One-Year Real-World Outcomes of Aflibercept 8 mg in Treatment-Naïve Neovascular Age-Related Macular Degeneration: A Swiss Retina Research Network Report

Abstract

Introduction

This study reports the 1-year efficacy of aflibercept 8 mg (afl8) in a real-world cohort of treatment-naïve patients with neovascular age-related macular degeneration (nAMD).

Methods

An observational, retrospective case series from nine centers of the Swiss Retina Research Network including treatment-naïve eyes with nAMD started on intravitreal afl8. Changes in visual acuity (VA), macular thickness, pigment epithelial detachment (PED) height, and retinal fluids were evaluated over 12 months and compared with baseline. Treatment intervals and safety data were recorded along with subgroup analyses based on macular neovascularization type, loading-phase completion, and treatment regimen.

Results

A total of 91 eyes met the inclusion criteria. After 1 year of treatment, VA changed by + 1.0 ± 13.2 letters (p = 0.018), mean CST changed by −110.0 ± 130.9 µm (p < 0.001), and mean PED height changed by −71.2 ± 104.8 µm (p < 0.001). After 12 months, 66.2% of eyes demonstrated a complete absence of macular fluids. Mean treatment interval was 13.9 ± 7.9 weeks, with 56.1% of eyes being extended to ≥ 12 weeks with an anatomy-driven approach. No functional or anatomical differences were observed between eyes receiving a clean loading phase and in terms of different macular neovascularization (MNV) subtypes. Two adverse events were observed.

Conclusions

The first 1-year real-world experience with afl8 in patients with nAMD revealed a robust anatomical response but only a variable and limited visual gain over 12 months due to a ceiling effect. Our findings confirm the fast and sustained macular drying reported from clinical trials, allowing for extended treatment intervals without compromising safety and efficacy.

Key Summary Points

Why carry out this study?

The CANDELA and PULSAR clinical trials demonstrated the noninferiority of aflibercept 8 mg compared with aflibercept 2 mg in naïve patients with neovascular age-related macular degeneration (nAMD).

Ongoing real-world studies are evaluating the safety and efficacy of aflibercept 8 mg in patients with nAMD, a population that is frequently undertreated relative to clinical trial participants.

What was learned from the study?

In this multicenter real-world cohort, aflibercept 8 mg achieved robust and sustained anatomical improvements with modest visual acuity changes and a favorable safety profile over 12 months.

More than half of treated eyes were extended to treatment intervals of ≥ 12 weeks using anatomy-guided strategies, supporting the feasibility of individualized interval extension in routine practice, while highlighting differences from protocol-driven clinical trials.

Introduction

Neovascular age-related macular degeneration (nAMD) is a leading cause of vision loss in elderly populations worldwide. Due to its strong association with aging, its prevalence is expected to rise substantially, representing a major global health burden [1]. The introduction of intravitreal antivascular endothelial growth factor (anti-VEGF) therapy has transformed the management of nAMD over the last two decades [2, 3]. However, the current standard of care still has limitations. These include high treatment demands, frequent monitoring and injections [4], and high economic and psychological strain [5–7].

Over the last years, aflibercept 2 mg has been extensively used as a first-line treatment supported by robust evidence from randomized clinical trials [2] and real-world evaluations [8, 9]. Despite these favorable outcomes, the concept of a higher-dose formulation was developed to further improve disease control and to allow for extended dosing intervals [10]. Increasing the aflibercept 2 mg dose to 8 mg has been shown in pharmacokinetic studies to sustain VEGF suppression below pathogenic levels for longer periods, potentially enabling greater treatment durability in clinical practice [10]. This rationale led to pivotal trials designed to confirm whether a higher dose could achieve equivalent efficacy with longer intervals between injections.

