Objective To report the effect of bevacizumab versus dexamethasone on hard exudates (HEX) in diabetic macular oedema (DME).

Design Post hoc analysis of 24-month data from the Randomised clinical trial of BEVacizumab OR DEXamethasone for diabetic macular oedema (BEVORDEX) phase 2 multicentre randomised clinical trial. Eyes with centre-involving DME resistant to or unlikely to benefit from macular laser therapy were included. Eyes were randomly assigned to bevacizumab every 4 weeks or Ozurdex dexamethasone implant (DEX) every 16 weeks, both as required. The 68 eyes from 48 patients that completed 24-month follow-up were analysed. Two masked graders assessed extent and location of HEX on baseline, 12-month and 24-month foveal-centred colour fundus photographs using validated grading software.

Results Macular HEX was present in 60% of study eyes. Of these, 21 eyes were treated with DEX and 20 eyes with bevacizumab. Both treatments led to reduction in area of macular HEX at 12 months and 24 months. There was greater regression of HEX from the foveal centre in DEX-treated eyes (median change +890 µm, IQR=1040 µm) than bevacizumab-treated eyes (median change +7.0 µm, IQR=590 µm) at 12 months (p=0.04) but the difference was no longer statistically significant (p=0.10) by 24 months (DEX +1400 µm, IQR=1590 µm; bevacizumab +20 µm, IQR=2680 µm). Reassuringly, no study eye developed HEX at the foveal centre, a visually devastating consequence of diabetic maculopathy.

Conclusions Bevacizumab and DEX were effective in reducing area of HEX in eyes with DME. DEX provided more rapid regression of HEX from the foveal centre although bevacizumab-treated eyes started to catch up by 24 months. Distance from the foveal centre as well as total area of macular HEX should be assessed when evaluating treatments for foveal-threatening HEX.

Trial registration number NCT01298076; Post-results.