Aflibercept biosimilar use and real-world outcomes in various retinal diseases: the AURA Korea study Authors: Jae Ryong Song, Jehwi Jeon, Seung Chul Baek, Ashish Sharma, Min Sagong, Christopher Seungkyu Lee, Se Joon Woo, Carl Regillo, Carl Regillo, Arshad M. Khanani, David Eichenbaum, Baruch D. Kuppermann, Anat Loewenstein, Francesco Bandello, Luke Nicholson, Clara Vazquez-Alfageme, David Boyer, Nilesh Kumar, Nikulaa Parachuri, Alay Banker, Kourous A. Rezaei, Peter K. Kaiser, Frank G. Holz, Sobha Sivaprasad, Taku Wakabayashi, Giuseppe Querques, Şengül Özdek, Kodjikian Laurent, Alper Bilgic, Paolo Lanzetta, Caroline R. Baumal, Nancy Holekemp, Taiji Sakamoto, Adnan Tufail, Nicolas Yannuzzi, Giulia Corradetti, Jay Sheth, Chitaranjan Mishra, Debdulal Chakraborty, Aleksandra Rachitskaya Published: 2026 ABSTRACT This multicenter, retrospective real-world study evaluated the efficacy and safety of aflibercept 2 mg biosimilar SB15 (Afilivu®) across multiple retinal diseases. A total of 1083 eyes from 984 patients received 2762 injections between March 2023 and February 2025 at three South Korean tertiary centers, comprising 188 treatment-naïve (17.4%) and 895 switching eyes (82.6%). Diagnoses included neovascular age-related macular degeneration (nAMD, n = 611), retinal vein occlusion (RVO, n = 182), diabetic macular edema (DME, n = 173), and other conditions (n = 117). Over a mean follow-up of 5.17 ± 2.76 months with 2.55 ± 1.09 injections, central foveal thickness decreased significantly by − 47.5 µm (95% CI − 55.2 to − 39.8; P < 0.001), while best-corrected visual acuity remained stable (mean change − 0.003 LogMAR; P = 0.572). Treatment-naïve eyes demonstrated greater anatomical improvement than switching eyes (− 93.9 vs. − 37.9 µm; P < 0.001). Retinal fluid resolved significantly across diagnostic groups: in nAMD, subretinal fluid decreased from 51.2 to 29.6% (P < 0.001); in RVO, intraretinal fluid reduced from 92.9 to 76.4% (P < 0.001); and in DME, intraretinal fluid decreased from 90.8 to 84.4% (P = 0.002). One case (0.09%) of rhegmatogenous retinal detachment occurred; no intraocular inflammation or other serious adverse events were observed. These findings support the real-world effectiveness and safety of SB15 (Afilivu®) in both treatment-naïve and switching populations across diverse retinal diseases.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-52212-4. Introduction Anti-vascular endothelial growth factor (VEGF) injections have changed the treatment landscape for retinovascular disorders, significantly improving visual outcomes for patients1. Aflibercept 2 mg (Eylea®, Regeneron Pharmaceuticals, Tarrytown, NY, USA / Bayer HealthCare, Leverkusen, Germany) is a recombinant fusion protein that binds VEGF-A, VEGF-B, and placental growth factor (PlGF), approved for the treatment of neovascular age-related macular degeneration (nAMD), diabetic macular edema (DME), and retinal vein occlusion (RVO)1. The VIEW 1 and VIEW 2 studies demonstrated that aflibercept 2 mg achieved equivalent efficacy and safety compared to ranibizumab, with the every-2-month dosing regimen offering the potential to reduce treatment burden2. Aflibercept has become a widely used first-line therapy for nAMD and DME, providing optimal visual gains and anatomic disease control3,4. Biosimilars are biological agents that are highly similar to an already-approved biological medicine (reference product) with no clinically meaningful differences in terms of quality, safety, and efficacy5. Biosimilars can reduce financial burdens on healthcare systems while providing clinically equivalent alternatives, leading to improved treatment access and better clinical outcomes6. Following the expiration of reference aflibercept 2 mg (Eylea) market exclusivity in 2024, multiple aflibercept biosimilars have been rapidly developed and approved7. SB15 (Afilivu®, Samsung Bioepis, South Korea) received approval from the Korean Ministry of Food and Drug Safety (MFDS) in February 2024, followed by approvals from the U.S. Food and Drug Administration (FDA) and the European Medicines Agency (EMA)8,9. Phase 3 randomized clinical trials demonstrated that SB15 showed equivalent efficacy and comparable safety, pharmacokinetics, and immunogenicity to reference aflibercept (AFL) in patients with nAMD10,11. While randomized controlled trials (RCTs) serve as the gold standard for demonstrating biosimilar efficacy and safety, strict inclusion and exclusion criteria may limit their generalizability to real-world clinical practice12. Real-world evidence can provide valuable insights into biosimilar performance across diverse patient populations, including those with comorbidities and concomitant medications that are often excluded from clinical trials6. Furthermore, real-world data is essential for resolving negative perceptions of biosimilars and building confidence among physicians and patients in their clinical implementation6,7. The purpose of this study was to evaluate the real-world efficacy and safety of the aflibercept 2 mg biosimilar SB15 (Afilivu®), thereby validating RCT findings in actual clinical settings and providing evidence to support its widespread clinical use. Methods Study design and participants This multicenter, retrospective, observational study was conducted at three tertiary referral centers in South Korea (Seoul National University Bundang Hospital, Yeungnam University Hospital, and Severance Hospital). The study included all patients who received at least one intravitreal injection of SB15 (Afilivu®) between March 