The CANDELA and PULSAR studies demonstrated the noninferiority of aflibercept 8 mg (afl8) compared with aflibercept 2 mg in terms of efficacy and safety in patients with nAMD [11, 12]. It also demonstrated a higher prevalence of patients being dry at week 16, indicating a superiority in drying efficacy. Despite similarly positive results of both formulations, the strict patient inclusion criteria and protocol-driven standardized regimens of clinical trials unavoidably limit the generalizability of their results. Validation from real-world studies is therefore essential to corroborate pivotal trial findings by translating them into routine clinical care, where therapeutic decisions are individualized according to patient characteristics and disease activity.

A few real-world observational studies have evaluated afl8, mainly focusing on the available outcomes until the end of the loading dose [13–16] or on intermediate term [17], often combining treatment-naïve and previously treated eyes. These early results indicate rapid anatomic improvements and fluid resolution, which are known predictors of long-term functional stability [18, 19]. Rapid fluid control could thus be indicative of longer durability, which is a critical endpoint in the treatment of nAMD. However, no published studies have yet reported 12-month outcomes in exclusively treatment-naïve eyes, leaving an important evidence gap regarding the real-world performance of afl8.

This multicenter study, conducted by the Swiss Retina Research Network (SRRN), aimed to evaluate the 12-month real-world outcomes of afl8 in treatment-naïve patients with nAMD, focusing on visual and anatomic efficacy, treatment durability, and safety.

Methods

This observational, retrospective, noncomparative, longitudinal multicentric study collected data from nine specialized retina centers in Switzerland: Swiss Visio Retina Research Center, Lausanne (A.A., N.B.); Berner Augenklinik, Bern (J.G.G., I.B.P., C.S.); Istituto delle Neuroscienze cliniche della Svizzera Italiana, Ente Ospedaliero Cantonale (M.M., G.G., A.P.); Augenarzt Praxisgemeinschaft Gutblick AG, Switzerland (M.R.M., A.T.); Jules Gonin Eye Hospital, University Hospital, Lausanne (C.M.E., E.C.D.O.F.); Stadtspital Triemli, Zürich (G.M.S., T.S.); Kantonsspital St. Gallen (A.E.); Talacker Augen Zentrum Zürich (MML, AK); and University Hospital Zürich, Zürich (S.Z., F.G.).

This study was approved by the ethics committees of all centers, under the lead of the Ethics Committee of the canton of Bern (registration no. 2024-01026). Coded patient data were included in the analysis on the basis of general or informed consent, in accordance with the International Council for Harmonisation E6 Good Clinical Practice Guideline, the Declaration of Helsinki, and Swiss federal laws.

A consecutive series of eyes of treatment-naïve patients with active nAMD (defined by the presence of either intraretinal fluid, subretinal fluid, exudates, and/or hemorrhage) who received intravitreal afl8 and had been followed for 12 ± 1 months by September 2025 were included in the study. The inclusion of both eyes was accepted if both eyes met the inclusion criteria. In patients with bilateral inclusion, intravitreal injections were administered according to routine clinical practice at each participating center. In most cases, bilateral injections were performed on the same day; however, same-day bilateral treatment was not mandated by protocol. Each eye was analyzed independently, and no inter-eye comparative analysis was performed. Exclusion criteria were the following: patient refusal to grant consent for the use of their coded data; subretinal bleeding of more than one optic nerve head diameter at baseline; preexisting macular damage with no residual functional potential; active systemic disease potentially interfering with treatment outcomes (e.g., local or systemic rheumatoid disease, and/or vasculitis requiring antiinflammatory or immunomodulatory treatment); any opacifications in the optic axis with relevance for visual function prohibiting ocular imaging and fundoscopy; and any intraocular surgery within 3 months from baseline (except YAG laser capsulotomy). Most patients initiated therapy with a planned loading phase of three consecutive monthly afl8 injections. In some cases, the loading schedule was adapted at the physician’s discretion, with an earlier transition to an individualized regimen once anatomical inactivity was achieved. After the loading or adjusted induction phase, patients were followed by a standard treat-and-extend (T&E) regimen with an interval adjustment of 2 weeks based on disease activity, or by monthly observations until relapse of disease activity (“observe and plan” regimen), after which the interval was adjusted by 2 weeks on the basis of disease activity.

