INTRAVITREAL TUMOR NECROSIS FACTOR INHIBITORS IN THE TREATMENT OF REFRACTORY DIABETIC MACULAR EDEMA A Pilot Study From the Pan-American Collaborative Retina Study Group LIHTEH WU, MD,* ERICK HERNANDEZ-BOGANTES, MD,* JOSE ´ A. ROCA, MD,† J. FERNANDO AREVALO, MD,‡ KAREN BARRAZA, MD,† ANDRES F. LASAVE, MD‡ Purpose: The purpose of this study was to report the short-term visual and anatomical outcomes after intravitreal injections of two different tumor necrosis factor a inhibitors in eyes with refractory diabetic macular edema. Methods: An interventional, retrospective, multicenter study of 39 eyes with refractory diabetic macular edema that were injected with adalimumab (n = 5 for 2 mg) or infliximab (n = 15 for 1 mg; n = 19 for 2 mg). The main outcome measures were the best-corrected visual acuity and the central macular thickness at 3 months of follow-up. Results: In the 1-mg infliximab group, the logarithm of the minimal angle of resolution best-corrected visual acuity improved from 1.49 6 0.58 at baseline to 1.38 6 0.56 at 3 months (P = 0.6991). In the 2-mg infliximab group, the logarithm of the minimal angle of resolution best-corrected visual acuity worsened from 0.76 6 0.54 to 1.03 6 0.69 at 3 months (P = 0.5995). In the adalimumab group, the logarithm of the minimal angle of resolution best-corrected visual acuity improved from 1.44 6 0.77 to 1.08 6 0.85 at 3 months (P = 0.2500). The central macular thickness in the 1-mg infliximab group decreased from 459 6 125 mm at baseline to 388 6 131 mm at 3 months (P = 0.1178). In the 2-mg infliximab group, the central macular thickness remained unchanged from 378 6 97 mm at baseline to 349 6 118 mm at 3 months (P = 0.2162). In the adalimumab group, the central macular thickness remained unchanged from 521 6 163 mm at baseline to 526 6 390 mm at 3 months (P = 0.1250). There were no systemic side effects reported in any of the patients. However, laboratory markers for autoimmunity were not done. None of the eyes injected with either adalimumab or 1 mg of infliximab had adverse ocular events. In the 2-mg infliximab group, 42% (8 of 19) of eyes developed severe uveitis. Three of these eyes (37.5%) required pars plana vitrectomy. The uveitis in the remaining five eyes resolved with topical steroid therapy. Conclusion: Both intravitreal adalimumab and infliximab do not appear to benefit eyes with refractory diabetic macular edema. Intravitreal injections of infliximab may elicit a severe intraocular inflammatory reaction. RETINA 31:298–303, 2011 D iabetic macular edema (DME) remains the most common cause of moderate visual loss in diabetic patients. 1 The Early Treatment Diabetic Retinopathy Study (ETDRS) showed that in eyes with DME, focal or grid macular photocoagulation reduced moderate visual loss by 50%. However, only 17% had some visual gain and less than 3% had a significant gain after 3 years of follow-up. 2 The Diabetic Retinopathy Clinical Research Network 3 has recently confirmed that macular laser photocoagulation remains the treatment of choice for eyes with DME. At 3 years of follow-up, the laser-treated eyes gained an average of 5 letters, and in 62% of eyes, some visual gain was obtained when compared with baseline. In 21% of 298