1, 2023, and February 28, 2025. Eligible patients were required to have a minimum follow-up period of one month, with no other intravitreal anti-VEGF agents administered concurrently with SB15 (Afilivu®) during the treatment period. For patients who subsequently switched to other intravitreal anti-VEGF agents, the follow-up period was defined as the duration from the first SB15 (Afilivu®) injection to the date of the last visit prior to switching, and data collected after the switch were not included in the analysis. Patients without visual acuity or optical coherence tomography (OCT) data at one month post-injection were excluded from the analysis. This study was approved by the Institutional Review Boards of Seoul National University Bundang Hospital (IRB No. B-2507-982-106), Yeungnam University Hospital (IRB No. 2025-10-018), and Severance Hospital (IRB No. 4-2025-0321), and was conducted in accordance with the tenets of the Declaration of Helsinki. Due to the retrospective nature of the study, the requirement for informed consent was waived by the Institutional Review Boards. Data collection Baseline demographic and clinical data were collected from electronic medical records, including age, sex, underlying diagnosis, treatment history (treatment-naïve or switching from reference aflibercept 2 mg), and comorbidities (diabetes mellitus, hypertension). Treatment-naïve eyes were defined as those receiving their first anti-VEGF therapy, while switching eyes had previously received reference aflibercept 2 mg or other anti-VEGF agents before transitioning to SB15 (Afilivu®). Treatment metrics included follow-up duration, total number of SB15 (Afilivu®) injections, number of previous intravitreal injections, and type of last intravitreal injection prior to switching. All patients received intravitreal injections of 2 mg (0.05 mL) SB15 (Afilivu®) following standard aseptic technique. Outcome measures The primary efficacy outcome was the change in best-corrected visual acuity (BCVA) from baseline to the final follow-up visit, measured using the Early Treatment Diabetic Retinopathy Study (ETDRS) chart and converted to logarithm of the minimum angle of resolution (LogMAR) units. The primary anatomical outcome was the change in central foveal thickness (CFT) measured by spectral-domain OCT from baseline to the final follow-up visit. Secondary outcomes included BCVA and CFT changes at 1, 2, 3, 4, 5, and 6 months post-treatment. All OCT scans were acquired using the Spectralis HRA + OCT device (Heidelberg Engineering, Heidelberg, Germany) with the macular volume cube scan protocol, and layer segmentation of each scan was reviewed by a designated retinal specialist at each institution with manual correction when automated segmentation was inaccurate. Fluid analysis assessed the presence of subretinal fluid (SRF), intraretinal fluid (IRF), and pigment epithelial detachment (PED) at baseline and final follow-up visits. Additional outcomes included documentation of prior anti-VEGF treatment agents in switching patients and evaluation of adverse events, including intraocular inflammation, endophthalmitis, and other ocular or systemic complications. Statistical analysis Continuous variables were expressed as mean ± standard deviation (SD). Categorical variables were presented as frequencies and percentages. The independent t-test or Mann–Whitney U test was used to compare continuous variables between treatment-naïve and switching groups, as appropriate. The chi-square test or Fisher’s exact test was employed for categorical variables. Changes in fluid parameters (SRF, IRF, PED) from baseline to final visit were analyzed using McNemar’s test for paired proportions. All statistical analyses were performed using Python (version 3.9, Python Software Foundation). A two-sided P value of less than 0.05 was considered statistically significant. Results Baseline characteristics A total of 1247 eyes from 1,147 patients were initially screened. After excluding 164 eyes with insufficient follow-up duration (< 1 month, n = 163), or missing baseline BCVA or CFT (n = 1), 1083 eyes (86.8%) from 984 patients who received 2762 aflibercept biosimilar injections were included in the final analysis. The cohort comprised 188 treatment-naïve eyes (17.4%) and 895 switching eyes (82.6%). The most common underlying diagnosis was neovascular age-related macular degeneration (nAMD; n = 611, 56.4%), followed by retinal vein occlusion (RVO; n = 182, 16.8%), diabetic macular edema (DME; n = 173, 16.0%), and other diagnoses (n = 117, 10.8%) (Table 1).Table 1Baseline characteristics of the study cohort.ParameterOverall (n = 1083)Treatment-naïve (n = 188)Switching (n = 895)P value†Diagnosis, n (%)0.502 AMD611 (56.4)107 (56.9)504 (56.3) RVO182 (16.8)31 (16.5)151 (16.9) DME173 (16.0)25 (13.3)148 (16.5) Other117 (10.8)25 (13.3)92 (10.3)Demographics Age (years)70.7 ± 12.669.5 ± 13.870.9 ± 12.30.155 Male sex, n (%)559 (51.6)101 (53.7)458 (51.2)0.578Comorbidities, n (%) HTN498 (46.0)92 (48.9)406 (45.4)0.4161 DM375 (34.6)63 (33.5)312 (34.9)0.7877Baseline ocular parameters BCVA (LogMAR)0.606 ± 0.5960.558 ± 0.5960.616 ± 0.5950.218 CFT (μm)375.5 ± 152.1413.7 ± 164.8367.6 ± 148.2 < 0.001Treatment metrics Follow-up duration (months)5.17 ± 2.764.42 ± 2.575.33 ± 2.77 < 0.001 Total aflibercept biosimilar injections2.55 ± 1.092.73 ± 1.052.51 ± 1.090.01 Previous anti-VEGF injections*N/AN/A15.10 ± 14.07N/AValues are presented as mean ± standard deviation (SD) or number (percentage). BCVA = best-corrected visual acuity; CFT = central foveal thickness; HTN = hypertension; DM = diabetes mellitus; AMD = age-related macular degeneration; RVO = retinal vein occlusion; DME = diabetic macular edema.