Clinical data were retrospectively collected from electronic medical records at baseline and at months 1, 3, 6, and 12 after treatment initiation. Data after switching to other molecules were not recorded, but reasons for switching were documented. Macular neovascularization (MNV) type was exclusively assessed at baseline. The presence of macular atrophy (MA) was assessed at baseline and after 12 months. MA was defined as the loss of outer retinal layers and/or the retinal pigment epithelium (RPE) on macular spectral-domain optical coherence tomography (SD-OCT) scans, assessed within 500 µm from the foveal center. The following parameters were retrieved at all timepoints: Snellen corrected visual acuity (VA), central retinal thickness (CRT), central subfield thickness (CST), maximal pigment epithelial detachment (PED) height, and presence/absence of intraretinal fluid (IRF) and subretinal fluid (SRF) in the central 1 mm² of the Early Treatment of Diabetic Retinopathy Study (ETDRS) grid (“central subfield”). For statistical analysis, Snellen VA was converted to ETDRS letters, where a Snellen VA of 1.0 corresponded to 85 ETDRS letters [20]. Hand motion and counting fingers were categorized using the Freiburg Visual Acuity Test (FraCT) guidelines [21]. Although VA values were converted to ETDRS letters for analysis, VA testing methods were not fully standardized across centers, reflecting routine clinical practice. The total number of injections and the last interval at the 1-year observation time point were also recorded.

Primary outcomes were the mean changes in visual acuity, macular thickness, and injection intervals at 1 year. Secondary outcomes included changes in macular fluids in OCT after the first injection, at 3–4 months, and at month 12, and ocular side effects. Moreover, subgroup analyses were performed on the basis of the MNV type and treatment regimen adopted. For eyes that underwent cataract surgery during the treatment period, subsequent data were censored from analysis after the date of surgery to exclude an inherent bias. Data from eyes switched to other anti-VEGF agents during the observation period were excluded from the analysis after the switch.

Descriptive statistics, subgroup comparisons, and correlation and regression analyses were performed. The Shapiro–Wilk test was used to test the distribution of data and revealed that the data were not normally distributed. Therefore, the data are presented as mean ± standard deviation (SD) as well as median and interquartile ranges (IQR). Nonparametric tests were used for comparisons. Longitudinal changes in functional and morphological parameters from baseline to 12 months were analyzed using the Wilcoxon signed-rank test for continuous data and the McNemar test for dichotomous variables. The Friedman test for dependent samples was used to compare multiple timepoints in the longitudinal analysis. The level of significance was set at p < 0.05. All statistical analyses were performed using SPSS Statistics V.27 (IBM Corp., Armonk, NY, USA) and R (version 4.5.1; R Foundation for Statistical Computing, Vienna, Austria).

Results

A total of 122 eyes with nAMD treated with afl8 were screened for inclusion. Of these, 31 eyes were excluded. Reasons for exclusion were follow-up less than 12 months (n = 13, 41.9%), patient’s refusal to grant consent for data use (n = 12, 38.7%), any intraocular surgery within 3 months prior to treatment initiation (n = 1, 3.2%), and visual acuity inferior to 0.1 at the time of diagnosis (n = 5, 16.1%).

The 91 remaining eyes (81 patients; 12.3% bilateral) were included in the study. Baseline demographic and clinical data for the study population are presented in Table 1.

Table 1.

