials investigating treatments for diabetic retinopathy, AMD, and retinal venous occlusive disease. He assisted in the collection, analysis and dissemination of important secondary outcomes in ophthalmic clinical trials. He has served as the Principal Investigator on 10 clinical trials, and been the protocol chair for the clinical trial conducted by the Diabetic Retinopathy Clinical Research Network (DRCR.net) comparing focal/ grid photocoagulation and intravitreal triamcinolone for DME. He has authored more than 40 peer-reviewed articles, 19 book chapters, and over 30 abstracts. He is currently Associate Editor of Archives of Ophthalmology, and is a reviewer for a number of journals (Ophthalmic Surgery and Lasers, Investigative Ophthalmology and Visual Science, Ophthalmology, Retina and American Journal of Ophthalmology). He has been recognized with several honors, including an Achievement Award and the Leadership Development Program Award, both from the American Academy of Ophthalmology.
This activity was developed for ophthalmologists treating patients with, or at risk for, diabetic macular edema (DME).
STATEMENT OF NEED
Macular edema is an important cause of poor central visual acuity in patients with diabetic retinopathy. DME is reversible in its early stages but chronic edema may lead to irreversible changes in the retina. Hence, rapid diagnosis and optimal treatment are critical for a patient’s health and well-being. This activity addresses these issues to aid clinicians in their care of patients.
At the conclusion of this initiative, participants will do the following:
- Optimally evaluate and diagnose diabetic macular edema (DME)
- Provide treatment for DME utilizing the latest clinical studies of intravitreal corticosteroids and anti-VEGF agents
- Assess safety issues in the treatment of diabetic macular edema
- Plan appropriate care for patients with diabetic macular edema
ACCREDITATION AND CERTIFICATION
The Annenberg Center for Health Sciences at Eisenhower is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians.
The Annenberg Center designates this educational activity for a maximum of 1.5 AMA PRA Category 1 Credits™. Physicians should only claim credit commensurate with the extent of their participation in the activity.
There is no charge for this activity. Statements of Credit will be mailed 4-6 weeks following activity participation, upon completion and return of the evaluation form to the Annenberg Center for Health Sciences (#4775), 39000 Bob Hope Drive, Rancho Mirage, CA 92270 or by fax to 760-773-4550.
It is the policy of the Annenberg Center to ensure fair balance, independence, objectivity, and scientific rigor in all programming. All faculty and planners participating in sponsored programs are expected to identify and reference off-label product use and disclose any relationship with those supporting the activity or any others with products or services available within the scope of the topic being discussed in the educational presentation.
In accordance with the Accreditation Council for Continuing Medical Education Standards, parallel documents from other accrediting bodies, and Annenberg Center policy, the following disclosures have been made:
Sophie J. Bakri, MD
Susan B. Bressler, MD
Bausch & Lomb
Scott M. Friedman, MD
Michael S. Ip, MD
Allergan, Inc. Genentech
Sirion Therapeutics, Inc.
The faculty for this activity have disclosed that there will be discussion about the use of products for non-FDA approved indications:
Additional content planners: In accordance with the Accreditation Council for Continuing Medical Education Standards, parallel documents from other accrediting bodies, and Annenberg Center policy, the following disclosures have been made:
The following have no significant relationship to
Richard Trubo (medical writer)
All staff at the Annenberg Center for Health Sciences at Eisenhower have nothing to disclose.
The ideas and opinions presented in this educational activity are those of the faculty and do not necessarily reflect the views of the Annenberg Center and/or its agents. As in all educational activities, we encourage practitioners to use their own judgment in treating and addressing the needs of each individual patient, taking into account that patient’s unique clinical situation. The Annenberg Center disclaims all liability and cannot be held responsible for any problems that may arise from participating in this activity or following treatment recommendations presented.
This activity is supported by an educational grant from Genentech.
This activity is an enduring material and consists of a monograph (print or online). Successful completion is achieved by reading the material, reflecting on its implications in your practice, and completing the assessment component.
The estimated time to complete the activity is 1.5 hours.
This activity was originally released in January 2011 and is eligible for credit through December 31, 2011.
This piece is based on a discussion among the faculty members and was written by a writer from the Annenberg Center. Faculty have final editorial control for the piece.
Diabetic eye disease is one of the leading causes of blindness in working-age people in the United States and throughout much of the developed world. Diabetic macular edema (DME) is a common manifestation of diabetic retinopathy, and is the result of retinal microvascular changes in individuals with diabetes.