†P values compare the treatment-naïve and switching groups using the independent t-test or χ2 test, as appropriate.Previous anti-VEGF injections are applicable only to the switching group. Table 1Baseline characteristics of the study cohort.ParameterOverall (n = 1083)Treatment-naïve (n = 188)Switching (n = 895)P value†Diagnosis, n (%)0.502 AMD611 (56.4)107 (56.9)504 (56.3) RVO182 (16.8)31 (16.5)151 (16.9) DME173 (16.0)25 (13.3)148 (16.5) Other117 (10.8)25 (13.3)92 (10.3)Demographics Age (years)70.7 ± 12.669.5 ± 13.870.9 ± 12.30.155 Male sex, n (%)559 (51.6)101 (53.7)458 (51.2)0.578Comorbidities, n (%) HTN498 (46.0)92 (48.9)406 (45.4)0.4161 DM375 (34.6)63 (33.5)312 (34.9)0.7877Baseline ocular parameters BCVA (LogMAR)0.606 ± 0.5960.558 ± 0.5960.616 ± 0.5950.218 CFT (μm)375.5 ± 152.1413.7 ± 164.8367.6 ± 148.2 < 0.001Treatment metrics Follow-up duration (months)5.17 ± 2.764.42 ± 2.575.33 ± 2.77 < 0.001 Total aflibercept biosimilar injections2.55 ± 1.092.73 ± 1.052.51 ± 1.090.01 Previous anti-VEGF injections*N/AN/A15.10 ± 14.07N/AValues are presented as mean ± standard deviation (SD) or number (percentage). BCVA = best-corrected visual acuity; CFT = central foveal thickness; HTN = hypertension; DM = diabetes mellitus; AMD = age-related macular degeneration; RVO = retinal vein occlusion; DME = diabetic macular edema.†P values compare the treatment-naïve and switching groups using the independent t-test or χ2 test, as appropriate.Previous anti-VEGF injections are applicable only to the switching group. Baseline characteristics of the study cohort. Values are presented as mean ± standard deviation (SD) or number (percentage). BCVA = best-corrected visual acuity; CFT = central foveal thickness; HTN = hypertension; DM = diabetes mellitus; AMD = age-related macular degeneration; RVO = retinal vein occlusion; DME = diabetic macular edema. †P values compare the treatment-naïve and switching groups using the independent t-test or χ2 test, as appropriate. Previous anti-VEGF injections are applicable only to the switching group. The mean age of the study population was 70.7 ± 12.6 years, with 51.6% male patients. Comorbidities included hypertension (46.0%) and diabetes mellitus (34.6%). Baseline demographics and comorbidities were comparable between treatment-naïve and switching groups (all P > 0.05). The mean baseline BCVA was 0.606 ± 0.596 LogMAR, and the mean baseline CFT was 375.5 ± 152.1 μm. Treatment-naïve eyes had significantly greater baseline CFT compared to switching eyes (413.7 ± 164.8 μm vs. 367.6 ± 148.2 μm, P < 0.001). The mean follow-up duration was 5.17 ± 2.76 months, with an average of 2.55 ± 1.09 SB15 (Afilivu®) injections administered. Switching patients had received a mean of 15.10 ± 14.07 previous anti-VEGF injections before transitioning to SB15 (Afilivu®). Among the switching eyes, the most common last agent before initiating aflibercept biosimilar was bevacizumab (Avastin, Genentech, USA) (50.2%), followed by Eylea (Aflibercept 2 mg) (24.1%) and Ranibizumab biosimilar (16.7%; Amelivu 13.4%, LucenBS 3.3%). During the follow-up period, 617 eyes (68.9%) discontinued SB15 (Afilivu®) and switched to other intravitreal agents. The most frequently switched agents were Eylea (37.2%; 2 mg 34.4%, 8 mg 2.8%), ranibizumab biosimilar (26.9%; Amelivu 2.6%, LucenBS 0.3%), and Avastin (19.8%) (Table 2).Table 2Last agents before initiating aflibercept biosimilar in previously treated patients and subsequent agents after biosimilar discontinuation.Agent typeLast agent before switching to aflibercept BS*Agent after switching from aflibercept BS†Avastin (Bevacizumab)438 (50.2)122 (19.8)Eylea 2 mg (Aflibercept)210 (24.1)212 (34.4)Eylea 8 mg (Aflibercept)–17 (2.8)Lucentis (Ranibizumab)20 (2.3)–Ranibizumab biosimilar146 (16.7)166 (26.9)Vabysmo (faricimab)22 (2.5)20 (3.2)Beovu (brolucizumab)16 (1.8)19 (3.1)Ozurdex (dexamethasone)11 (1.3)6 (1.0)Triamcinolone acetonide†10 (1.1)5 (0.8)Total873 (100.0)617 (100.0)* Previously treated group: Last intravitreal injection before initiating aflibercept biosimilar treatment. Among the 895 eyes in the switching cohort, 22 eyes had no documented prior anti-VEGF agent in the electronic medical records (patients previously treated at outside institutions without a specified agent) and were excluded from this tabulation, yielding an analyzable denominator of n = 873.†Patients who switched from aflibercept biosimilar to other agents during follow-up period. Non-pharmacological interventions (PDT, n = 3; PPV, n = 1) were excluded from analysis. Table 2Last agents before initiating aflibercept biosimilar in previously treated patients and subsequent agents after biosimilar discontinuation.Agent typeLast agent before switching to aflibercept BS*Agent after switching from aflibercept BS†Avastin (Bevacizumab)438 (50.2)122 (19.8)Eylea 2 mg (Aflibercept)210 (24.1)212 (34.4)Eylea 8 mg (Aflibercept)–17 (2.8)Lucentis (Ranibizumab)20 (2.3)–Ranibizumab biosimilar146 (16.7)166 (26.9)Vabysmo (faricimab)22 (2.5)20 (3.2)Beovu (brolucizumab)16 (1.8)19 (3.1)Ozurdex (dexamethasone)11 (1.3)6 (1.0)Triamcinolone acetonide†10 (1.1)5 (0.8)Total873 (100.0)617 (100.0)* Previously treated group: Last intravitreal injection before initiating aflibercept biosimilar treatment. Among the 895 eyes in the switching cohort, 22 eyes had no documented prior anti-VEGF agent in the electronic medical records (patients previously treated at outside institutions without a specified agent) and were excluded from this tabulation, yielding an analyzable denominator of n = 873.