Research report: Early outcomes of switching from bevacizumab to aflibercept-jbvf in exudative macular diseases: a real-world study

The visual benefit, however, did not reach statistical significance, and the researchers themselves caution that the anatomical change cannot be confidently attributed to the new drug alone.

The study, led by Tugce Horozoglu Ceran, Selma Beyatli and Mustafa Dogan, enrolled 41 patients contributing 53 eyes with three common causes of fluid accumulation in the retina: diabetic macular edema, macular neovascularization secondary to neovascular age-related macular degeneration, and macular edema from retinal vein occlusion.

The primary question was whether the switch produced measurable changes in two key metrics within one month. The first was best-corrected visual acuity. The second was central subfield thickness, measured on optical coherence tomography scans. Both were compared against a preswitch baseline taken immediately before the aflibercept-jbvf injection, after the bevacizumab loading phase.

The anatomical result was clear in direction, if modest in size. Median central subfield thickness fell from 312 micrometers at baseline, with an interquartile range of 275 to 354 micrometers, to 291 micrometers at week 4, with an interquartile range of 255 to 322 micrometers. The mean change was a reduction of 22.6 plus or minus 68.1 micrometers, with a 95 percent confidence interval of minus 40.9 to minus 4.2 micrometers and a p value of 0.001. When the researchers used generalized estimating equation models to account for the fact that some patients contributed both eyes, the thickness reduction remained statistically significant at p equals 0.005.

The functional picture was less convincing. Median best-corrected visual acuity was 0.699 logMAR at both the preswitch baseline and week 4, meaning the group median did not move at all. The mean change was a numerical improvement of 0.051 plus or minus 0.230 logMAR, with a confidence interval spanning zero, from minus 0.114 to plus 0.013, and a p value of 0.129. After adjustment for within-patient intereye correlation, the visual acuity change was still not significant, with a beta of minus 0.051 and a p value of 0.065. The authors describe this as a small numerical trend toward improvement that failed to reach statistical significance in either analysis.

The cohort itself was skewed toward diabetic disease: 30 of the 53 eyes, or 56.6 percent, had diabetic macular edema, 18 eyes, or 34.0 percent, had macular neovascularization, and only five eyes, or 9.4 percent, had retinal vein occlusion. The authors explicitly flag that these are biologically and clinically distinct entities, so the pooled results should be read as descriptive cohort-level estimates within a shared treatment pathway rather than evidence of a uniform treatment effect across conditions. Any diagnosis-specific breakdowns were treated as exploratory.

Perhaps the most important caveat comes from the study design itself. This was a single-arm study with no comparator group, meaning every eye received the same sequence of drugs and there was no control arm continuing bevacizumab or receiving placebo. The authors note that the observed anatomical change cannot be attributed specifically to aflibercept-jbvf and may partly reflect a delayed or cumulative response to the prior bevacizumab injections, regression to the mean, or natural variability in retinal thickness over time. In other words, the four-week improvement in swelling sits in exactly the window where a lagged effect of the previous drug could plausibly appear.

Safety data were reassuring within the limits of the small sample: no serious adverse events were observed over the study period. The authors also stress that a four-week endpoint says nothing about how durable any response might be or about longer-term treatment burden, both of which matter greatly for chronic diseases that typically require years of repeated injections.

The findings suggest the switch was temporally associated with a modest early anatomical gain in this heterogeneous group, without a statistically significant visual gain, and without any signal of serious harm. Whether that pattern holds with a comparator design, larger disease-specific cohorts and longer follow-up remains an open question that this study, by its own design, cannot answer.

Subject of Research: Early anatomical and visual outcomes of switching from intravitreal bevacizumab to aflibercept-jbvf in patients with exudative macular diseases

Article Title: Early outcomes of switching from bevacizumab to aflibercept-jbvf in exudative macular diseases: a real-world study

Article References: Horozoglu Ceran, T., Beyatli, S., & Dogan, M. (2026). Early outcomes of switching from bevacizumab to aflibercept-jbvf in exudative macular diseases: a real-world study. BMC Ophthalmology. https://doi.org/10.1186/s12886-026-05491-4

Image Credits: AI Generated

DOI: 10.1186/s12886-026-05491-4

Keywords: aflibercept-jbvf, bevacizumab, anti-VEGF, diabetic macular edema, neovascular age-related macular degeneration, retinal vein occlusion, central subfield thickness, best-corrected visual acuity, biosimilar, intravitreal injection, BMC Ophthalmology, single-arm study