Since a report described the use of intravitreal triamcinolone acetonide (TA) for the treatment of diabetic macular edema in 2001,1 retinal physicians have used corticosteroids for DME with considerable frequency. However, until recently, no randomized controlled trials have compared their use to the gold standard: focal/grid laser photocoagulation as described in the Early Treatment Diabetic Retinopathy Study.2
In the ETDRS, laser photocoagulation reduced the three-year incidence of moderate vision loss by 50 percent. With the use of intravitreal TA, multiple reports 3-5 have described transient improvement to both the retinal thickness and visual acuity, although significant side effects, including cataract and secondary open- angle glaucoma, were also reported.

Therefore, the Diabetic Retinopathy Clinical Research Network (, a consortium of more than 150 clinical centers and 500 physicians across the United States, developed a well-powered randomized controlled trial to compare these two treatment modalities directly. In the trial of 840 eyes, intravitreal TA (1 mg or 4 mg) was compared with focal/grid photocoagulation in patients with DME. After a two-year period, photocoagulation produced superior visual acuity and retinal thickness measurements and had fewer side-effects than TA.9 This was regarded as strong support for focal/grid photocoagulation as the benchmark against which other therapies for DME should be measured.

The results of this trial raise an important question: Is therapy with intravitreal corticosteroids now ever justified for treating DME given the inferior efficacy to photocoagulation and significant side-effect profile? This article will explore the science behind corticosteroids, potential side-effects, current application and trials of various formulations of corticosteroids, and the future of corticosteroids for the treatment of DME.


Science Behind Corticosteroids

Regimens of systemic corticosteroid therapy were widely used for treatment of neovascular retinal diseases 30 years ago, but as the adverse effects of systemic corticosteroids became known, their use was abandoned. Angiostatic steroids have been proposed to inhibit angiogenesis by altering the capillary basement membrane composition, suppressing its dissolution and inhibiting endothelial cell migration, in addition to their capacity for regulating the participation of inflammatory cells in the neovascular process. Recently, there has been a great deal of interest in angiostatic steroids for the treatment of diabetic macular edema.


Mechanism of Action

Macrophages are inflammatory cells that play a significant role during angiogenesis by secreting angiogenic cytokines and growth factors. Recently, they have also been found to produce INF-gamma, which is involved in angiogenic inhibition, and it has therefore been proposed that the role of macrophages during the process of neovascularization might be that of angiogenic on/off switch. Glucocorticoid steroids possess the ability of regulating macrophage activity via a glucocorticoid receptor (GR), which inhibits inflammatory programs of gene expression. Macrophages activated by glucocorticoids exhibit a different growth factor expression pattern than classically activated cells, with considerable differences in expression of angiogenic proteins like Tumor Necrosis Factor-alpha, Insulin-like Growth Factor-1, Platelet-derived Growth Factor-A and M-Kinase. It is therefore possible that a corticosteroid environment results in a change of macrophage behavior affecting not only the inflammatory response, but also the neovascular process.

In the retina, activation of the complement system leads to the secretion of angiogenic factors by recruited inflammatory cells and to the buildup of cells and extracellular deposits resulting in local ischemia and subsequent release of angiogenic stimuli by the retinal pigment epithelium. Oxidative stress has been shown to induce expression of Vascular Endothelial Growth Factor and Transforming Growth Factor-beta, as well as of the fibrosis-promoting Connective Tissue Growth Factor by RPE cells. Treatment of cells with TA, a corticosteroid with known anti-angiogenic effect, affects the expression of all three cytokines. The observed upregulation of VEGF and TGF-beta is reduced by treatment with TA in a concentration-dependent manner, while upregulation of CTGF is accelerated.

The potent anti-angiogenic effect observed after treatment of retinal cells with TA probably occurs due to the fact that this corticosteroid affects multiple pathways, increasing the probability of effectively disrupting the neovascular process. In addition to down-regulating the initial angiogenic stimulus by modifying expression of VEGF and TGF-beta, TA can also inhibit the degradation of capillary basal membrane. This degradation process constitutes a critical component of choroidal angiogenesis in vivo and it is carried out by matrix metalloproteinases that become activated in choroidal endothelial cells. Treatment of these cells with TA has been described as resulting in metalloproteinase downregulation and subsequent inhibition of endothelial cell migration and tube formation.

In addition to degradation of endothelial basal membrane, a breakdown of extracellular matrix also occurs during the invasive phase of angiogenesis, allowing endothelial cells to proliferate and migrate. This process is mediated by the urokinase-type plasminogen activator (uPA), which can be regulated by corticosteroids via activation of the PA inhibitor (PAI).12 This corticosteroid-induced suppression of PA function leads to a halt in endothelial cell proliferation and migration, and consequently to their inability to participate in new blood vessel formation.