Baseline patient demographic and study eye characteristics

	Number	n (%)
Patient demographics
Total eyes (patients)	91 (81)
Sex
Female		51 (63)
Male		30 (37)
	Median [IQR]
Age, years	80.0 [76, 85.5]
Baseline clinical features
Lens state
Phakic (from baseline till end of observation)		26 (28.6)
Pseudophakic		61 (67.0)
Cataract surgery after treatment initiationa		3 (3.3)
Treatment regimen (induction phase)
Loading dose (×3 monthly IVT)		59 (64.8)
Treatment regimen (maintenance phase)
Observe and plan		29 (31.9)
Treat and extend		62 (68.1)
MNV type
Type 1		49 (53.8)
Type 2		14 (15.4)
Mixed (type 1 + 2)		8 (8.8)
Type 3		14 (15.4)
PCV		4 (4.4)
Not gradableb		2 (2.2)
Visual acuity (ETDRS)		68.4 ± 14.3 (69.9, 63.0–80.2)
Central retinal thickness (mm)		375.0 ± 163.7 (329.0, 267.0–451.5)
Central subfield thickness (mm)		390.6 ± 142.4 (349.5, 290.8–469.3)
Presence of macular atrophy		36 (39.6)
Presence of PED		67 (73.6)
Maximal PED height in the central 3 mm (mm)		226.1 ± 181.2 (170.0, 113.0–270.0)
Presence of macular fluid (central 1 mm)		68 (76.4)
Subretinal		37 (41.5)
Intraretinal		15 (16.8)
Both SRF/IRF		16 (18.1)

Functional and anatomical data are presented as mean ± SD [median, IQR]

IQR interquartile range, IRF intraretinal fluid, IVT intravitreal injection, PED pigment epithelium detachment, MNV macular neovascularization, SRF subretinal fluid

^aFor eyes undergoing cataract surgery during the treatment period, data were censored after surgery

^bMNV type was classified as “not gradable” in cases with insufficient image quality or ambiguous lesion morphology precluding reliable classification

Functional Outcomes

Mean VA changed from 68.4 ± 14.3 (median 69.9, IQR 63–80.2) ETDRS letters at baseline to 69.6 ± 14.7 (median 75, IQR 65.1–80.2) ETDRS letters after a single afl8 injection, further to 70.2 ± 15.5 and 71.5 ± 15.6 ETDRS letters at 3–4 and 6 months of follow-up, respectively. After 12 months, a mean VA of 68.8 ± 19.9 (median 75.0, IQR 65.1–85.0) ETDRS letters was recorded (+ 1.0 ± 13.2 letters from baseline, p = 0.018) (Fig. 1).

Fig. 1.

Functional outcomes. Changes in mean ± SD of visual acuity (VA) in ETDRS letters from baseline to 12-month follow-up

Eyes were further classified on the basis of the extent of visual change from baseline to 12 months. At 1 year, VA was stable between −4 and +4 ETDRS letters in 26 eyes (28.6%), and an improvement of 5, 10, and 15 letters was recorded in 6 (6.6%), 10 (11%), and 9 (9.9%) eyes, respectively. A 5-letter VA drop was recorded in 7 eyes (7.7%), and 13 eyes (14.3%) lost 10 letters or more over 1 year.

VA at 12 months was positively correlated with baseline VA (r(71) = 0.75, p < 0.001) and negatively correlated with baseline MA (r(68) = −0.37, p = 0.002).

Anatomical Outcomes

Overall, CRT decreased from a baseline of 375.0 ± 163.7 (median 329.0, IQR 267.8–451.5) µm to 265.6 ± 80.3 (median 243.0, IQR 214.0–308.0) µm after one injection, to 266.6 ± 81.2 (median 237.0, IQR 210.0–297.8) µm at 3–4 months (p < 0.001), to 261.3 ± 69.4 µm at 6 months (p < 0.001), and finally to 265.2 ± 73.0 (median 247.5, IQR: 208.5–301.5) µm at month 12 (p < 0.001) (Fig. 2A). The mean change in CRT was of −108.1 ± 148.2 (median −62.0, IQR −133.0 to −11.0) µm at month 12 (p < 0.001). Similarly, a statistical significant change in CST was observed at all timepoints: from a baseline value of 390.6 ± 142.4 (median 349.5, IQR 290.8–469.3) µm to 283.5 ± 75.4 (median 270.5, IQR 235.3–311.0) µm after one injection, to 283.3 ± 78.4 (median 269.0, IQR 240.0–318.5) µm at 3–4 months (p < 0.001), and to 277.3 ± 72.8 (median 266.5, IQR 230.5–316.0) µm at month 12 (p < 0.001) (Fig. 2B). The mean CST change was of −110.0 ± 130.9 (median −62.0, IQR −165.0 to −30.0) µm at month 12 (p < 0.001). VA at 12 months was negatively correlated with baseline CRT (r(70) = −0.43, p = 0.002) and CST (r(70) = −0.49, p < 0.001). Maximal PED height decreased from a baseline of 226.1 ± 181.2 (median 170.0, IQR 113–270) µm to 139.3 ± 79.9 (median 125.0, IQR 92.0–151) µm at month 12 (p < 0.82).

