Natural history of diabetic macular edema and factors predicting outcomes in sham-treated patients (MEAD study).

PURPOSE To describe the natural history of diabetic macular edema (DME) with respect to best-corrected visual acuity (BCVA) and central retinal thickness (CRT) outcomes and to identify baseline patient characteristics and systemic factors associated with improvement or worsening of outcomes in sham-treated patients. METHODS The study population was sham-treated patients (n = 350) in the 3-year MEAD registration study of dexamethasone intravitreal implant for treatment of DME. Patients had center-involved DME and received sham intravitreal injections in the study eye at ≥ 6-month intervals. Potential prognostic factors for outcomes were evaluated using multiple linear regression analysis. RESULTS Visual and anatomic outcomes were poorer in patients who left the study early (n = 198) than in study completers (n = 152). Mean change in BCVA from baseline at the last visit with available data was + 0.9 letters; 37.5% of patients had no change in BCVA, 23.2% had gained > 10 letters, and 16.0% had lost > 10 letters. Older age and baseline diabetic retinopathy score > 6 were associated with worse BCVA outcomes; thicker baseline CRT and larger number of hypertension medications used were associated with larger reductions in CRT during the study. CONCLUSIONS BCVA and CRT outcomes were variable in this population of DME patients with generally good glycemic control. In DME patients without active treatment, older age and baseline diabetic retinopathy score > 6 were associated with less improvement in BCVA; thicker baseline CRT and a larger number of antihypertensive medications used predicted better improvement in CRT. TRIAL REGISTRATION The MEAD study trials are registered at ClinicalTrials.gov with the identifiers NCT00168337 and NCT00168389.