The prevalence of diabetic eye disease increases with the duration of diabetes. Data from the Wisconsin Epidemiologic Study of Diabetic Retinopathy, a population-based study in southern Wisconsin, showed that in people whose age at the time of diagnosis of diabetes was less than 30 years old and who were taking insulin, the rate of macular edema in those who had a diabetes duration of 20 or more years was 29%. In those people whose age at diagnosis was 30 years or older, the prevalence rate of macular edema was 28% in both type 1 and type 2 diabetes if their duration of diabetes was 20 or more years.1
From a systemic standpoint, the key steps that can be taken to decrease the risk of diabetic eye disease, including DME, in a diabetic patient are to achieve tight glycemic control and maintain it over a long duration, as well as work toward excellent control of blood pressure.
DIAGNOSIS AND EVALUATION
There is no consensus on how to begin the diagnostic process for DME. Many ophthalmologists start with an examination using slit lamp biomicroscopy and a non-contact method such as a 78-diopter or 90-diopter lens examination with the slit-lamp biomicroscope. If there is uncertainty about whether the patient has DME, you may move to placement of a fundus contact lens. You would look for microaneurysms, intraretinal hemorrhage and lipid, areas of diffuse retinal thickening, and cystoid edema.
“My preferred method of evaluation for diabetic macular edema is to bypass the indirect lens and go straight to contact lens methodology. It’s easy to put on the lens if you use topical anesthesia and a hardcontact- lens wetting solution, without disturbing the cornea and potentially interfering with the clarity of images that might be done afterward. Also, the relative immobility of the eye and the eyelids that is achieved appears to be superior to that of using an indirect lens. Most important, the optical properties of the contact lens method are superior for stereopsis, which should facilitate the doctor’s ability to appreciate retinal thickening more so than with the 78 or the 90 diopter lens. The more stable view and better stereopsis often make it possible to see minute areas of lipid more clearly than with an indirect lens.
“You’re as good with the contact lens method as
your experience permits. So if it’s something you’re
accustomed to doing all the time, it’s faster than
using an indirect lens.”
Susan B. Bressler, MD
The advent of optical coherence tomography has altered the way many clinicians approach DME. OCT, in fact, has revolutionized the diagnosis and management of macular disease, including DME. Rather than moving directly to contact lens examinations as an initial step, some clinicians rely instead on OCT to facilitate the diagnosis. In part, that’s because some ophthalmologists find the contact lens exam to be more cumbersome, and have concerns that it may affect the quality of photography and/or angiography.
When diabetic patients come into your office, should all of them be screened with OCT? The scientific data aren’t conclusive in this area. However, many retinal specialists believe that if a patient has been diagnosed with diabetes and has at least minimal non-proliferative retinopathy, OCT can be sensitive in picking up subtle areas of macular edema that possibly can’t be seen clinically. This is important because these patients may benefit from earlier treatment or more frequent follow-up.
“I used to do a lot of contact lens examinations.
But with the advent of OCT, I am doing much fewer
of them. In fact, at this point, I almost never or very
rarely do a contact lens exam, for reasons that include
patient comfort level, practicality and efficiency in
Michael S. Ip, MD
A DRCR (Diabetic Retinopathy Clinical Research Network) study currently under peer review looked at the prevalence of subclinical DME – that is, a central subfield thickness measurement that is 2 standard deviations or greater above the mean in eyes (≥ 225 µm) that otherwise looked to be clinically absent of macular edema. It found a low prevalence of subclinical DME of less than 4%.
When a patient’s visual acuity is declining, then you may find OCT useful in confirming what has already been seen clinically, or to search for what you suspect is present.
“I’m judicious about my use of OCT in patients who
have 20/20 visual acuity and are asymptomatic. We
already know from many studies that as the retinopathy
level progresses along the ETDRS severity scale,
the chances of edema increase. So if I have a patient
who has less than moderately severe non-proliferative
disease, for example, and he/she has 20/20 vision,
and I don’t detect any macular edema on clinical
examination, then I have to ask myself the question,
‘Why would I do OCT? What would I be screening
for?’ Even if I found some subclinical DME, I don’t
think I’d treat this patient. I believe that when we do
OCT, we do so to detect subclinical DME and then
perhaps change our follow-up pattern for patients,
maybe advising them to come in more frequently.
Michael S. Ip, MD
“If a patient has a decrease in vision and there’s
no apparent cause such as cataract, then I would
definitely recommend imaging to include OCT
because of the information we can learn from it. It’s
also important to look for other causes of decreased
vision beyond diabetes.”
Sophie J. Bakri, MD
If you already know that macular edema is present, you must determine whether it is center-involved or non-center-involved. Data indicate that if the center is involved, the visual prognosis is worse than in non-center involved cases. The question is whether OCT is needed to assist in classifying eyes as center or non-center involved DME or whether differentiation on clinical exam is sufficient. Most clinical trials require OCT assessment to confirm abnormal thickening in the central subfield as criteria confirming center-involved DME.