†Patients who switched from aflibercept biosimilar to other agents during follow-up period. Non-pharmacological interventions (PDT, n = 3; PPV, n = 1) were excluded from analysis. Last agents before initiating aflibercept biosimilar in previously treated patients and subsequent agents after biosimilar discontinuation. * Previously treated group: Last intravitreal injection before initiating aflibercept biosimilar treatment. Among the 895 eyes in the switching cohort, 22 eyes had no documented prior anti-VEGF agent in the electronic medical records (patients previously treated at outside institutions without a specified agent) and were excluded from this tabulation, yielding an analyzable denominator of n = 873. †Patients who switched from aflibercept biosimilar to other agents during follow-up period. Non-pharmacological interventions (PDT, n = 3; PPV, n = 1) were excluded from analysis. Primary outcomes: visual and anatomical changes Overall, the mean change in BCVA from baseline to final follow-up was − 0.003 ± 0.329 LogMAR (P = 0.572), indicating stable visual acuity. The mean change in CFT was − 47.5 ± 128.5 μm (P < 0.001), representing a significant anatomical improvement (Table 3).Table 3Treatment outcomes: treatment-naïve versus switching group.ParameterTreatment-naïve (n = 188)Switching (n = 895)P value†Baseline characteristics Age (years)69.5 ± 13.870.9 ± 12.30.155 Baseline BCVA (LogMAR)0.558 ± 0.5960.616 ± 0.5950.572 Baseline CFT (μm)413.7 ± 164.8367.6 ± 148.2 < 0.001Primary visual outcome Change in BCVA (LogMAR) − 0.032 ± 0.3920.000 ± 0.3090.221 Median (IQR)0.000 (− 0.155 to 0.125)0.000 (− 0.097 to 0.097)Primary anatomical outcome Change in CFT (μm) − 93.9 ± 169.4 − 37.9 ± 116.1 < 0.001 Median (IQR) − 53.0 (− 158.0 to − 8.5) − 17.0 (− 75.0 to 12.0)Negative values for change in BCVA (LogMAR) indicate visual improvement; negative values for change in CFT indicate a reduction in retinal thickness. IQR = interquartile range.†P values compare the treatment-naïve and switching groups using the independent t-test or Mann–Whitney U test for continuous variables, as appropriate. Table 3Treatment outcomes: treatment-naïve versus switching group.ParameterTreatment-naïve (n = 188)Switching (n = 895)P value†Baseline characteristics Age (years)69.5 ± 13.870.9 ± 12.30.155 Baseline BCVA (LogMAR)0.558 ± 0.5960.616 ± 0.5950.572 Baseline CFT (μm)413.7 ± 164.8367.6 ± 148.2 < 0.001Primary visual outcome Change in BCVA (LogMAR) − 0.032 ± 0.3920.000 ± 0.3090.221 Median (IQR)0.000 (− 0.155 to 0.125)0.000 (− 0.097 to 0.097)Primary anatomical outcome Change in CFT (μm) − 93.9 ± 169.4 − 37.9 ± 116.1 < 0.001 Median (IQR) − 53.0 (− 158.0 to − 8.5) − 17.0 (− 75.0 to 12.0)Negative values for change in BCVA (LogMAR) indicate visual improvement; negative values for change in CFT indicate a reduction in retinal thickness. IQR = interquartile range.†P values compare the treatment-naïve and switching groups using the independent t-test or Mann–Whitney U test for continuous variables, as appropriate. Treatment outcomes: treatment-naïve versus switching group. Negative values for change in BCVA (LogMAR) indicate visual improvement; negative values for change in CFT indicate a reduction in retinal thickness. IQR = interquartile range. †P values compare the treatment-naïve and switching groups using the independent t-test or Mann–Whitney U test for continuous variables, as appropriate. When divided by treatment history, treatment-naïve eyes showed a mean BCVA change of − 0.032 ± 0.392 LogMAR(P = 0.265), while switching eyes demonstrated a mean change of 0.000 ± 0.309 LogMAR(P = 0.998). The difference between the two groups was not statistically significant (P = 0.221). Both treatment-naïve and switching eyes demonstrated statistically significant CFT reduction from baseline (− 93.9 ± 169.4 μm and − 37.9 ± 116.1 μm, respectively; both P < 0.001). Treatment-naïve eyes exhibited significantly greater CFT reduction compared to switching eyes (P < 0.001). Analysis by diagnostic category showed baseline BCVA and CFT variations across diagnostic groups (Table 4). Mean baseline BCVA was 0.680 ± 0.648 LogMAR in nAMD, 0.541 ± 0.496 LogMAR in RVO, and 0.481 ± 0.456 LogMAR in DME, while mean baseline CFT was 353.1 ± 152.2 μm in nAMD, 438.8 ± 171.5 μm in RVO, and 394.4 ± 116.8 μm in DME. From baseline to the final visit, visual acuity remained stable within each diagnostic group: nAMD showed a mean BCVA change of 0.009 ± 0.303 LogMAR (P = 0.456), RVO 0.013 ± 0.347 LogMAR (P = 0.624), and DME − 0.026 ± 0.323 LogMAR (P = 0.291). However, all diagnostic groups demonstrated statistically significant CFT reduction from baseline (RVO: − 72.0 ± 183.7 μm; DME: − 44.1 ± 101.5 μm; nAMD: − 38.9 ± 113.7 μm; all P < 0.001). The magnitude of CFT reduction differed significantly among the three groups (P = 0.01), with RVO patients showing the greatest anatomical improvement.Table 4Visual and anatomical outcomes after aflibercept biosimilar according to underlying diagnosis.ParameterOverall (n = 1,062)AMD (n = 611)RVO (n = 182)DME (n = 173)Baseline characteristics Baseline BCVA (LogMAR)0.602 ± 0.5910.680 ± 0.6480.541 ± 0.4960.481 ± 0.456 Baseline CFT (μm)375.2 ± 151.8353.1 ± 152.2438.8 ± 171.5394.4 ± 116.8Primary visual outcome Change in BCVA (LogMAR) − 0.006 ± 0.3250.009 ± 0.3030.013 ± 0.347 − 0.026 ± 0.323 Median (IQR)0.000 (− 0.097 to 0.097)0.000 (− 0.097 to 0.125)0.000 (− 0.097 to 0.125)0.000 (− 0.125 to 0.070) P value*0.5720.4560.6240.291Primary anatomical outcome Change in CFT (μm) − 47.5 ± 128.5 − 38.9 ± 113.7 − 72.0 ± 183.7 − 44.1 ± 101.5 Median (IQR) − 22.0 (− 88.0 to 8.0) − 17.5 (− 74.2 to 9.0) − 35.0 (− 150.5 to 14.5) − 21.0 (− 83.2 to 7.2) P value* < 0.001 < 0.001 < 0.001 < 0.001Values represent changes from baseline to the final visit. BCVA = best-corrected visual acuity; CFT = central foveal thickness; AMD = age-related macular degeneration; RVO = retinal vein occlusion; DME = diabetic macular edema; IQR = interquartile range.Negative values for change in BCVA (LogMAR) indicate visual improvement; negative values for change in CFT indicate a reduction in retinal thickness.