Controversies in Care

DME: When Should You Initiate Local Treatment?


DME: When Should You Initiate Local Treatment?


Michael Colucciello, MD, is a partner at South Jersey Eye Physicians and a clinical associate at the University of Pennsylvania/Scheie Eye Institute, Philadelphia. He is a member of the Retina Society and the American Society of Retina Specialists. He has no financial disclosures to report. Dr. Colucciello can be reached via e-mail at

Carl D. Regillo, MD, is professor of ophthalmology at Thomas Jefferson University, director of the Wills Eye Clinical Retina Research Unit, and chief of the Retina Service of Wills Eye Hospital in Philadelphia. He has no financial disclosures to report.

Michael Elman MD, is president of Elman Retina Group and assistant professor of ophthalmology at the Johns Hopkins University School of Medicine, both in Baltimore. He has no financial disclosures to report.

The results from the trials in the landmark ETDRS from the mid-1980s to the mid-1990s helped physicians reduce vision loss from DME, a leading cause of disability in working-age adults.

In an era before OCT macular imaging, the ETDRS results provided guidelines that helped physicians decide when to initiate local ocular treatment (at that time, laser photocoagulation) in any given patient.

We evaluated anatomical changes based on contact lens biomicroscopy of the macula; we administered focal macular laser photocoagulation if we noted any one of three criteria: retinal thickening 500 μm or less from the center of the foveal avascular zone; lipid exudates associated with retinal thickening 500 μm or less from the center of the foveal avascular zone; or retinal thickening at least one disc area in dimension within one disc diameter of the foveal avascular zone.

Today, we are treating DME more successfully than in the pre-OCT era, not only due to the use of intravitreal anti-VEGF agents, but also because of improved evaluation and surveillance of macular anatomy with OCT.

OCT allows much more precise evaluation of macular edema than contact lens biomicroscopy, and with much better sensitivity. We can detect subtle edema; we can characterize where cystic edema exists and can precisely measure its location in relation to the fovea. We can also detect progression or improvement compared to previous exams with high precision.

Because we no longer use the signs consistent with the ETDRS term “clinically significant macular edema” primarily to indicate initiation of local (intravitreal pharmacologic, laser, vitrectomy peeling of taut vitreous cortex) treatment of DME, what anatomical changes should trigger treatment initiation in the OCT era? Also, does visual acuity matter?

In this issue, we are most fortunate to have the insights of Carl Regillo, MD, and Michael Elman, MD.


Figure 1. The decision to initiate local treatment is clear when diffuse DME with lipid exudation is present.

Initiating Local Treatment of DME In the OCT Era


Before the era of OCT, we based treatment of DME mainly on the definitions, study results, and subsequent guidelines introduced by the ETDRS.1 Based on the results of this study, we were to consider focal laser photocoagulation for eyes with clinically significant macular edema (CSME), which is edema that is close to or involves the center of the fovea. In effect, it is DME that threatens or affects VA.

Although laser photocoagulation can improve visual acuity, it is not as effective as anti-VEGF therapy. So it made sense when it was the only proven effective therapy to try to prevent decreased vision by treating CSME outside the foveal center in patients with good or even normal VA.

Now that we have more effective therapy, does it make sense to treat nonfoveal CSME at all, or just wait until CSME starts to decrease VA to some degree before initiating therapy as was done in the various prospective clinical studies that investigated pharmacotherapy for treating DME, such as protocol I, the BOLT study, and the RIDE/RISE trials?2–4

Furthermore, what about DME that involves the fovea center but is OCT incidentally pick it up in the mild or asymptomatic phase?


Figure 2. The decision to treat becomes muddled when subtle extra- or juxtafoveal DME, discernable only with SD-OCT, exists — especially after optimal control of blood sugar, blood pressure, and cholesterol have been achieved.

Knowing the Limits

Unfortunately, limits exist regarding what we can extrapolate from the available published studies. All of the pharmacotherapy studies included patients only with center-involving DME and significantly decreased BCVA, such as 20/32, 20/40, or worse by EDTRS measurement techniques which, in our clinics, could be Snellen VA as bad as 20/50 or so.

Does it make sense to intervene earlier, such as when foveal-threatening CSME or early foveal involvement exists but with better vision?

The latter scenario is relatively common, especially now with SD-OCT, which can detect small amounts of center-involving DME before we can see it on clinical examination by even the best ophthalmoscopic techniques. In such cases, the VA may or may not be affected, and the patient may or may not have any significant symptoms.

Because DME is not typically rapidly progressive and may fluctuate or even improve with systemic metabolic control, it is may be best to observe small amounts of center-involving DME with good VA rather than commit a patient to a course of anti-VEGF therapy with its associated risks, costs, and burden.

It often takes at least several anti-VEGF treatments to improve a given level of DME, and the condition is rarely ever “cured” with one or two injections of bevacizumab (Avastin, Genentech, South San Francisco, CA) or ranibizumab (Lucentis, Genentech).

The Art of Medicine

This is where the art and science of managing patients converge because no one study can guide us in this scenario. Whether to treat small amounts of center-involving DME will depend on many individual patient factors, such as symptoms, status of the fellow eye, status of systemic metabolic control, predicted reliability for compliance with home vision monitoring and follow-up, etc.

