Diabetic Retinopathy Treatment: Are the Standards Changing?

An overview of the Diabetic Retinopathy Clinical Research Network's studies of DME.

ALI A. ZAIDI, MD ∙ ALEXANDER J. BRUCKER, MD

The management of diabetic retinopathy remains a challenge. Diabetic macular edema (Figure 1) is a major cause of vision loss. The Early Treatment of Diabetic Retinopathy Study (ETDRS) showed a 50% reduction in the risk of moderate vision loss (15 or more letters) at three years with focal laser photocoagulation.¹ However, only 16% of patients with a baseline vision of 20/40 or worse improved by 15 letters at three years. The low rate of visual improvement has fueled interest in other treatment modalities.

Figure 1. Diabetic macular edema.

The Diabetic Retinopathy Clinical Research Network (DRCR) is a research organization funded primarily through the National Eye Institute, but also with the support of other sponsors, such as the Juvenile Diabetes Research Foundation, and companies, such as Genentech (South San Francisco, CA) and Allergan (Irvine, CA). The DRCR has conducted rigorous clinical studies evaluating surgical and pharmacologic therapy for the treatment of DME. This article reviews and summarizes the various studies carried out by the DRCR to date.

The studies of the DRCR can be found on the internet by going to its Web site at www.drcr.net. The studies are presented below according to their protocol designations, which have been alphabetized and are sequential. As not all of the trials have reported data, some letters in the alphabetic sequence are missing.

PROTOCOL A — COMPARISON OF MODIFIED ETDRS AND MILD MACULAR GRID LASER PHOTOCOAGULATION STRATEGIES FOR DME

The exact mechanism by which laser photocoagulation reduces DME is not known. Direct treatment of microaneurysms may cause their closure and thereby reduce leakage. However, indirect effects of photocoagulation may also play a role, including decreased retinal blood flow, improved oxygenation, and changes in the retinal pigment epithelium (RPE).^2-4

To evaluate the importance of the indirect effects of laser photocoagulation, the DRCR conducted a randomized clinical trial comparing two different laser techniques in patients previously un treated for DME. In one group, patients received a modified ETDRS protocol, which involved both direct treatment of microaneurysms and grid laser to areas of thickened retina only. In the second group, patients received mild macular grid (MMG) protocol, which involved more diffuse treatment distributed throughout the macula in both thickened and unthickened retina without direct treatment of microaneurysms. Theoretically, this approach could avoid the complications of treating microaneurysms close to the fovea.

The primary outcome measurement of this study was changes in optical coherence tomography (OCT) measurements and the secondary outcome measurement was change in visual acuity. After 12 months of follow-up, eyes treated with the modified ETDRS protocol had slightly greater reduction in retinal thickening and a trend toward better visual acuity outcome than those treated with MMG. The authors concluded that a larger long-term trial would be unlikely to show benefit of MMG over the modified ETDRS treatment.

PROTOCOL B — A RANDOMIZED TRIAL COMPARING INTRAVITREAL TRIAMCINOLONE TO FOCAL/GRID PHOTOCOAGULATION FOR DME

Macular edema is caused by the breakdown of the blood-retinal barrier, which may be mediated by multiple factors, including ischemia-induced release of vascular endothelial growth factor (VEGF) and retinal leukostasis causing inflammation.⁵ The rationale for the use of corticosteroids to treat DME stems from observations that corticosteroids may inhibit both VEGF and inflammation.⁶

To evaluate the efficacy of corticosteroids, the DRCR conducted a phase 3, randomized, controlled clinical trial comparing laser photocoagulation to intravitreal triamcinolone acetonide (1 mg vs 4 mg) for the treatment of DME.⁷ The main outcome measurements were visual acuity and retinal thickness as measured by OCT. At four months, patients receiving intravitreal triamcinolone had better visual acuity compared to laser photocoagulation.

However, these findings were reversed by 16 months, and laser photocoagulation remained superior to triamcinolone at three-year follow-up (Figure 2). Twenty-six percent of patients treated with laser photocoagulation gained 15 or more letters of vision at three years, compared to 21% of patients treated with triamcinolone. Conversely, only 8% of patients treated with laser photocoagulation lost 15 or more letters of vision, compared to 16% of patients treated with triamcinolone.

