BEVORDEX - A multicentre randomized clinical trial of intravitreal dexamethasone versus intravitreal bevacizumab for persistent diabetic macular edema

Abstract

Purpose: To report the 24 month outcomes of the first head to head randomized clinical trial of Ozurdex® slow release dexamethasone intravitreal implant (DEX) versus intravitreal bevacizumab (BEV) for diabetic macular edema (DME).

Methods: This phase II, prospective, multicentre, randomized, single-masked clinical trial was performed at 4 Australian sites, Clinicaltrial.gov#NCT01298076. We enrolled 88 eyes of 61 patients, of which 42 eyes were randomized to receive BEV 4 weekly and 46 to DEX 4 monthly, both pro re nata. The primary outcome was the proportion of eyes that improved by 10 LogMAR letters. Secondary outcomes included mean change in visual acuity (VA), injection frequency, regression of hard exudates and adverse events. Results were analyzed using linear regression with generalized estimation equations methods to account for between eye correlation.

Results: VA improvement of 10 or more letters was found in 19 of 42 (45%) eyes treated with BEV compared with 20 of 46 (43%) DEX treated eyes (P=0.99). Only 1 of the 42 BEV treated eyes lost 10 or more letters, whereas 5 of 46 DEX treated eyes had this degree of VA loss (P=0.25). The mean improvement in VA was 9.6 letters, (95% confidence interval [CI], 6.9-12.3) for BEV treated eyes and 5.6 (95% CI, 0.7-10.5) for DEX treated eyes (P=0.16). The mean central macular thickness at 24 months was not significantly greater for BEV than DEX eyes (343.3µ vs. 311.7µ, P=0.28). BEV treated eyes received a mean of 13.2 injections vs. 5.0 for DEX treated eyes. There was greater regression of hard exudates in DEX compared to BEV treated eyes (P=0.02); however no study eye developed foveal plaquing. In phakic eyes, cataract progressed by 2 grades in 3/36 (8%) BEV eyes and 13/30 (43%) DEX eyes (P=0.02). During the study 12 DEX treated eyes and 2 BEV treated eyes had an intraocular pressure greater than 25mmHg recorded at least once during follow-up visits (P=0.01).

Conclusions: Both DEX and BEV can improve vision in eyes with DME. We found no significant difference between the 2 groups with respect to 10 letter VA gain, mean VA gain or central macular thickness at 24 months. However, more DEX treated eyes lost vision and BEV treated eyes had a higher treatment load.