For example, it is certainly reasonable to treat a patient with anti-VEGF therapy when the involved eye has mildly decreased VA, such as 20/25, and moderate central retinal thickness on OCT, especially if the fellow eye has poor vision from aggressive DME.

What about the patient with CSME that is non–center-involving DME and who has VA of 20/20? In this case, the OCT will show thickening only outside the foveal center. No consensus exists here, but options include observation for progression or focal laser to leaking microaneurysms per ETDRS guidelines. The latter approach offers the potential to prevent foveal involvement and the need for a subsequent course of injections.

Even today with anti-VEGF therapy, corticosteroids, and SD-OCT, laser for nonfoveal CSME is still a reasonable choice. Performed properly, it carries a very low risk of complications and the potential for adverse visual effects.


In summary, the treatment algorithm for managing CSME is evolving, and OCT is playing a large role in this paradigm shift, along with new, more effective pharmacotherapeutic options.

At this time, an amalgamation of the old laser studies and the new OCT-guided anti-VEGF and steroid studies guide our decision on whether to treat and how. For non–center-involving CSME, it comes down mainly to observation or focal laser. For center-involving CMSE (confirmed by OCT), it will be observation (for very good VA and few or no symptoms) or pharmacotherapy (anti-VEGF or, in select cases, corticosteroids).

Finally, in some circumstances, vitreomacular interface pathology, such as VMT or ERM might appear in eyes with CSME, detected or confirmed by OCT. In such cases, additional options to consider at some point in the course of therapy include ocriplasmin (Jetrea, ThromboGenics, Iselin, NJ) for VMT and vitrectomy with membrane peeling for ERM.


1. Early Treatment Diabetic Retinopathy Study Research Group. Photocoagulation for diabetic macular edema: Early Treatment Diabetic Retinopathy Study report number 1. Arch Ophthalmol. 1985;103:1796-1806.

2. Diabetic Retinopathy Clinical Research Network; Elman MJ, Aiello P, Beck RW, et al. Randomized trial evaluating ranibizumab plus prompt or deferred laser or triamcinolone plus prompt laser for diabetic macular edema. Ophthalmology. 2010;117:1064-1077.

3. Michaelides M, Kaines A, Hamilton RD, et al. A prospective randomized trial of intravitreal bevacizumab or laser therapy in the management of diabetic macular edema (BOLT study) 12 month results: report 2. Ophthalmology. 2010;117:1078-1086.

4. Nguyen QD, Brown DM, Marcus DM, et al; RISE and RIDE Research Group. Ranibizumab for diabetic macular edema: results of 2 phase III randomized trials: RISE and RIDE. Ophthalmology. 2012;119:789-801.

SD-OCT Essential in DME Decisions


Anti-VEGF treatment and the OCT have revolutionized the evaluation and treatment of DME. The results of expertly designed and executed randomized clinical trials have offered much guidance on the indications for treatment and parameters for retreatment. In these studies, both OCT and VA were key factors for initiating and assessing the response to current treatments.

The Protocol I first established the value of ranibizumab treatment for center-involving DME (>250 μm Stratus-equivalent OCT) with vision loss of 20/32 to 20/320 BCVA on an ETDRS equivalent vision chart.

Furthermore, the study showed the physician could safely defer initial focal/grid laser treatment that inadequately responded to ranibizumab alone. Moreover, at three years, focal/ grid laser treatment at the initiation of intravitreal ranibizumab was no better, and possibly worse, than deferral of laser treatment for 24 weeks or more in eyes with DME involving the fovea with vision impairment.

Using the algorithm, OCT together with VA guide the clinical decision process for the initiation, continuation, resumption, and cessation of treatment. Five-year follow-up data in this study, due to be published later this year, will report on the safety and efficacy of ranibizumab treatment with and without focal/grid laser and address the effects of delaying treatment with ranibizumab.

This raises the question: Can eyes treated with laser alone for two years “catch up” with the addition of ranibizumab?

Data From Trials

In the RIDE and RISE studies, eyes initially assigned to sham treatment could cross over to ranibizumab at the 24-month follow-up. Although the macular edema rapidly responded to ranibizumab in these cases, as shown on OCT, VA did not improve to the same degree seen in eyes initially treated with ranibizumab.

This finding suggests that delaying ranibizumab treatment for two years can adversely affect visual recovery once treatment begins, despite the anatomic improvement seen on OCT. In the RIDE and RISE studies, two years seemed too long to start anti-VEGF treatment.

The’s Protocol V will determine the best approach for eyes with center-involved DME on SD-OCT and very good vision (≥20/25). The protocol will randomize eyes to prompt focal/grid laser and deferred aflibercept, prompt aflibercept, or observation and deferred aflibercept.

For eyes with non-central DME on OCT, laser remains an excellent choice, based on the ETDRS experience. While clinicians increasingly administer anti-VEGF treatment in these cases, the definitive treatment remains unknown.

Not all macular thickening in diabetic eyes can be ascribed to DME alone. Using OCT, we can clearly detect VMT, as well as tangential traction from ERMs. The edema in these eyes may be more amenable to vitrectomy with membrane peeling, particular if unresponsive to pharmacologic treatment.

OCT Remains Essential

Spectral-domain OCT may be the single most important test available today, providing critical anatomic structural details and quantification of treatment response. The total clinical picture dictates the proper treatment approach in each patient. In addition to OCT, this includes careful clinical examination, VA, and fluorescein angiography. RP