Many retinal specialists believe that it is important to quantitate the amount of edema present and, in cases initially sparing the fovea, determine whether there’s encroachment toward the center over time. This information will help measure the patient’s progress, and assist you in making treatment decisions.
“If you’re going to initiate treatment based on DRCR
protocol I results, then you would not need to get an
OCT every month during treatment since the patient
will receive monthly injections for 4 months regardless
of the OCT findings. So for a patient with non-center
involved macular edema, I would certainly consider
repeating an OCT at the next follow-up visit, which
may be 4 months later, if I elected laser treatment or
observation when I initially encountered the patient.
If you go strictly by the research, you may decide to
increase the frequency of OCTs after the first 4 months
of intravitreal pharmacotherapy. But if normal visual
acuity is achieved during treatment, I would probably
see the patient back in another 2 to 4 months,
and then repeat the OCT at that time. In the DRCR
protocol I, OCTs were obtained frequently during
follow-up looking for increased retinal thickening.
But when patients are not in a trial, and their visual
acuity has not deteriorated from the previous visit –
and clinically they don’t have edema – you may want
to consider not obtaining OCT on that visit.
Scott M. Friedman, MD
“If we are following a patient outside the context of
protocol I or a clinical trial, it’s never easy to determine
when you want to reinject the pharmacotherapy. So
I would say that we need a lot of OCTs, although we
don’t need them on every visit. However, in those
patients who have some response, and it’s equivocal
whether or not the DME is getting better – or if the
DME is resolved – then I think OCT would be very
useful to help determine whether you would re-treat
the patient at that visit.”
Michael S. Ip, MD
When selecting an OCT, popular choices include time-domain and spectral-domain instruments. In terms of morphology, spectral-domain OCTs may provide more information due to higher resolution. Nevertheless, if you already rely on time-domain OCT, the purchase of a spectral-domain instrument may not be necessary since it won’t provide that much more information for the general ophthalmologist that he or she can use in making therapeutic decisions. Volume scans may provide as much information as you need to determine treatment algorithms and strategy, and do so rapidly. In a busy practice of a busy clinician, the best evidence to date indicates that looking at a volume scan with particular emphasis on the central subfield is adequate to treat most patients.
When spectral-domain OCT is used, it may be advisable to order a macular cube scan to determine the thickness in the 9 macular subfields (as defined by the standard ETDRS grid) so the patient can be followed longitudinally and change detected. Some clinicians may order a higher resolution raster scan as well to check whether there are any vitreomacular interface abnormalities, which could affect patient management.
You may choose to scroll through a patient’s B scans yourself, knowing that it can take just seconds and feeling that you’d like to view the automated placement of the inner and outer retinal boundary lines to decide whether the information on the map should be accepted, rather than taking it at face value.
“We always do raster scans on every OCT machine. We actually do a cube as well on the Spectralis. Typically, the way my practice runs is that the photographer will print out or include in the medical record those scans that show fluid. And if I don’t see fluid and I think that there might be, I examine the cube myself.
“When I look at scans, I don’t just look at the
thickness numbers. I like to see what the morphology
is outside the fovea. I want to know if there is
non-clinically significant macular edema, and if it
leads me to see the patient more frequently.”
Sophie J. Bakri, MD
“At viewing stations which are available to me in
every patient management room, I’ll scroll through
however many B scans were obtained (cube and
raster). It can be as few as 19 and as many as
128 in the cube alone.”
Susan B. Bressler, MD
Color Fundus Photography / Fluorescein
In the pre-OCT era, many ophthalmologic practices used stereoscopic color fundus photographs more often than they do today. In fact, fundus photographs have been supplanted by OCT in many practices. Quality of stereoscopic photographs can be quite variable, and detection of retinal thickening is highly dependent on good quality images.
Many clinicians order color photographs only at baseline and only in certain patients. Proponents of color fundus photography also point out that certain morphologic features that can be seen on color photographs are not easily visible on OCT (including hard exudates).
Like color fundus photography, fluorescein angiography (FA) has a more limited role now than in the past, thanks to the availability of OCT. FA takes more time than fundus photography, and is more invasive. It can also reduce efficiency in the clinic if you are in a high volume fluorescein center. But still, it may make sense in some circumstances. In current times, in patients who have reduced vision but no significant edema, color photographs and angiography at baseline might be able to identify features suggestive of macular ischemia.
