The study by Rahimy and associates assessed the short-term visual and anatomic outcomes of patients with persistent diabetic macular edema (DME) who were converted from bevacizumab (Avastin; Genentech, Inc, South San Francisco, California, USA) and/or ranibizumab (Lucentis; Genentech, Inc) to aflibercept (Eylea; Regeneron Pharmaceuticals, Inc, Tarrytown, New York, USA). The article has several shortcomings that prevent the validation and extrapolation of their results and that can be specifically summarized as follows.

The study was retrospectively conducted with 56% of the eyes lost over 4.6 months of subsequent follow-up. Of note, the primary outcome analysis was carried out for all eyes through the second visit following aflibercept conversion, which averaged 2.5 months.

The follow-up data were limited in length after transition, with the future possibility of tachyphylaxis or tolerance occurring in response to aflibercept.

There were no data on the anatomic types of macular edema (subretinal fluid/cystic changes within neurosensory retina) before and after switching to aflibercept. Nothing was stated regarding the percentage of eyes with central macular thickness (CMT) ≤320 μm at the end of the follow-up.

There was a disparity between changes of the visual acuity and CMT. Although improving vision was minimal (mean logMAR change of −0.05), decreasing CMT was significant, to 348.7 μm. Importantly, these values are more than the cutoff (315.2 μm) for the upper level of normal foveal thickness (270.2 ± 22.5 μm) plus 2 standard deviations. We believe that persistence of high values of the CMT highlights unresolved macular edema and indicates that the disease process is still active and progressive, requiring further treatment with antiangiogenic agents.

The presumed pharmacologic advantages of aflibercept over bevacizumab or ranibizumab (eg, a higher binding affinity for vascular endothelial growth factor [VEGF]-A and activity against VEGF-B and placental-derived growth factor [PIGF]) were not confirmed by the poor results of this series, namely, incomplete resolution of the DME; an approximately 3 Early Treatment Diabetic Retinopathy Study (ETDRS) letters gain in visual acuity; and increase in the number of eyes with epiretinal membranes from 18 to 20 and of those with vitreomacular tractions from 2 to 4. Most likely, there was a permanent VEGF receptor 2–mediated breakdown of the inner blood-retinal barrier. Additionally, there was also a permanent VEGF receptor 1–mediated rupture of the retinal pigment epithelium junctions induced by high vitreous levels of PIGF in patients with diabetic retinopathy. This permanent retinal capillaropathy caused by long-standing duration of DME was temporarily relieved by reduction of the edematous component with antiangiogenic agents. However, the pathology was incurable owing to irreversible ischemic changes to the macular ganglion cell complex, close to the foveola, with macular edema being a minor factor.

Altogether, regardless of the anti-VEGF agents used (bevacizumab/ranibizumab/aflibercept), the efficacy of therapy depends primarily on the precociousness of the therapy after DME diagnosis. Therefore, therapy with antiangiogenic agents has to be promptly applied as soon as possible after DME onset. Every delay of therapy adversely influences the deterioration of visual functions that are difficult to restore, even with subsequent treatment.