Angiostatic Corticosteroids in DME

While patients with PDR are highly sensitive to anti-VEGF treatments, some patients with retinal edema show only partial responses and may require higher doses, suggesting that other VEGF-independent mechanisms may be involved in edema and require different therapeutic approaches. Recent reports imply that resolution of edema and improvements in VA in patients with DME may be better associated with an anti-inflammatory, rather than antiangiogenic, effect of corticosteroids. In one study, intravitreal triamcinolone yielded significantly better results (reduction of edema and improvement in VA) than bevacizumab on DME, evidence implying that other, non– VEGF-mediated mechanisms sensitive to corticosteroid but not bevacizumab treatment could be involved in development of edema. A comparison study of intravitreal ranibizumab or triamcinolone, alone or in combination with laser treatment for DME (LRT for DME), has recently begun.


Potential Side Effects

Corticosteroids administered intravitreally bypass the blood-ocular barrier, achieving therapeutic levels in the eye while avoiding systemic side effects. However, in view of the chronic nature of most of the neovascular diseases and the relatively rapid clearance of corticosteroids from the posterior segment, multiple injections are frequently necessary, increasing the risk of infection, hemorrhage and retinal detachment.

Some corticosteroid compounds, such as TA, have been found to have a longer intravitreal half-life than the originally studied dexamethasone,13 but multiple injections are still usually required to achieve disease remission. In addition, some of the most common and serious complications of intravitreal corticosteroid administration are not related to the injection procedure per se, but rather the individual sensitivity to prolonged presence of corticosteroids in the vitreous. One of the most common adverse effects is ocular hypertension, which occurs in 40 to 50 percent of eyes treated between one and two months following the procedure. Most of the episodes of intraocular pressure elevation respond favorably to topical medical treatment, but approximately 1 percent of patients eventually require filtering surgery.

Another ocular side effect that occurs with frequency during treatment is rapid cataract progression. A meta-analysis revealed a nearly 10 percent rate of significant cataract formation over one year and 25 percent at two years.14 The exact molecular mechanisms behind these complications are not known, but they are thought to be related to the sustained effects on cell metabolism. This has prompted an intense search for a new generation of corticosteroid compounds, which are specifically angiostatic yet free of hormonal activity.

In rabbits, doses of TA at 1 mg or lower were safe; however, doses between 4 and 20 mg produced destruction of photoreceptor outer segments and the migration of macrophage-like cells into the subretinal space.15 This could explain the initial improvement in visual acuity followed by decreased efficacy with follow-up injections that was observed in the DCRC study.

Rabbit data also suggested that intravitreal TA might inhibit the ocular immune response and make the eye more susceptible to infection.16 Studies report 0.05 to 0.6 percent and 0.8 to 9.3 percent infectious and "sterile" endophthalmitis rates respectively.14,17,18 Although the true cause of "sterile" endophthalmitis is unknown, hypotheses include the 0.99% benzyl alcohol bacteriostatic preservative or possible endotoxin in some commercially available TA.19-21 However, "sterile" endophthalmitis has now been described in benzyl alcohol-filtered TA preparations.


Formulations of Corticosteroids

Until recently, Kenalog (40 mg/ml, Bristol-Myers Squibb) was used off-label as the TA compound of choice for a variety of intraocular indications. However, a cluster of "sterile" endophthalmitis cases,17 (Grabe H, et al. IOVS 2008;49:ARVO E-Abstract 973) and a letter from the manufacturer sent to heath-care providers in late 2006 warning physicians to refrain from intraocular use, have led to the development and use of commercial preservative-free compounds. Triescence (40 mg/ml, Alcon) and Trivaris (80 mg/ml, Allergan) are preservative-free synthetic TA formulations that were approved by the Food and Drug Administration for indications including visualization during vitrectomy and the treatment of sympathetic ophthalmia, temporal arteritis, uveitis and ocular inflammatory conditions unresponsive to topical corticosteroids, on November 30, 2007 and June 16, 2008 respectively.

Various corticosteroid formulations are undergoing additional clinical investigation for DME. The I-vation TA (SurModics and Merck) is a metallic coil with a polymer/triamcinolone coating that is screwed through the pars plana after a small cut in the conjunctiva and a 25-ga. needle puncture. The drug can be delivered at variable rates. The 15-month preliminary results of a three-year Phase I safety and preliminary efficacy study revealed improvement to visual acuity and retinal thickness compared to baseline. Although the incidence of ocular hypertension was relatively small, lenticular opacities resulted in 36 percent of subjects (Eliott D, et al. IOVS 2008;49:ARVO E-Abstract 3501). Based on the promising Phase I results, a Phase IIb clinical trial was announced in June 2008; however, Merck recently suspended enrollment given the results of the DRCR.