Figure 2. Protocol B: Visual acuity results in laser-treated vs triamcinolone-treated eyes.

The inferior effect of triamcinolone on visual acuity could be not attributed to steroid-induced cataract, because subgroup analysis of patients who were pseudophakic at baseline showed similar results. Further, the retinal thickness findings paralleled visual acuity findings. There was no difference in the main outcomes between the two doses of triamcinolone. At three years, patients in the 4 mg triamcinolone group had a much higher incidence of cataract surgery (83% vs 31%) and intraocular pressure rise of 10 mm Hg or more (31% vs 3%) compared to laser photocoagulation.

One subgroup of patients, those with the most severe visual acuity loss at baseline (20/200 to 20/320), had superior visual acuity gains in response to 4 mg of triamcinolone compared to laser photocoagulation. However, the number of patients in this subgroup was too small to draw meaningful conclusions. The authors concluded that laser photocoagulation in this patient population is the most effective evidence-based treatment for DME.

A recent exploratory study using data from this trial compared the effect of triamcinolone vs laser photocoagulation on the rate of progression of diabetic retinopathy.⁸ The main outcome was progression to proliferative diabetic retinopathy or worsening of two or more severity levels of diabetic retinopathy based upon fundus photographs. This study found that patients receiving 4 mg of intravitreal triamcinolone had a reduced risk of progression of diabetic retinopathy at three years. However, due to the high incidence of adverse effects, such as cataract formation and elevated IOP, the authors concluded that the use of triamcinolone solely for reducing the progression of diabetic retinopathy is not warranted.

PROTOCOL D — EVALUATION OF VITRECTOMY FOR DME

The vitreous may contribute to DME through various mechanisms. The posterior hyaloid may cause vitreomacular traction. Removal of the vitreous may not only remove tractional forces, but may also improve retinal oxygenation and remove harmful growth factors.^9-11

The DRCR has completed a prospective cohort study evaluating the use of vitrectomy for the treatment of DME. Patients were included if the presence of vitreomacular traction was associated with DME or if the investigator felt that the DME would not respond to laser treatment. The main outcomes measured were visual acuity, retinal thickening on OCT, and complications. The results of this study have not yet been published, but Julia A. Haller, MD, presented the findings of a cohort of 87 patients at the American Academy of Ophthalmology meeting in 2009.¹²

All patients in this cohort had vitreomacular interface abnormalities. Half underwent internal limiting membrane removal and 61% had epiretinal membrane removal. Sixty-four percent of patients received steroids (intravitreal or peribulbar) at the end of the surgery. None of the patients underwent cataract removal.

At six months, 37% of patients improved >10 letters of visual acuity. However, 23% of patients had a decrease in >10 letters of visual acuity. In terms of retinal thickening on OCT, there was a mean decrease of 153 µm in central retinal thickness (Figure 3), and 42% of patients had resolution of central DME.

Figure 3. Protocol D: Changes in central retinal thickness.

The most frequent surgical complications were elevated intraocular pressure (8%), vitreous hemorrhage (6%), and retinal detachment (3%). The study was limited by lack of controls, inclusion of patients based upon investigator discretion rather than a reading center, variable surgical techniques, and the confounding effect of steroids. One-year follow-up of the entire cohort (n=227) may provide additional information, including subgroup analysis of which patients benefited the most and which surgical interventions are most important.

PROTOCOL E — A RANDOMIZED TRIAL OF PERIBULBAR TRIAMCINOLONE ACETONIDE WITH AND WITHOUT FOCAL PHOTOCOAGULATION FOR MILD DME: A PILOT STUDY

A phase 2, randomized, controlled clinical trial evaluated peribulbar triamcinolone acetonide with and without laser photocoagulation for the treatment of mild DME.¹³ The study compared anterior subtenon's injection (20 mg) of triamcinolone, posterior subtenon's injection (40 mg) of triamcinolone, laser photocoagulation, and combination therapy. The main outcomes measurements were visual acuity and retinal thickness as measured by OCT. At 24 weeks, there were no significant differences in the main outcomes between the various groups (Figure 4). The authors concluded that peribulbar steroids are unlikely to be of benefit in the treatment of DME.

