Aflibercept in diabetic macular edema refractory to previous bevacizumab: outcomes and predictors of success

Purpose To evaluate functional and anatomical outcomes after aflibercept in patients with diabetic macular edema (DME) with poor response to bevacizumab. Methods We retrospectively reviewed patients with DME recalcitrant to bevacizumab who were switched to aflibercept between January and December 2015. All patients had a minimal follow-up of three months before the conversion and underwent at least three injections of bevacizumab. Functional outcome consisted in best corrected visual acuity (VA). Anatomical outcomes were demonstrated through central macular thickness (CMT) measured by optical coherence tomography. Results Forty-nine eyes of 34 subjects were reviewed. Mean VA improved from 0.55 +/- 0.32 logMAR to 0.46 +/- 0.33 logMAR (p = 0.038). Mean CMT decreased from 473 +/- 146 mu m to 349 +/- 85 mu m (p < 0.001). Twelve eyes (24%) demonstrated absence of macular edema after aflibercept. Previous bevacizumab exposure did not correlate with different outcomes. The variation of VA in response to aflibercept was significantly superior in the group with poorer VA before the switch (mean variation of -0.097 +/- 0.21 logMAR) when compared to eyes with VA < 0.4 logMAR (mean variation of +0.019 +/- 0.090 logMAR; p = 0.036). The same scenario was verified for anatomical outcomes as eyes with poor vision before the switch (>= 0.4 logMAR) achieved superior reduction in CMT in response to aflibercept (mean CMT variation of -157 +/- 171 mu m versus -49.5 +/- 39.9 mu m; p < 0.01). Pre-switch CMT was a predictor of CMT reduction after switching (B = -0.945; confidence interval 95% -1.1; -0.76; p < 0.001). Conclusions Conversion to aflibercept for persistent DME resulted in functional and anatomical improvements and these outcomes were not influenced by previous bevacizumab exposure. Pre-switch CMT was a predictor of anatomical changes after aflibercept.