Posurdex (Allergan) is a biodegradable dexamethasone implant delivered into the vitreous cavity in an office-based procedure. A Phase II trial evaluating the effects of
Posurdex on all types of macular edema, including diabetic macular edema, completed in 2003 revealed improved visual acuity and macular thickness. Allergan is currently enrolling 860 patients in a Phase III trial evaluating the efficacy of two doses of Posurdex and a 250-patient Phase II trial comparing focal laser photocoagulation with combined laser photocoagulation and Posurdex implants. The results are expected in several years.

Retisert (Bausch & Lomb) is a 0.59-mg surgically introduced nondegradable fluocinolone acetonide implant currently FDA approved only for eyes with uveitis. The three-year results of a randomized safety and efficacy clinical trial enrolling 197 patients with DME revealed a statistically significant improvement in visual acuity and macular edema over the standard of care. However, the side-effects are significant, with 95 percent of phakic eyes requiring a cataract extraction and 28 percent of eyes a glaucoma filtering procedure. (Pearson P, et al. IOVS 2006;47:ARVO E-Abstract 5442)

Iluvien (formerly Medidur FA, Alimera Sciences) is a narrow 3.5 x 0.37 mm cylindrical nondegradable fluocinolone acetonide implant that can be injected in an office-based procedure with a 25 ga. injector. The implant allows for the sustained release of FA in a low-dose (0.23 micrograms/day) or high-dose (0.45 micrograms/day) insert that lasts for 36 or 24 months respectively. Alimera is currently conducting two simultaneous Phase III clinical trials known as the FAME (Fluocinolone Acetonide in Diabetic Macular Edema) study, which completed enrollment of 956 patients in the United States, Canada, Europe and India in October 2007. The safety and efficacy of the implant for diabetic macular edema will be assessed after two years of follow-up.


Future of Corticosteroids for DME

The DRCR results apply to monotherapy for DME, but it is likely that combinations of therapies will be the best options in the future, especially in refractory cases. Combination therapies may be important for minimizing repeated administration and the risk of side effects in chronic disorders such as DME—i.e., cataract or glaucoma in the case of steroids, laser-related field defects and scotoma in the case of laser treatment, and intravitreal injection in general. The Laser-Ranibizumab-Triamcinolone for Diabetic Macular Edema (LRT for DME) study is a National Eye Institute-sponsored Phase III clinical trial of nearly 700 patients that began enrollment in 2007 and will directly compare laser alone with laser combined with triamcinolone, laser combined with ranibizumab, and ranibizumab alone. Similarly, as mentioned, Allergan is studying laser versus a combined dexamethasone implant and laser. It will be some time before quality studies answer whether corticosteroids hold promise as combination therapies, but the initial transient improvement in visual acuity and macular thickness reported in monotherapy provides hope for future applications involving this class of therapeutics.

In current clinical practice, most retinal specialists would treat focal DME with laser initially. In cases refractory to laser, intravitreal TA (4 mg/0.1cc) is often employed. Many retina specialists will consider intravitreal TA after the patient fails to respond to the first laser treatment, but others attempt a second or third session of macular laser photocoagulation before considering intravitreal TA. Generally, after three to four sessions of macular laser photocoagulation (i.e., after 300 to 400 spots), most retina specialists would not continue to offer laser photocoagulation, but would seek alternative treatments such as intravitreal TA, intravitreal bevacizumab or, as a last resort, vitrectomy with or without stripping of the internal limiting membrane. In cases that are refractory to both laser and TA, some retina specialists retreat with bevacizumab with or without half-dose 2 mg TA (to limit the risk of ocular hypertension and limit the amount of volume injected concurrently), as there may be mechanistic rationale for combining a target-specific anti-VEGF agent with a mechanistically "broad-spectrum" angiostatic corticosteroid. Medicare and most insurers reimburse for intravitreal TA used for DME, but do not reimburse for bevacizumab for this indication, although the out-of-pocket costs for bevacizumab are not unreasonable. Patients who present initially with severe diffuse DME are sometimes initially treated with combination intravitreal TA plus laser treatment, either on the same visit or on consecutive visits. Given these increasingly complex choices, the entire ophthalmic community eagerly awaits the results of DME clinical trials involving TA, bevacizumab, ranibizumab and other agents in order to provide more definitive guidance for this very common and potentially devastating disorder.