Figure 4. Protocol E: Changes in central retinal thickness. No significant differences remain at 24 weeks.

Completed Steroid Studies in the DRCR

These clinical trials evaluating the use of corticosteroids to treat DME have provided valuable insights. They have illustrated the importance of long-term follow-up. Although patients receiving intravitreal triamcinolone had superior visual acuity and retinal thickness results initially, these findings no longer held true with longer follow-up. These results have dampened enthusiasm for the use of triamcinolone acetonide for the treatment of DME.

However, the DRCR studies also have limitations. It is possible that the dosages and dosing frequency of triamcinolone were not optimal. More aggressive therapy, however, is also likely to result in more adverse effects. It is also possible that some subgroups of patients, such as those with severe vision loss (20/200 to 20/320), may benefit from triamcinolone. This topic could not be adequately addressed by the DRCR studies due to inadequate numbers of patients with severe vision loss.

These studies also do not address the management of patients with DME refractory to laser photocoagulation. For example, in the intravitreal triamcinolone vs laser photocoagulation study, 29% of patients in the laser group lost vision at two years. The next step in the management of these patients has not been established. This paper reviews only the DRCR studies. There are several other large, ongoing clinical trials that are using steroids in a variety of forms for the treatment of DME. The results of these formulations and deliveries may show efficacy in the treatment of diabetic macular edema.

PROTOCOL H — A PHASE 2 RANDOMIZED CLINICAL TRIAL OF INTRAVITREAL BEVACIZUMAB FOR DME

Given the likely role of VEGF in the pathophysiology of DME, anti-VEGF agents may be of benefit in patients with DME. The DRCR network conducted a phase 2, randomized, controlled clinical trial evaluating the short-terms effects of bevacizumab (Avastin, Genentech) for the treatment of DME.¹⁴ This study compared focal laser alone vs intravitreal bevacizumab (1.25 mg vs 2.5 mg) vs combination therapy. Patients were assigned to one of five groups: focal laser at baseline; 1.25 mg intravitreal bevacizumab at baseline and six weeks; 2.5 mg intravitreal bevacizumab at baseline and six weeks; 1.25 mg intravitreal bevacizumab at baseline and sham injection at six weeks; and 1.25 mg intravitreal bevacizumab at baseline, focal laser at three weeks, and 1.25 mg intravitreal bevacizumab at six weeks. The main outcome measurements were visual acuity and retinal thickness as measured by OCT.

Patients receiving bevacizumab (1.25 mg or 2.5 mg) achieved an average one line greater improvement relative to focal laser at 12 weeks. Patients receiving intravitreal bevacizumab initially showed a greater reduction in central retinal thickness compared to focal laser, but no significant differences in retinal thickness were observed beyond three weeks. The reduction in retinal thickness appeared to plateau after three weeks, suggesting that the six-week interval between intravitreal injections used in the protocol may have been too long.

Only half of the eyes receiving bevacizumab showed a positive response (>11% reduction in central retinal thickness). No difference in retinal thickness or visual acuity was observed between the two doses of bevacizumab. No additional benefit was seen in patients receiving combination therapy of intravitreal bevacizumab followed by focal laser. This study was limited by the short-term follow-up (12 weeks) and small sample size (109 patients). These results could not be generalized to conclude that there is a clinically meaningful benefit in treating DME with intravitreal bevacizumab.

PROTOCOL I — INTRAVITREAL RANIBIZUMAB OR TRIAMCINOLONE ACETONIDE IN COMBINATION WITH LASER PHOTOCOAGULATION FOR DME

The DRCR is currently conducting a randomized controlled clinical trial evaluating the use of intravitreal ranibizumab (Lucentis, Genentech) for DME. Patients are randomized into one of four groups: laser photocoagulation and sham injection; intravitreal ranibizumab and immediate laser photocoagulation; intravitreal ranibizumab and deferred laser photocoagulation; or intravitreal triamcinolone and immediate laser photocoagulation. Randomization of patients into this trial is complete and the results of the trial should be published in the near future.