*Paired t-test comparing baseline and final visit. Table 4Visual and anatomical outcomes after aflibercept biosimilar according to underlying diagnosis.ParameterOverall (n = 1,062)AMD (n = 611)RVO (n = 182)DME (n = 173)Baseline characteristics Baseline BCVA (LogMAR)0.602 ± 0.5910.680 ± 0.6480.541 ± 0.4960.481 ± 0.456 Baseline CFT (μm)375.2 ± 151.8353.1 ± 152.2438.8 ± 171.5394.4 ± 116.8Primary visual outcome Change in BCVA (LogMAR) − 0.006 ± 0.3250.009 ± 0.3030.013 ± 0.347 − 0.026 ± 0.323 Median (IQR)0.000 (− 0.097 to 0.097)0.000 (− 0.097 to 0.125)0.000 (− 0.097 to 0.125)0.000 (− 0.125 to 0.070) P value*0.5720.4560.6240.291Primary anatomical outcome Change in CFT (μm) − 47.5 ± 128.5 − 38.9 ± 113.7 − 72.0 ± 183.7 − 44.1 ± 101.5 Median (IQR) − 22.0 (− 88.0 to 8.0) − 17.5 (− 74.2 to 9.0) − 35.0 (− 150.5 to 14.5) − 21.0 (− 83.2 to 7.2) P value* < 0.001 < 0.001 < 0.001 < 0.001Values represent changes from baseline to the final visit. BCVA = best-corrected visual acuity; CFT = central foveal thickness; AMD = age-related macular degeneration; RVO = retinal vein occlusion; DME = diabetic macular edema; IQR = interquartile range.Negative values for change in BCVA (LogMAR) indicate visual improvement; negative values for change in CFT indicate a reduction in retinal thickness.*Paired t-test comparing baseline and final visit. Visual and anatomical outcomes after aflibercept biosimilar according to underlying diagnosis. Values represent changes from baseline to the final visit. BCVA = best-corrected visual acuity; CFT = central foveal thickness; AMD = age-related macular degeneration; RVO = retinal vein occlusion; DME = diabetic macular edema; IQR = interquartile range. Negative values for change in BCVA (LogMAR) indicate visual improvement; negative values for change in CFT indicate a reduction in retinal thickness. *Paired t-test comparing baseline and final visit. Longitudinal analysis demonstrated consistent CFT reduction across all time points and diagnostic categories (Fig. 1). Treatment-naïve patients exhibited more pronounced anatomical improvements compared to switching patients throughout the follow-up period.Fig. 1BCVA and CFT Changes by Diagnosis and Treatment History. Longitudinal changes in best-corrected visual acuity (BCVA, LogMAR) and central foveal thickness (CFT, μm) following aflibercept 2 mg biosimilar SB15 (Afilivu®) treatment. (A) BCVA changes across diagnostic groups (nAMD, RVO, DME, Other) over 6 months. (B) BCVA comparison between treatment-naïve and switching patients. (C) CFT changes by diagnosis showing anatomical improvement in all groups. (D) CFT comparison between treatment-naïve and switching patients demonstrating greater reduction in treatment-naïve eyes. Error bars represent standard deviation. BCVA and CFT Changes by Diagnosis and Treatment History. Longitudinal changes in best-corrected visual acuity (BCVA, LogMAR) and central foveal thickness (CFT, μm) following aflibercept 2 mg biosimilar SB15 (Afilivu®) treatment. (A) BCVA changes across diagnostic groups (nAMD, RVO, DME, Other) over 6 months. (B) BCVA comparison between treatment-naïve and switching patients. (C) CFT changes by diagnosis showing anatomical improvement in all groups. (D) CFT comparison between treatment-naïve and switching patients demonstrating greater reduction in treatment-naïve eyes. Error bars represent standard deviation. Fluid analysis Fluid analysis revealed significant changes in retinal fluid patterns from baseline to final visit (Fig. 2). In nAMD patients (n = 611), the proportion of eyes with SRF decreased from 51.2% at baseline to 29.6% at final visit (P < 0.001), and IRF decreased from 38.0 to 26.4% (P < 0.001). PED remained relatively stable (59.7% at baseline vs. 59.1% at final visit, P = 0.82). In RVO patients (n = 182), IRF showed the most dramatic reduction, decreasing from 92.9% at baseline to 76.4% at final visit (P < 0.001). SRF remained minimal throughout the study period. Similarly, in DME patients (n = 173), IRF decreased significantly from 90.8% at baseline to 84.4% at final visit (P = 0.002), while SRF showed minimal changes. Representative cases demonstrating anatomical improvement following aflibercept biosimilar treatment in nAMD, RVO, and DME are shown in Supplementary Fig. 1.Fig. 2Retinal Fluid Changes by Diagnosis. Prevalence of subretinal fluid (SRF, purple) and intraretinal fluid (IRF, blue) at baseline and final visit across all diagnostic groups; pigment epithelial detachment (PED, teal) is shown only for nAMD. (A) nAMD patients (n = 611) showing significant reduction in SRF and IRF. (B) RVO patients (n = 182) demonstrating marked decrease in IRF. (C) DME patients (n = 173) with modest IRF reduction. Asterisks indicate statistical significance (P < 0.05). Retinal Fluid Changes by Diagnosis. Prevalence of subretinal fluid (SRF, purple) and intraretinal fluid (IRF, blue) at