“If you plan to treat a patient with macular photocoagulation, I think that FA might be able to help guide your treatment. Some research suggests that fluorescein angiographic guided laser treatment of diabetic macular edema may be superior to treatment that does not use FA. So in rare cases, I may use it. The treatable lesions on FA are leaking microaneurysms, areas of diffuse leakage and areas of non-perfusion in the macula. And fluorescein is good at identifying those. But overall, I’ve really reduced my use of FA, and am very judicious in its use for DME.
“Overall, the disadvantages of fluorescein
angiography usually outweigh any of the advantages.”
Michael S. Ip, MD
Despite a decrease in utilization of fluorescein angiography in DME, FA is still chosen by some ophthalmologists for patients with macula edema, reduced visual acuity, and who are likely to be treated with pharmacologic intervention. A baseline FA in these patients will let you know up front whether there is ischemia as a contributing factor in the posterior pole. But overall, in the era of OCT, fewer fluorescein angiographies are being done now than in previous times.
“I am more comfortable with a fluorescein angiogram
at presentation before I start treating DME as it
facilitates my decision-making process as to which
treatment I may adopt initially. And if I elect to treat
the patient with laser alone in a non-center-involved
case, or laser as an adjuvant to pharmacologic
therapy, I’m going to also use that fluorescein to
guide laser treatment.”
Susan B. Bressler, MD
A recent DRCR Network publication looked for correlations among a variety of features on fluorescein angiography in patients with DME – specifically, how well they correlated with levels of acuity or various parameters on OCT, or whether they were predictive of how vision would change over time. The researchers concluded that FA wasn’t particularly useful, especially in comparison with what OCT can tell you about the same parameters. There’s a stronger correlation between OCT thickness and the presenting level of visual acuity, and a correlation between OCT change over time and how vision will change as well, as compared to relationships between fluorescein angiography and vision.
DME THERAPIES —
CONSIDERING THE OPTIONS
Focal/grid photocoagulation remains a good initial treatment option for many patients with DME, and may often still be considered the gold standard by some. It is safe, durable, and has been proven effective in many clinical trials. There are certainly patients with non-center-involved DME who respond very well to focal/grid laser alone. Some ophthalmologists believe that its main drawbacks in this age of pharmacotherapy are that the initial “wow effect” associated with many medications is absent and that a small proportion (but higher than eyes managed with pharmacotherapy) of patients managed with laser experience vision decline despite treatment.
“If you have a patient who is willing to have some patience, laser can often be a good option. I would argue that among patients who have center-involved diabetic macular edema and who have relatively good levels of visual acuity, focal/grid laser photocoagulation is still a very good initial choice.
“Nearly 100% of my patients with DME receive
some sort of laser – whether it’s at initial presentation
or some time during follow-up. So I definitely haven’t
thrown away my laser in the clinic.”
Michael S. Ip, MD
In terms of the OCT’s ability to demarcate areas of retinal thickening, some clinicians use the OCT as a road map to illustrate areas of retinal thickening to direct their laser treatment only to those regions. Some ophthalmologists feel comfortable applying laser exclusively to areas of clinically detectable thickening viewed during the procedure while a contact lens is on the eye, while others rely on topographic OCT maps to provide information on where most of the thickening can be found and addressed during laser treatment.
The corticosteroid triamcinolone acetonide as monotherapy was studied in a DRCR Network clinical trial, where it was compared with focal/grid photo-coagulation. In that trial, 840 study eyes (693 subjects) with DME were randomized to 1 mg or 4 mg of preservative-free intravitreal triamcinolone vs focal/ grid photocoagulation. Retreatment was administered for new or persistent edema at 4-month periods.2
At 4 months, the mean visual acuity was better in the eyes managed with 4 mg triamcinolone than in the 1 mg triamcinolone group or in the laser cohort. But by 1 year, no significant differences in visual acuity existed among the groups. At 16 months and continuing to 3 years, those in the laser group showed better mean visual acuity than the other cohorts. Not only was the use of the laser more effective, but it had a better safety profile, particularly with respect to the development of cataracts and glaucoma.2
(See Figures 1 and 2 for depictions of changes in visual acuity and central subfield thickening in the DRCR.net trial described above).
A subset analysis of pseudophakic eyes (n=273) in Protocol I evaluated patients who had received intravitreal triamcinolone along with prompt focal/grid laser photocoagulation, and found that they did as well as 2 ranibizumab groups in terms of achieving significant visual acuity benefits and reducing retinal thickening. Furthermore, this was accomplished with significantly fewer intravitreal injections.3 As a result, many clinicians now think that intravitreal triamcinolone should be considered only in pseudophakic eyes and only in conjunction with prompt laser treatment.
