The Interval between Treatments of Bevacizumab and Dexamethasone Implants for Diabetic Macular Edema Increased over Time in the BEVORDEX Trial.

PURPOSE

To report increasing retreatment interval of intravitreal bevacizumab and dexamethasone implants for diabetic macular edema (DME) in the BEVORDEX trial.

DESIGN

Multicenter randomized clinical trial.

PARTICIPANTS

Sixty-eight eyes from 47 patients who completed 2 years of follow-up.

METHODS

The BEVORDEX study (www.clinicaltrials.gov identifier, NCT01298076), set in Australia, was the first head-to-head trial of bevacizumab (Avastin; Genentech, South San Francisco, CA), with retreatment considered after 4 weeks, versus a slow-release dexamethasone implant (Ozurdex; Allergan Inc., Irvine, CA), with retreatment possible after 16 weeks, for center-involving DME. Study eyes were assessed every 4 weeks for retreatment according to prespecified visual acuity and central macular thickness criteria. In this post hoc analysis, changes in treatment interval over time were examined using mixed-effects regression models. We assessed whether the mean treatment interval changed over time and if this depended on baseline characteristics or the treatment received.

RESULTS

Of the 68 eyes from 47 patients, 67 study eyes received at least 1 retreatment (1 eye in the dexamethasone implant group did not require retreatment over 24 months). Thirty-two eyes received bevacizumab and 35 eyes received dexamethasone implants. Study eyes received a mean of 14.6 injections (standard deviation [SD], 7.8 injections) and 5.6 injections (SD, 1.4 injections) in the bevacizumab and dexamethasone groups, respectively, over 2 years. The mean retreatment interval over the 2-year follow-up period was 70.8 days (SD, 43.8 days) for the bevacizumab group and 145 days (SD, 45.4 days) for the dexamethasone implant group. The mean treatment interval increased over time for both drugs (P = 0.016), independent of which treatment was received (P = 0.808). Longer treatment interval over time was associated with younger age (P = 0.037) and better baseline visual acuity (P = 0.026), but not with gender (P = 0.907) or baseline central macular thickness (P = 0.900).

CONCLUSIONS

The increase in treatment interval for both intravitreal bevacizumab and dexamethasone implants over time has implications when informing patients about potential treatment burden for DME, planning intravitreal injections services, and designing future clinical trials. For drugs with a disease-modifying effect, fixed-interval dosing may not be required beyond an initial loading phase.