Multicentre Real-World Evaluation of Intravitreal Bevacizumab, Aflibercept, and Ranibizumab for Diabetic Macular Edema.

PURPOSE: To assess the real-world effectiveness and treatment patterns of intravitreal bevacizumab, aflibercept (2 mg), and ranibizumab in patients newly diagnosed with diabetic macular edema (DME).

METHODS

A retrospective, multicenter observational study with a one-year follow-up was conducted at public clinics in Israel and Canada. A total of 309 eyes from 277 treatment-naive patients with DME were analyzed, including both eyes from 32 patients. Eyes were grouped by first-line therapy: bevacizumab (n=199), aflibercept 2 mg (n=63), or ranibizumab (n=47). The primary endpoint was change in visual acuity (VA) at one year. Secondary endpoints included change in central retinal thickness (CRT), injection frequency, final injection interval, and treatment response metrics. Generalized estimating equation (GEE) models were used to account for inter-eye correlation.

RESULTS

At 12 months, VA improved by 6 ETDRS letters with no significant difference between groups in the primary endpoint (p=0.501). In adjusted GEE analyses using bevacizumab as the reference group, aflibercept and ranibizumab were associated with greater CRT reduction (-60.31 µm and -57.24 µm, respectively), longer final injection intervals (+3.52 and +2.58 weeks), and fewer injections over 12 months (-2.05 and -2.21; all p<0.001). Aflibercept was also associated with lower odds of non-response (OR 0.08, p=0.010), and both aflibercept and ranibizumab were associated with lower odds of non-extension (OR 0.08 and OR 0.02, respectively; both p<0.001).

CONCLUSIONS

In this real-world multicenter cohort, VA improved after anti-VEGF treatment, without a significant between-group difference in the primary visual endpoint. Aflibercept (2 mg) was associated with more favorable secondary anatomical and treatment-burden metrics in adjusted analyses. However, given center-determined treatment allocation and residual baseline CRT imbalance, these findings should be interpreted as exploratory real-world associations rather than definitive evidence of pharmacological superiority. These results highlight the gap between randomized trial efficacy and outcomes observed in routine clinical practice, supporting the need for further practice-based DME research.