Patients who have an aversion to a large number of intravitreal injections, either because they don’t tolerate the injections or because they live far away and it’s hard for them to come in to receive frequent injections, may be candidates for this regimen. However, frequent monitoring and management of steroid-associated complications may be needed.
“I believe that the use of intravitreal triamcinolone in DME should be limited to combination therapy with focal/grid laser photocoagulation and in general only in pseudophakic eyes. I also would not use it for eyes with a history of open-angle glaucoma or steroid-response glaucoma.
“Some clinicians use a compounded formulation
of triamcinolone, or they turn to the trade
formulation of triamcinolone (Triesence) even
though its labeling is not for DME.”
Michael S. Ip, MD
“In my state (Florida), it’s hard to get reimbursed
for the trade version of triamcinolone. At the same
time, it’s hard to justify paying 8 to 10 times more
for Triesence compared to pharmacy-compounded
versions of triamcinolone, when there’s no increased
efficacy. The standard care in my community is
to use off-label compounded preservative-free
triamcinolone. One would have to prove to me
increased benefit to justify the increased cost of
Scott M. Friedman, MD
Dexamethasone (Ozurdex) is another steroid preparation with anti-inflammatory properties since it interferes with inflammatory cytokines. It is sometimes used off-label as monotherapy in the treatment of DME. Typically, dexamethasone is chosen for patients who have responded poorly to laser or anti-VEGF treatment. These individuals may have a history of well managed glaucoma or ocular hypertension, or they may be phakic, as the safety profile reported to date suggests fewer steroid-associated complications of cataract and glaucoma relative to triamcinolone.
Even so, the evidence is limited in this area. Ozurdex is a drug that some clinicians keep in their armamentarium for cases where patients have failed other therapies. But reimbursement has been a challenge as it is not approved for treatment of DME.
“My experience with dexamethasone (Ozurdex) has
been limited because it is not reimbursed. But I
would like to see a head-to-head trial of Ozurdex
versus ranibizumab, although it probably isn’t going
to happen soon. I would also like to see combination
therapies of Ozurdex plus anti-VEGF versus Ozurdex
alone or anti-VEGF alone. With the government moving
toward increased funding for comparative trials,
perhaps that’s something we will get in the future.”
Sophie J. Bakri, MD
Fluocinolone acetonide is an injectable steroid implant. Although it is not yet available for DME, it is anticipated that there will be some use for this agent in eyes with DME even though there is a scarcity of comparative efficacy research which may help define the precise role of this drug.
Two phase 3, randomized, double-masked trials, collectively known as FAME (Fluocinolone Acetonide in Diabetic Macular Edema), were conducted to evaluate the efficacy and safety of fluocinolone. Recent presentations have included the following:
- The 2 studies enrolled 953 subjects who had DME involving the fovea, with central retinal thickness of ≥250 µm on OCT; all individuals had undergone prior laser treatment, and their visual acuity ranged from 19 to 68 letters (approximate Snellen equivalent 20/40-20/400) on the ETDRS chart. They were randomized in a 2:2:1 ratio to receive low-dose fluocinolone inserts, high-dose drug inserts, or placebo.4
- After 24 months, 29% in each group of patients allocated to fluocinolone showed a gain in visual acuity of at least 15 letters from baseline, compared with 16% of patients in the placebo group.
- The fluocinolone subjects showed an acceleration of the formation of cataracts. At 24 months, about 75% of initially phakic patients had undergone cataract surgery during follow-up. The treatment groups experienced an increase in intraocular pressure, and trabeculectomy occurred in 2.1% and 5.1% in the low-dose and high-dose treatments, respectively.4
Vascular endothelial growth factor (VEGF) antagonists have received the attention of retinal specialists and researchers in recent years, not only for neovascular age-related macular degeneration (AMD) but also for DME.
One of the available anti-VEGF agents, ranibizumab, is a recombinant, humanized, monoclonal antibody fragment that inhibits VEGF, and has been increasingly studied and used for DME. Although there are scientific data to support its use for this disease, it is problematic in some states where it is not a reimbursable treatment.