Dr. Comer is on the Vitreoretinal Service at the Kellogg Eye Center, University of Michigan. Dr. Ciulla practices in the Retina Section at Midwest Eye Institute, where he also conducts clinical research. Contact him at 201 Pennsylvania Pkwy, Indianapolis, Ind. 46280 Phone: (317) 817-1822; fax: (317) 817-1898 or e-mail: thomas ciulla@yahoo .com.

1. Jonas JB, Sofker A. Intraocular injection of crystalline cortisone as adjunctive treatment of diabetic macular edema. Am J Ophthalmol 2001;132:425-7.

2. Early Treatment Diabetic Retinopathy Study Research Group. Photocoagulation for diabetic macular edema. Early Treatment Diabetic Retinopathy Study report number 1. Arch Ophthalmol 1985;103: 1796-806.

3. Jonas JB, et al. Intravitreal triamcinolone acetonide for diabetic macular edema: A prospective, randomized study. J Ocul Pharmacol Ther 2006;22 (3):200-7.

4. Gillies MC, et al. Intravitreal triamcinolone for refractory diabetic macular edema: Two-year results of a double-masked, placebo-controlled, randomized clinical trial. Ophthalmology 2006;113:1533-8.

5. Nicolo M, et al. Intravitreal triamcinolone acetonide as primary treatment for diffuse diabetic macular edema: A prospective noncomparative interventional case series. Eur J Ophthalmol 2006;16(1):129-33.

6. Martidis A, et al. Intravitreal triamcinolone for refractory diabetic macular edema. Ophthalmology 2002;109:920-7.

7. Jonas JB, Kreissig I, Degenring R. Intraocular pressure after intravitreal injection of triamcinolone acetonide. Br J Ophthalmol 2003;87:24-7.

8. Jonas JB, et al. Frequency of cataract surgery after intravitreal injection of high-dosage triamcinolone acetonide. Eur J Ophthalmol 2005;15:462-4.

9. Diabetic Retinopathy Clinical Research Network. A randomized trial comparing intravitreal triamcinolone acetonide and focal/grid photocoagulation for diabetic macular edema. Ophthalmology 2008; 115:1447-1459.

10. Matsuda S, Gomi F, Oshima Y, Tohyama M, Tano Y. Vascular endothelial growth factor reduced and connective tissue growth factor induced by triamcinolone in ARPE19 cells under oxidative stress. Invest Ophthalmol Vis Sci 2005;46(3):1062-1068.

11. Wang YS, Friedrichs U, Eichler W, Hoffmann S, Wiedemann P. Inhibitory effects of triamcinolone acetonide on bFGF-induced migration and tube formation in choroidal microvascular endothelial cells. Graefe's Arch Clin Exp Ophthalmol 2002;240(1):42-48.

12. Blei F, Wilson EL, Mignatti P, Rifkin DB. Mechanism of action of angiostatic steroids: suppression of plasminogen activator activity via stimulation of plasminogen activator inhibitor synthesis. J Cell Physiol 1993;155:568-578.

13. Blankenship GW. Evaluation of a single intravitreal injection of dexamethasone phosphate in vitrectomy surgery for diabetic retinopathy complications. Ophthalmology 1991;229:62-65.

14. Jager RD, et al. Risks of intravitreous injection: A comprehensive review. Retina 2004;24:676-698.

15. Yu S-Y, et al. Retinal toxicity of intravitreal triamcinolone acetonide. A morphological study. Retina 2006;26:531-536.

16. Bucher RS, et al. Effect of intravitreal triamcinolone acetonide on susceptibility to experimental bacterial endophthalmitis and subsequent response to treatment. Arch Ophthalmol 2005;123: 649-653.

17. Jonisch J, et al. Increased incidence of sterile endophthalmitis following intravitreal preserved triamcinolone acetonide. Br J Ophthalmol 2008; 92(2):1051-4.

18. Bhavsar AR, et al. The risk of endophthalmitis following intravitreal treamcinolone injection in the DRCRnet and SCORE clinical trials. Am J Ophthalmol 2007;144(3):454-6.

19. Morrison VL, et al. Intravitreal toxicity of the kenalog vehicle (benzyl alcohol) in rabbits. Retina 2006;26:339-344.

20. Carrero JL, Barcia MG, and Flores IP. Sterile endophthalmitis after benzyl alcohol-filtered triamcinolone acetonide injection [letter]. Arch Ophthalmol 2008;126:142-3.

21. Roth DB, et al. Noninfectious endophthalmitis associated with intravitreal triamcinolone injection. Arch Ophthalmol 2003;121:1279-1282.