To compare the effect of an intravitreal injection of bevacizumab, an anti-vascular endothelial growth factor (VEGF) antibody, with that of triamcinolone acetonide, a corticosteroid for reduction of diabetic macular edema (DME).

Prospective, comparative interventional case series.

Twenty-eight eyes of 14 patients with bilateral DME participated in this study. In each patient, one eye received an intravitreal injection of 4 mg triamcinolone acetonide and the other eye received 1.25 mg bevacizumab. The clinical course of best-corrected visual acuity (VA) with a logarithm of the minimum angle of resolution chart and averaged foveal thickness using optical coherence tomography was monitored for up to 24 weeks after the injection.

Before the injection, foveal thickness and VA were 522.3 ± 91.3 μm and 0.64 ± 0.28 μm in the triamcinolone-injected eye, and 527.6 ± 78.8 μm and 0.61 ± 0.18 μm in the bevacizumab-injected eye, respectively; there was no significant difference between the eyes. One week after the injection, both eyes showed significant regression of macular edema. The triamcinolone-injected eye (342.6 ± 85.5 μm and 0.33 ± 0.21 μm) showed significantly better results than the bevacizumab-injected eye (397.6 ± 103.0 μm and 0.37 ± 0.17 μm). However, both eyes showed the recurrence of macular edema with time, even at 24 weeks. Triamcinolone (410.4 ± 82.4 μm and 0.47 ± 0.25 μm) kept better results than bevacizumab (501.6 ± 92.5 μm and 0.61 ± 0.17 μm).

With the generally used concentration, intravitreal injection of triamcinolone acetonide showed better results in reducing DME and in the improvement of VA than that of bevacizumab, suggesting that the pathogenesis of DME is not only attributable to VEGF-dependency, but is also attributable to other mechanisms suppressed by corticosteroid.