Real-world visual acuity outcomes for patients with naïve neovascular age-related macular degeneration treated with aflibercept, ranibizumab, or bevacizumab in the Republic of Korea.

BACKGROUND

To compare the visual outcomes of different anti-vascular endothelial growth factor (VEGF) drugs, including aflibercept, ranibizumab, and bevacizumab, in a real-world setting in Korea.

METHODS

We collected data from patients who received monotherapy using one of these three anti-VEGF drugs as naïve treatment after being diagnosed with neovascular age-related macular degeneration. The number of injections and visual acuity (VA) outcomes of each cohort were obtained and pairwise comparisons were performed using propensity score matching.

RESULTS

A total of 254 aflibercept, 238 ranibizumab, and 282 bevacizumab treatment-naïve eyes were included. The mean VA change at 3 years for all cohorts combined was -1.8 letters, and the mean number of injections was 9.4. In the direct comparison of the three drugs, the mean change in the VA letter score was +2.0 letters for aflibercept and -11.7 letters for bevacizumab (P < 0.001). The number of aflibercept injections was significantly higher than the number of bevacizumab injections (P = 0.002). The visual outcomes for aflibercept and ranibizumab were +4.7 letters and -1.9 letters, respectively, and comparable results were obtained (P = 0.13). The VA outcomes for ranibizumab and bevacizumab were also not significantly different (P = 0.09). The numbers of injections for aflibercept, ranibizumab, and bevacizumab were 10.8, 6.7, and 8.8, respectively. Significant differences were observed between the injection frequencies comparisons of aflibercept and ranibizumab and ranibizumab and bevacizumab (P < 0.001 and P = 0.002, respectively).

CONCLUSIONS

In the Korean clinical medical environment, which included various confounding factors, especially socioeconomic ones, the aflibercept VA outcome was significantly better than that of bevacizumab, and aflibercept injections were the most numerous. These real-world data imply that the drug effect as well as the environment in which the drug can be sufficiently used affected patient final VA scores.