The Diabetic Retinopathy Clinical Research Network has reported that when ranibizumab plus prompt or deferred laser therapy was given to patients with center-involved DME, their vision fared better than those receiving laser alone. The researchers reported:
• 854 study eyes (691 patients) were part of the study. The participants were diagnosed with type 1 or 2 diabetes and had compromised vision (Snellen equivalent 20/32-20/320, mean approximately 20/64) from center-involved DME. Stratus OCT confirmed a central subfield thickness of ≥250 µm. Participants were randomized to one of 4 groups:
- Sham injections plus prompt laser treatment
- 0.5 mg of ranibizumab plus prompt laser
- 0.5 mg of ranibizumab plus deferred laser, or
- 4 mg triamcinolone injections plus prompt laser
• At 1 year, the average change in visual acuity letter score relative to baseline was significantly greater in the ranibizumab plus prompt laser (+9 ± 11, P<0.001), and the ranibizumab plus deferred laser (+9 ± 12, P<0.001) groups, compared with the sham plus prompt laser (+3 ± 13) group. In addition, rates of moderate vision loss (≥ 3 lines) were lower in the ranibizumab groups relative to the laser group (2%, 2%, and 15%). The median number of ranibizumab treatments was 8 in the ranibizumab plus prompt laser group and 9 in the ranibizumab group with deferred laser. Among patients who were pseudophakic at study entry, those receiving triamcinolone plus laser had results comparable to the 2 ranibizumab groups, but IOP increased in more patients treated with triamcinolone than ranibizumab. Ranibizumab was well tolerated in terms of ocular and systemic safety. (See Figure 3, depicting the mean change in visual acuity at follow-up, and Figure 4, showing ≥2 step improvement in the logarithmic transformation of OCT central subfield thickness from baseline.)3
The Ranibizumab for Edema of the mAcula in Diabetes (READ) study provided another look at the effects of ranibizumab in patients with diabetic macular edema and vision compromise in the range of 20/40 to 20/320 (mean 20/80). This prospective, randomized, interventional trial enrolled 126 patients, and randomized them 1:1:1 to one of the following:
• Injections of ranibizumab 0.5 mg at baseline and months 1, 3 and 5 (group 1)
• Focal/grid laser photocoagulation at baseline and, if needed, at month 3 (group 2)
• A combination of ranibizumab 0.5 mg followed by focal/grid laser at baseline and month 3 (group 3).
At month 6 and thereafter, all eyes could receive ranibizumab as often as every other month. At the 24-month visit, the mean improvement in bestcorrected visual acuity (BCVA) was 7.7 letters, 5.1 and 6.8 letters in the 3 groups, respectively. The percentage of patients with 20/40 or better Snellen equivalent at 24 months was 45%, 44% and 35% in the 3 groups. Thus, the injections of ranibizumab produced patient benefits for at least 24 months. When ranibizumab was combined with focal/grid laser, the amount of residual edema at month 24 appeared less than in the group that did not receive laser, and fewer injections of ranibizumab were administered, relative to the ranibizumab only group.
Bevacizumab is another anti-VEGF agent that is a humanized monoclonal antibody, and it may be an alternative to ranibizumab. Though there are not as much data for bevacizumab as there are for ranibizumab, level 2 and 3 evidence appears strong for its use in DME. The body of data at this point suggests that both ranibizumab and bevacizumab may be useful in DME. A large scale prospective study comparing the 2 agents in DME has not been performed.
“In patients with center-involved diabetic macular edema and decreased vision, you may consider treating with ranibizumab (if it is a covered service in your community outside of clinical trials). If there are financial barriers to the use of ranibizumab, you might turn to bevacizumab instead.
“Outside of clinical trials, if patients have good visual acuity (greater than 20/32) and the center is involved, I treat with anti-VEGF therapy because I think the benefits greatly outweigh the risks, including cost and the low chance of side effects. It’s better to preserve visual acuity than wait until the visual acuity deteriorates and try to improve it.
“At the same time, with all injections, including
anti-VEGFs, we have to keep in mind the morbidity
of giving an injection to a patient with 20/20 vision,
because of the small but real chance of severe side
effects such as endophthalmitis.”
Scott M. Friedman, MD
In the DRCR Network’s phase II trial of bevacizumab, 121 DME patients were randomized to one of 5 treatment arms:
- Focal photocoagulation at baseline (group A)
- Intravitreal injection of 1.25 mg of bevacizumab at baseline and at week 6 (group B)
- Intravitreal injection of 2.5 mg of bevacizumab at baseline and at week 6 (group C)
- Intravitreal injection of 1.25 mg of bevacizumab at baseline, plus a sham injection at week 6 (group D)
- Intravitreal injection of 1.25 mg of bevacizumab at baseline and at week 6, plus photocoagulation at week 3 (group E).