baseline and final visit across all diagnostic groups; pigment epithelial detachment (PED, teal) is shown only for nAMD. (A) nAMD patients (n = 611) showing significant reduction in SRF and IRF. (B) RVO patients (n = 182) demonstrating marked decrease in IRF. (C) DME patients (n = 173) with modest IRF reduction. Asterisks indicate statistical significance (P < 0.05). To further characterize fluid response by prior treatment exposure, we compared complete fluid resolution rates between treatment-naïve and switching eyes among eyes with the respective fluid type at baseline (Table 5). Treatment-naïve eyes demonstrated significantly higher rates of complete SRF resolution compared with switching eyes (70/98 [71.4%] vs. 152/322 [47.2%], P < 0.001), as well as higher complete IRF resolution (36/82 [43.9%] vs. 116/507 [22.9%], P < 0.001). Rates of complete PED resolution were low in both groups and did not differ significantly (2/43 [4.7%] vs. 4/374 [1.1%], P = 0.119).Table 5Retinal fluid outcomes according to treatment history.ParameterTreatment-naïveSwitchingP value*Subretinal fluid (SRF) Baseline, n (%)98 (53.3)322 (36.1) Final, n (%)31 (16.8)213 (23.9) Complete resolution, n/N (%)70/98 (71.4)152/322 (47.2)< 0.001Intraretinal fluid (IRF) Baseline, n (%)82 (44.6)507 (56.8) Final, n (%)53 (28.8)421 (47.2) Complete resolution, n/N (%)36/82 (43.9)116/507 (22.9)< 0.001Pigment epithelial detachment (PED) Baseline, n (%)43 (23.4)374 (41.9) Final, n (%)43 (23.4)373 (41.8) Complete resolution, n/N (%)2/43 (4.7)4/374 (1.1)0.119Values are presented as number (percentage). Baseline and final values were calculated among eyes with paired baseline and final OCT data for each fluid parameter. Complete resolution was defined as absence of the corresponding fluid at the final visit among eyes with that fluid present at baseline.*P values compare complete resolution rates between treatment-naïve and switching eyes and were calculated using the chi-square test or Fisher’s exact test, as appropriate. Table 5Retinal fluid outcomes according to treatment history.ParameterTreatment-naïveSwitchingP value*Subretinal fluid (SRF) Baseline, n (%)98 (53.3)322 (36.1) Final, n (%)31 (16.8)213 (23.9) Complete resolution, n/N (%)70/98 (71.4)152/322 (47.2)< 0.001Intraretinal fluid (IRF) Baseline, n (%)82 (44.6)507 (56.8) Final, n (%)53 (28.8)421 (47.2) Complete resolution, n/N (%)36/82 (43.9)116/507 (22.9)< 0.001Pigment epithelial detachment (PED) Baseline, n (%)43 (23.4)374 (41.9) Final, n (%)43 (23.4)373 (41.8) Complete resolution, n/N (%)2/43 (4.7)4/374 (1.1)0.119Values are presented as number (percentage). Baseline and final values were calculated among eyes with paired baseline and final OCT data for each fluid parameter. Complete resolution was defined as absence of the corresponding fluid at the final visit among eyes with that fluid present at baseline.*P values compare complete resolution rates between treatment-naïve and switching eyes and were calculated using the chi-square test or Fisher’s exact test, as appropriate. Retinal fluid outcomes according to treatment history. Values are presented as number (percentage). Baseline and final values were calculated among eyes with paired baseline and final OCT data for each fluid parameter. Complete resolution was defined as absence of the corresponding fluid at the final visit among eyes with that fluid present at baseline. *P values compare complete resolution rates between treatment-naïve and switching eyes and were calculated using the chi-square test or Fisher’s exact test, as appropriate. Safety Among 2762 intravitreal injections administered to 1,083 eyes, no cases of intraocular inflammation (IOI) were observed throughout the study period. One case (0.09%) of rhegmatogenous retinal detachment occurred one-month post-injection in a patient with pre-existing peripheral retinal pathology, requiring pars plana vitrectomy with successful anatomical outcome. No other serious ocular or systemic adverse events related to aflibercept biosimilar were observed. However, minor ocular adverse events such as redness, pain, and watering were not systematically captured due to the retrospective nature of the study and routine clinical practice settings. The safety profile was consistent with the phase 3 clinical trial of SB15. Discussion In recent years, the anti-VEGF biosimilar landscape for retinal diseases has expanded rapidly. For ranibizumab, many biosimilars have received regulatory approval worldwide—including Razumab (Intas, India; DCGI 2015), SB11 (Byooviz/Amelivu; Samsung Bioepis, US FDA/EMA/MFDS 2021–2022), FYB201 (Cimerli/Ongavia/Ranivisio; Formycon/Bioeq, 2022), CKD-701 (LucenBS; Chong Kun Dang, 2022), Xlucane (Ximluci; Xbrane, 2022–2023), and QL1205 (Rimmyrah; Qilu, 2024)—with real-world evidence demonstrating efficacy and safety comparable to reference ranibizumab across nAMD, RVO, and DME13–16. For aflibercept 2 mg, development accelerated following the expiration of the reference product’s core market exclusivity in May 2024, and some biosimilars have since obtained approval from major regulatory agencies: MYL-1701P (Yesafili; Biocon Biologics/Viatris; EMA 2023, US FDA 2024), SB15 (Opuviz/Afilivu®; Samsung Bioepis; MFDS and US FDA 2024, EMA 2024), FYB203 (Ahzantive/Baiama; Formycon; US FDA 2024, EMA 2025), ABP-938 (Pavblu; Amgen; US FDA 2024, EMA 2025), SOK583A19 (Enzeevu/Afqlir; Sandoz; US FDA and EMA 2024), and CT-P42 (Eydenzelt; Celltrion; MFDS 2024, EMA 2025), with additional candidates (QL1207, P041, SCD411, ALT-L9, AVT06, OT-702, 