Document Type


Publication Date



This article is the author's final published version in the Journal of Personalized Medicine, Volume 13, Issue 3, 2023, Article number 544.

The published version is available at Copyright © 2023 by the authors. Licensee MDPI, Basel, Switzerland.


This retrospective cohort study aimed to assess the systemic effects of three commonly available anti-vascular endothelial growth factor intravitreal injections in patients with diabetes, using data taken from a multi-institutional database in Taiwan. Patient data were sourced from the multi-institutional Chang Gung Research Database. Participants were divided into groups based on treatment with bevacizumab, ranibizumab, or aflibercept. Baseline characteristics were matched among the groups by the inverse probability of treatment weighting. The incidence rate of outcome events was calculated as the number of events divided by 100 person-years of follow-up. The cumulative incidence function was used to estimate the incidence rate of the outcome events among groups. The incidence of ischemic stroke was higher in the ranibizumab group than the bevacizumab and aflibercept groups (1.65, 0.92, and 0.61 per 100 person-years, respectively). The incidence of major adverse lower-limb events was higher in the bevacizumab group (2.95), followed by ranibizumab (2.00) and aflibercept (0.74). Major bleeding was relatively higher in bevacizumab (12.1) compared to ranibizumab (4.3) and aflibercept (3.8). All-cause death was higher for both bevacizumab (3.26) and aflibercept (2.61) when compared to ranibizumab (0.55), and all-cause admission was found to be highest with bevacizumab (58.6), followed by aflibercept (30.2), and ranibizumab (27.6). The bevacizumab group demonstrated a greater decrease in glycated hemoglobin compared to the baseline level (−0.33%). However, a few differences in the clinical condition between the groups were still observed after matching. In conclusion, this study suggests that different anti-vascular endothelial growth factor agents may be associated with various and differing systemic adverse events. The differences might also be attributed to differences in patient characteristics and clinical status.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.

Supplemental Material.pdf (203 kB)
Table S1: Baseline characteristics of patients after inverse probability of treatment weighting