Researchers reported that bevacizumab can reduce DME in some eyes. Although the study was not designed to demonstrate whether bevacizumab has benefit for DME, the data showed that groups B and C had a greater decline in central subfield thickness at 3 weeks, compared with group A. At 12 weeks, there was about a 1-line better median visual acuity in these groups as well, relative to laser only. The combination of bevacizumab and focal photocoagulation produced no short-term benefit.7
Another trial, the 2-year BOLT study, supported the use of intravitreal bevacizumab in clinically significant diabetic macular edema. Eighty patients were randomized to receive either bevacizumab or laser therapy. The bevacizumab group gained a median of 8 ETDRS letters over 12 months, compared with a loss of 0.5 ETDRS letters in the laser group (P = 0.0002).8
“At this time, I substitute bevacizumab for
ranibizumab. One important reason for doing so is
that in my state (Wisconsin), ranibizumab is not
covered by third-party payors for DME. So unless
patients want to pay personally for the drug, it is not
readily available for them. At the same time, I’m very
comfortable using bevacizumab off-label. I tell
patients that there’s a lot of evidence to support its
use. It’s generally a community standard. But we
really need a randomized controlled trial comparing
these 2 drugs for DME.”
Michael S. Ip, MD
Using Anti-VEGF Therapy: Other Factors
When choosing anti-VEGF therapy, should you start it as monotherapy? Or should you combine it immediately with prompt focal/grid laser or some other therapy? Later, should there be a switch to a different treatment?
There are differing opinions on the answer to these questions.
“Personally, I would give 3 or 4 injections, and then
start the laser, when there is much less edema and I
can use a lower-powered laser, and I have gained as
much visual acuity as possible initially and I know the
patient’s potential. I think that after about 4 injections,
the patient may become more reluctant to keep going
with this, and the clinician sometimes has to offer
another treatment. So I’ll go for visual acuity initially
with ranibizumab, and then see if it will last longer. I
would monitor the patient with OCT. And if the patient
doesn’t respond to laser, then I would continue the
patient on monthly ranibizumab.”
Sophie J. Bakri, MD
“I always use anti-VEGF therapy initially. And
depending on the clinical circumstances – if patients
have predominantly focal edema with significant
retinopathy including hard exudate with localized
aneurysms – I think those patients are highly
amenable to adding treatment with focal/grid laser.
If there’s more diffuse thickening, it’s probably less
beneficial to add focal/grid laser.”
Scott M. Friedman, MD
“I prefer to start with ranibizumab monotherapy and
defer the laser. In Protocol I when following participants
through 2 years, fewer patients assigned to the
ranibizumab plus deferred laser groups ever got
exposed to laser and they only required one additional
injection in year 1 and one additional injection in year 2,
relative to the group that received the combination of
ranibizumab plus prompt laser. Had there been a more
substantial difference in the number of injections
required when using it as monotherapy with deferred
laser, I might feel otherwise. I think avoiding the
morbidity of laser, since it didn’t result in fewer
injections or superior vision results, is a good idea.”
Susan B. Bressler, MD
If you look at the safety profile of anti-VEGF agents, you would find no increased incidence of systemic complications compared with laser when treating DME. For example, in clinical trials, there is no increased risk of significant cardiovascular complications in patients receiving ranibizumab as compared to patients receiving sham injections.
“Protocol I has about 400 individuals who have
been ranibizumab-exposed and followed for 2 years.
Trials like RESTORE (A Randomized, Double-Masked,
Multicenter, Laser-Controlled Phase 3 Study Assessing
the Efficacy and Safety of Ranibizumab) and RESOLVE
(Randomized, Double-Masked, Multicenter, Phase 2
Study Assessing the Safety and Efficacy of Two
Concentrations of Ranibizumab) have another 200 to
300, and READ-2 has a little over 100. So with about
700 patients total exposed, nothing has raised our
pulse about systemic complications. That, of course,
doesn’t rule out that if thousands of patients were
followed, you might find a small but statistically
significant difference in some systemic complications.
But we’ll probably never be in a position to have
such data at our fingertips.”
Susan B. Bressler, MD
For that reason, you might feel comfortable using these agents in diabetic patients, with few limitations. Some clinicians will restrict treatment in patients who have had a stroke or heart attack in recent weeks, or in patients who have had recent surgery or are anticipating surgery. In these patients in particular, you may consider initiating or resuming treatment after a period of recovery.
“In patients who have bilateral DME, it would be extremely rare that I would manage both eyes in a diabetic patient with intravitreal therapy. With the number of treatment options available, we can perhaps do laser in one eye and injection in the other. However, if they insist on having bilateral injections, then I will discuss the risks with them.
“In patients who have bilateral DME for whom
you want to start injection treatment in both eyes,
I advise them to have each eye injected on separate
days. But if we needed to inject both eyes, I would
do both on the same day if there were extenuating
Sophie J. Bakri, MD
When patients travel long distances for treatment, you may decide to offer bilateral injections on the same day. However, 100% reimbursement is not available for bilateral injections performed on the same day.