PROTOCOL J — INTRAVITREAL RANIBIZUMAB OR TRIAMCINOLONE ACETONIDE AS ADJUNCTIVE TREATMENT TO PANRETINAL PHOTOCOAGULATION FOR PROLIFERATIVE DIABETIC RETINOPATHY

Another study is evaluating intravitreal ranibizumab or triamcinolone acetonide as adjunctive treatments to panretinal photocoagulation and focal treatment in proliferative diabetic retinopathy eyes that also have diabetic macular edema. Patients are randomized into one of three groups: intravitreal ranibizumab at baseline and four weeks; intravitreal triamcinolone at baseline and a sham injection at four weeks; or sham injections at baseline and four weeks. All patients will receive focal laser shortly after the randomized injection, followed by panretinal photocoagulation within 14 days of the initial injection. The results of these studies should elucidate the role of ranibizumab in the treatment of diabetic retinopathy. Randomization of patients into this trial is complete and the results of this trial should be published in the fall of 2010.

CONCLUSION

The clinical studies conducted by the DRCR network have greatly added to our knowledge on the management of DME. These studies have re-established laser photocoagulation as the gold standard for the treatment of DME. Triamcinolone acetonide appears to have a limited role, if any. Ongoing studies will establish the role of anti-VEGF agents in the treatment of DME. Vitrectomy may play a role in certain subgroups of patients with DME. Future studies to be carried out by the DRCR network will continue to add to our understanding of diabetic retinopathy and new ways of treating this major course of visual impairment. RP

REFERENCES

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. Wilson DJ, Finkelstein D, Quigley HA, et al. Macular grid photocoagulation: an experimental study on the primate retina. Arch Ophthalmol. 1988;106:100-105.
3. Arnarsson A, Stefansson E. Laser treatment and the mechanism of edema reduction in branch retinal vein occlusion. Invest Ophthalmol Vis Sci. 2000;41:877-879.
4. Ogata N, Tombran-Tink J, Jo N, et al. Upregulation of pigment epithelium-derived factor after laser photocoagulation. Am J Ophthalmol. 2001;132:427-429.
5. Ciulla TA, Amadar AG, Zinman B. Diabetic retinopathy and diabetic macular edema: pathophysiology, screening, and novel therapies. Diabetes Care. 2003;26:2653-2664.
6. Nauck M, Karakiulakis G, Perruchoud A, et al. Corticosteriods inhibit the expression of the vascular endothelial growth factor gene in human vascular smooth muscle cells. Eur J Pharmacol. 1998;341:309-315.
7. Diabetic Retinopathy Clinical Research Network. Three year follow-up of a randomized trial comparing focal/grid photocoagulation and intravitreal triamcinolone for the treatment of diabetic macular edema. Arch Ophthalmol. 2009;127:245-251.
8. Diabetic Retinopathy Clinical Research Network. Exploratory analysis of diabetic retinopathy progression through 3 years in a randomized clinical trial that compares intravitreal triamcinolone acetonide with focal/grid photocoagulation. Arch Ophthalmol. 2009;127:1566-1571.
9. Lewis H, Abrams GW, Blumenkranz MS, et al. Vitrectomy for diabetic macular traction and edema associated with posterior hyaloidal traction. Ophthalmology. 1992;99:753-759.
10. Stefansson E. Ocular oxygenation and the treatment of diabetic retinopathy. Surv Ophthalmol. 2006;51:364-380.
11. Stolba U, Binder S, Gruber D, et al. Vitrectomy for persistent diffuse diabetic macular edema. Am J Ophthalmol. 2005;140:295-301.
12. Haller, JA. Vitrectomy for diabetic macular edema. Paper presented at: Annual meeting of the American Academy of Ophthalmology; October 24-27, 2009; San Francisco, CA.
13. Diabetic Retinopathy Clinical Research Network. Randomized trial of peribulbar triamcinolone acetonide with and without focal photocoagulation for mild diabetic macular edema: a pilot study. Ophthalmology. 2007;114:1190-1196.
14. Diabetic Retinopathy Clinical Research Network. A phase II randomized clinical trial of intravitreal bevacizumab for diabetic macular edema. Ophthalmology. 2007;114:1860-1867.