9MW0813, RBS-001) either approved in individual jurisdictions or under regulatory review7. However, real-world evidence supporting long-term efficacy and safety of these biosimilars across heterogeneous retinal indications remains limited. This multicenter, retrospective real-world study evaluated the efficacy and safety of SB15 (Afilivu®) in 1,083 eyes across diverse retinal diseases. The study demonstrated that SB15 (Afilivu®) maintained stable visual acuity and achieved significant anatomical improvement from baseline to final follow-up. Treatment-naïve eyes exhibited more pronounced anatomical improvements (CFT reduction: − 93.9 ± 169.4 μm) compared to switching eyes (− 37.9 ± 116.1 μm, P < 0.001). Notably, SB15 (Afilivu®) demonstrated significant efficacy in reducing SRF and IRF in nAMD patients and IRF in RVO and DME patients. Aside from one case of rhegmatogenous retinal detachment requiring surgical repair with successful outcome, no serious adverse events were reported, confirming the favorable safety profile observed in phase 3 clinical trials. Our findings contribute real-world outcomes for SB15 (Afilivu®) across treatment-naïve and switching cohorts in three major retinal diseases. The findings of this real-world study align with the phase 3 randomized clinical trial of SB15, which demonstrated equivalent efficacy to reference aflibercept in patients with nAMD10. Moreover, the anatomical improvements observed in our study are consistent with other aflibercept biosimilar trials. The INSIGHT trial evaluating MYL-1701P in DME patients reported a mean CFT reduction of approximately 112 μm at week 817. While our DME cohort showed a mean reduction of 44.1 μm. The relatively smaller magnitude of CFT reduction in our study may reflect the predominance of switching patients (82.6%) who had already received extensive prior anti-VEGF treatment, suggesting a more treatment-experienced and potentially refractory population compared to the treatment-naïve cohorts in phase 3 trials. Furthermore, recent real-world evidence on ranibizumab biosimilar demonstrated similar findings, with stable visual acuity and reduced CFT across multiple retinal diseases14–16. Our study extends these observations to aflibercept biosimilar, confirming that biosimilar anti-VEGF agents maintain efficacy and safety in real-world clinical settings. Our results are comparable to other Korean studies using original reference aflibercept 2 mg. Choi et al. reported a mean of 6.8 injections per year with the treat-and-extend (TAE) regimen in treatment-naive exudative AMD patients, with significant improvements in BCVA and CMT18. Moon et al. demonstrated maintained efficacy over 2 years using an O-TAE strategy, with 5.11 ± 1.69 injections in the first year and 3.84 ± 2.39 in the second year19. These findings support the effectiveness of the TAE approach in real-world clinical practice. Furthermore, with the recent approval of aflibercept biosimilars, real-world data are emerging. Avadzadeh et al. reported early real-world experience with aflibercept biosimilar (Tyalia) from Iran, showing similar efficacy and safety profiles to the reference aflibercept 2 mg20 This study provides important real-world evidence supporting the clinical implementation of SB15 (Afilivu®) in routine ophthalmologic practice. The abbreviated regulatory approval pathway for biosimilars involves smaller phase 3 trials with shorter follow-up compared to reference biologics, raising some concerns among clinicians about their real-world performance5,7. Our findings demonstrate that SB15 (Afilivu®) is efficient and safe in various retinal diseases across diverse patient populations, including those with comorbidities and extensive prior treatment history, thereby addressing a critical evidence gap. The inclusion of a large switching cohort (n = 895, 82.6%) in our study is particularly relevant for current clinical practice, as many patients are being transitioned from the reference aflibercept 2 mg to SB15 (Afilivu®) following regulatory approvals and institutional policy changes. The comparable safety profile and maintained efficacy in switching patients provide reassurance to clinicians and patients regarding interchangeability, which has been a source of hesitation in biosimilar adoption21. As healthcare systems worldwide face increasing economic pressures, the validated efficacy and safety of aflibercept biosimilar in real-world settings supports its broader adoption as a cost-effective alternative to reference biologics. The safety profile in our real-world cohort was consistent with phase 3 clinical trial findings. One case of rhegmatogenous retinal detachment occurred among 2762 aflibercept biosimilar injections administered during the study period (0.036%, 3.6 per 10,000 injections). This event occurred in a patient with the pre-existing peripheral retinal lesion and was attributed to the mechanical effect of intravitreal injection rather than the pharmacological properties of the biosimilar itself. The observed rate is comparable to published large-scale registry data reporting an incidence of 1.36 per 10,000 anti-VEGF injections (95% CI 0.95–1.89), with no significant association between RRD risk and type of VEGF inhibitors22. The patient underwent successful pars plana vitrectomy and maintained stable retinal attachment for 10 months post-operatively, demonstrating favorable anatomical outcome despite this rare complication. The high rate of treatment discontinuation and switching was observed in this study, and this should be interpreted