Skip to main content
  • Retina/Vitreous

    Review of: Aflibercept monotherapy or bevacizumab first for diabetic macular edema

    Jhaveri D, Glassman A, Ferris F, et al. New England Journal of Medicine, in press 2022

    The initial treatment options for diabetic macular edema (DME) may be limited to bevacizumab by insurance company step therapy mandates for some patients. This raises the question of the clinical impact of using bevacizumab first in patients, followed by a switch to other therapies. This study helps to answer that question by comparing the outcomes of patients who receive aflibercept from the outset to those who first receive bevacizumab for 12 weeks but, after no or minimal improvement, are switched to aflibercept.

    Study design

    This prospective, multicenter study randomly assigned eyes with center-involving DME and a Snellen visual-acuity (VA) equivalent of 20/320 to 20/50 to receive either aflibercept or bevacizumab. After 12 weeks, eyes that received bevacizumab were switched to aflibercept based on the protocol-specified criteria: persistent center-involved DME with central subfield thickness (CST) above the eligibility threshold; recent improvement in eye condition as defined by VA improvement of no more than 4 letters and CST decrease of no more than 10% as compared with each of the 2 preceding visits or between each of the 2 preceding visits; and, suboptimal vision as defined by VA of 20/50 or worse before 24 weeks or 20/32 or worse at 24 weeks or later. The primary outcome was the time-averaged change in the VA letter score, as observed in the area under the curve (AUC) during the 104-week trial period for the VA change from baseline and divided by the length of follow-up. Secondary outcomes included the percent of patients meeting switching criteria in the bevacizumab-first group, retinal CST, and safety outcomes.


    A total of 312 eyes (270 adults) were randomized. The aflibercept monotherapy group treated 158 eyes while 154 eyes received bevacizumab first. Of the eyes in the bevacizumab-first group, 70% were switched to aflibercept during the 2-year trial period. There was no statistically significant difference in the mean change in VA from baseline during the 2-year trial period. AUC indicated that VA was 15.0 ± 8.5 letters in the aflibercept-monotherapy group and 14.0 ± 8.8 letters in the bevacizumab-first group. The mean CST change from baseline to 2 years was −192 ± 143 μm in the aflibercept-monotherapy group and −198 ± 160 μm in the bevacizumab-first group. Over 2 years, eyes in the aflibercept-monotherapy group received a mean of 14.6 ± 4.1 injections and eyes in the bevacizumab-first group received 16.1 ± 4.1 injections. At least 1 serious systemic adverse event occurred in 60 of 116 patients (52%) in the aflibercept-monotherapy group, in 40 of 112 (36%) in the bevacizumab-first group, and in 18 of 42 (43%) patients with 2 study eyes. At least 1 hospitalization occurred in 56 patients (48%) in the aflibercept-monotherapy group, in 36 (32%) in the bevacizumab-first group, and in 18 (43%) among patients with 2 study eyes.


    Treating physicians were not blinded and patients could infer their treatment group based on billing information.   

    Clinical significance

    This study demonstrates that although the mean VA and CST outcomes were more favorable in the aflibercept monotherapy compared to the bevacizumab-first strategy throughout the first year, rescue treatment with aflibercept based on the protocol criteria resulted in similar mean visual and anatomical differences by the end of year 2. A large percent of patients (70%) in the bevacizumab group required a switch to aflibercept. Several clinical factors are important to consider when determining the ideal medication for an individual patient.