“I have a lot of rural patients who travel long
distances to get to their appointments with me. That
may be okay in the summer but not in the winters in
Minnesota, so these are real considerations in rural
areas. I want to offer the patient the best treatment,
but I also try to minimize the treatment burden. When
caregivers and patients are saying this is too much
for both them and the elderly patients, then I think
you have to start negotiating a compromise.”
Sophie J. Bakri, MD
What About Vitrectomy?
There is consensus that if you have a patient with diabetic macular edema without incontrovertible evidence of vitreomacular traction, you should not offer straight to vitrectomy as a first-line therapy. The eyes that potentially benefit from vitrectomy are those with significant vitreomacular traction. Some clinicians use vitrectomy as a last resort when eyes haven’t responded to multiple intravitreal injections as well as one would hope.
The role of vitrectomy in DME needs to be elucidated further.
Once treatment has been initiated with anti-VEGF therapy, retreatment with the same agent following Protocol I retreatment algorithms is typically done. This protocol requires monthly visits throughout the first year during which it uses the 4-2-7 rule. Initially, patients are given an injection monthly for 4 consecutive months, regardless of functional or anatomic status at each of these visits.
Thereafter, therapy is based on retinal thickness and vision parameters, as well as change from the last visit. If patients are deemed to meet “success” criteria, meaning an OCT <250 µm central subfield thickness or vision equivalent of 20/20, you may withhold treatment as early as the fifth month exam. Otherwise, patients generally receive 2 more injections during the fifth and sixth month of management.
Treatments in months 7 through 12 are based on continued improvement. If treatment results in success criteria at any point between 7 and 12 months, treatment can be deferred. However, if success criteria are not met and patients are showing continued improvement, then another injection is given. Continued improvement means a 5-letter increase in acuity or a 10% decrease in CSF thickening on OCT, relative to the last visit. If stabilization of both vision and thickness emerge then treatment may be deferred, monthly follow-up continued and treatment resumed if retinal thickening recurs or vision deteriorates.
In the second year of the protocol I, retreatment continues as long as successive improvement occurs at each visit. Once stabilization or success criteria are met, the frequency of visits falls to every 2 months and then to every 4 months if the eye continues to do well. However, if vision falls or edema increases, then treatment and monthly follow-up resume.
How would you respond if in following the retreatment protocol, the patient actually experiences vision loss, even if the OCT is progressively getting better (while still showing thickening)? Would you continue to treat in this setting of worsening vision, or are you going to conclude that this approach just isn’t working and it’s time to try something else?
“I would continue to re-treat these patients based on
the treatment guidelines from DRCR Protocol I until
they reach futility or failure. I think you need to push
the treatment until you have definite evidence that
they aren’t improving. Yet, especially if they’ve
already had maximal laser, there’s really not a whole
lot else that you’re left with at this point. So push the
anti-VEGF treatment until you’re very sure that there’s
not going to be any more improvement or potential
Scott M. Friedman, MD
In patients in this situation, you should continue monitoring, particularly since in terms of diabetic edema injections, they are relatively new and it is imperative that the patients be followed closely. In the opinion of many clinicians, patients with clinically treated macular edema that’s relatively stable should continue to be evaluated every 4 to 6 months, and should be followed for the rest of their lives.
With emphasis on regular monitoring, the patient volume in many clinics is expected to increase. Your own office staff may need to become more efficient, and there may be longer waiting times for your patients. Even though you may have an increased patient load, it is anticipated that DME is different than neovascular AMD in that the DME patients probably won’t be getting injections for their entire lives. In the DRCR.net trial evaluating ranibizumab, the number of injections per year decreased from a median of 8 to 9 in the first year to 2 to 3 in the second year.
Considering Cost and Patient Issues
In your discussions of therapeutic options with patients, cost may surface and resurface as a factor in the decision-making process. If a particular treatment is appropriate, but it is expensive and not covered by third-party payors, you should include it in your discussions with patients. However, as described earlier, patients may decline costly treatments if they cannot afford them.
Cost-effectiveness is another issue to consider, but we don’t have solid data yet on which therapeutic options are the most cost-effective.
Treatment burden should also be considered in patient discussions and in therapy selection.
Diabetic macular edema is a serious problem among many Americans, and as the number of patients with diabetes increases, so too will the prevalence of DME in all likelihood. Some patients have incomplete responses to available treatments, and thus newer approaches continue to be studied. A number of the therapies for the treatment of DME described in this monograph — from anti-VEGF injections to triamcinolone to dexamethasone — may dramatically improve the lives of many people. With a growing number of options for the management of DME, higher rates of success than more traditional treatments may be on the horizon.