in the context of the unique regulatory and legal environment surrounding the use of aflibercept biosimilars in Korea during the study period. Following its approval in February 2024, the commercial launch and ongoing supply of the SB15 (Afilivu®) were temporarily stopped due to a preliminary injunction relating to patent issues, causing uncertainty regarding the availability of the drug and decisions regarding its inclusion in institutional formulas. Although the Seoul High Court overturned this injunction following the Korean Patent Court’s invalidation of the relevant aflibercept formulation patent, allowing market re-entry in late 2025, the interim period was characterized by forced non-medical switching driven by legal and supply constraints rather than clinical inefficacy23. Although the specific clinical rationale for each switch was not explicitly documented in the electronic medical records, the distribution of prior and subsequent anti-VEGF agents (Table 2) may provide indirect insight into real-world switching patterns during this period. Therefore, the high switching rate observed in this real-world cohort likely reflects external, policy- and legal-related factors rather than an inadequate anatomical or functional response to the biosimilar. This emphasizes the importance of interpreting real-world SB15 (Afilivu®) usage patterns within their broader healthcare and regulatory context. This study has several limitations that warrant consideration. First, the retrospective design introduces potential selection bias and limits the ability to control for confounding variables. Second, the lack of standardized treatment protocols across participating centers may introduce heterogeneity in treatment regimens and follow-up intervals, although this also reflects the diversity of real-world clinical practice. Third, the study population was limited to three tertiary referral centers in South Korea, which may limit generalizability to other geographic regions and healthcare settings. In addition, owing to the retrospective design, the specific clinical reason for each treatment switch or discontinuation could not be uniformly captured from the electronic medical records. Finally, the predominantly switching cohort (82.6%) with extensive prior anti-VEGF exposure may limit the assessment of biosimilar efficacy in treatment-naïve disease. Future studies with larger treatment-naïve populations and longer follow-up would help establish the durability of aflibercept biosimilar efficacy and inform optimal treatment strategies. In conclusion, this multicenter real-world study demonstrates that aflibercept 2 mg biosimilar SB15 (Afilivu®) maintains stable visual acuity and achieves significant anatomical improvements across multiple retinal diseases in routine clinical practice. The findings validate phase 3 clinical trial results and support the efficacy and safety of aflibercept biosimilar in both treatment-naïve and switching patient populations. The overall favorable safety profile, with one rare complication successfully managed and significant anatomical improvements across diagnostic categories, supports the clinical utility of this biosimilar. These real-world data address critical evidence gaps and provide reassurance to clinicians and patients regarding biosimilar performance outside the controlled clinical trial setting. Supplementary Information Supplementary Information 1: Figure 1. Representative spectral-domain optical coherence tomography (OCT) images illustrating anatomical changes following treatment with aflibercept 2 mg biosimilar (SB15, Afilivu®) in different retinal diseases. (A) Neovascular age-related macular degeneration (nAMD): Baseline OCT demonstrates subretinal fluid (SRF) and intraretinal fluid (IRF). After SB15 (Afilivu®) treatment, marked resolution of SRF and reduction of IRF are observed, accompanied by decreased central foveal thickness. (B) Retinal vein occlusion (RVO): Baseline OCT shows pronounced IRF consistent with macular edema and shallow SRF. Follow-up OCT reveals substantial reduction of IRF and SRF and improved retinal contour after treatment. (C) Diabetic macular edema (DME): Baseline OCT displays diffuse IRF and retinal thickening. Post-treatment OCT demonstrates partial resolution of IRF with corresponding anatomical improvement. These representative cases illustrate the anatomical efficacy of SB15 (Afilivu®)across multiple retinal diseases in real-world clinical practice. Supplementary Information 1: Figure 1. Representative spectral-domain optical coherence tomography (OCT) images illustrating anatomical changes following treatment with aflibercept 2 mg biosimilar (SB15, Afilivu®) in different retinal diseases. (A) Neovascular age-related macular degeneration (nAMD): Baseline OCT demonstrates subretinal fluid (SRF) and intraretinal fluid (IRF). After SB15 (Afilivu®) treatment, marked resolution of SRF and reduction of IRF are observed, accompanied by decreased central foveal thickness. (B) Retinal vein occlusion (RVO): Baseline OCT shows pronounced IRF consistent with macular edema and shallow SRF. Follow-up OCT reveals substantial reduction of IRF and SRF and improved retinal contour after treatment. (C) Diabetic macular edema (DME): Baseline OCT displays diffuse IRF and retinal thickening. Post-treatment OCT demonstrates partial resolution of IRF with corresponding anatomical improvement. These representative cases illustrate the anatomical efficacy of SB15 (Afilivu®)across multiple retinal diseases in real-world clinical practice.