Samantha Fraser-Bell, Lyndell Lim, Anna Campain,Ian McAllister, Goff Quin, Wedad Salem,Christine Aroney, Mark Gillies,

Purpose:
Intravitreal injections of slow-release ster-oids, such as dexamethasone (DEX implant, Ozurdex),and inhibitors of vascular endothelial growth factor,such as bevacizumab (Avastin), have been shown to beefficacious for centre-involving diabetic macularoedema (DMO). The purpose of this study was tocompare bevacizumab to DEX implant for the treat-ment of DMO at one year.

Method:
Phase II, prospective, randomised trial. Weenrolled 88 eyes of 61 patients; 42 eyes receivedbevacizumab monthly, 46 DEX implant 4 monthly,both PRN. Primary outcome was the proportion ofeyes with BCVA improvement of 10 LogMAR letters.Secondary outcomes included mean change in bestcorrected visual acuity (BCVA), and central macularthickness (CMT). Results were analysed using linearregression with GEE methods.

Results:
Improvement in BCVA of 10 or more lettersoccurred in 17/42 (40%) eyes treated with bevacizumab and 19/46 (41%) of eyes with DEXimplant (p = 0.83). None of the bevacizumab eyes lost10 or more letters, whereas 5/46 (11%) DEX implantdid, mostly due to unoperated cataract. Meanimprovement in BCVA was 8.9 letters, (95% CI, 6.27–11.6) for bevacizumab eyes and 5.6 (95% CI, 0.90–10.4) for DEX implant eyes (p = 0.24). The mean CMTdecreased 122 ?m for bevacizumab and 187 ?m forDEX, (p = 0.015).

Conclusions:
Both DEX implant and bevacizumabcan improve vision in eyes with DMO. DEX implantachieved similar rates of VA improvement comparedwith bevacizumab, with superior anatomic outcomesand fewer injections. However, more DEX implanttreated eyes lost vision, mainly due to cataract.