Fig. 2.

Anatomical outcomes. Changes in mean ± SD of central retinal thickness (CRT) (A) and central subfield thickness (CST) (B) from baseline to 12-month follow-up

Qualitative fluid assessments revealed retinal fluid presence in the central 1 mm zone in 71 out of 89 eyes (78.1%) at baseline. For the 18 remaining cases, 16 cases had fluid in the outer zones, and 2 had hemorrhages. After one afl8 injection (n = 87), 49 eyes (55.1%) achieved complete anatomical dryness. At month 12 (n = 71), 47 eyes (51.6%) were dry, while 24 eyes had residual macular fluid; specifically, 7 eyes (29.2%) had SRF only, 15 eyes (62.5%) had IRF only, while 2 eyes (8.3%) had both intra- and subretinal fluids (Fig. 3).

Fig. 3.

Qualitative analysis of fluid dynamics. Proportion of eyes with subretinal fluid (SRF) only, intraretinal fluid (IRF) only, both SRF and IRF, or dry macula from baseline through 12-month follow-up

After 1 year of treatment with afl8, MA was present in a total of 42 eyes (46.2%). New MA developed in 10 eyes. Two eyes were classified as type 1 MNV and 2 eyes as PCV at baseline, while 3 eyes were categorized as type 2, and 3 eyes as type 3 MNV.

Treatment Burden and Discontinuation

During the 12-month observation period, 73 eyes (80.3%) were maintained on afl8, reaching a mean interval of 13.9 ± 7.9 (median 12.0, IQR 8.3–14.9) weeks after 7.1 ± 1.9 (range 3–13) injections. Eight eyes (11.0%) had a final interval < 8 weeks, while 24 eyes (32.9%) were on an interval between 8 and 11 weeks. A total of 41 eyes (56.1%) achieved an interval of ≥ 12 weeks, and 25 eyes (34.2%) reached a final interval from 12 to 15 weeks, while 16 eyes (21.9%) achieved a final interval of > 16 weeks (Fig. 4). In eight eyes (8.8%), the treatment regimen was converted over follow-up from a T&E to a pro re nata approach after demonstrating complete and sustained disease inactivity over multiple visits. None of these eyes showed relapse during the 12-month observation period.

Fig. 4.

Durability outcomes. Proportion of patients with a treatment interval of less than 8 weeks, between 8 and 11 weeks, between 12 and 15 weeks, or greater than 16 weeks at 12 months (A). Proportion of eyes with no macular fluid at month 12 within each treatment interval category (B)

Treatment was discontinued in 18 eyes (19.7%). Reasons for discontinuation are listed in Table 2. Reasons for switching to other agents were the occurrence of adverse events (n = 2, 16.7%), insufficient response to afl8 (n = 8, 66.7%), and patient’s wish (n = 2, 16.7%).

Table 2.

Reasons for treatment discontinuation over follow-up (n = 18 eyes of n = 98 eyes total)

Reason	n (%)
Switch to another agent	12 (13.2)
Aflibercept 2 mg	3 (25)
Faricimab	7 (58.3)
Ranibizumab	2 (16.7)
Intraocular inflammation	1 (1.1)
No potential for visual improvement	1 (1.1)
Patient deceased	4 (4.4)

During the observation period, ocular adverse events occurred in two eyes (2.2%): one eye (1.1%) developed a retinal pigment epithelium tear, which was detected at the 3-month follow-up; one eye (1.1%) developed intraocular inflammation (anterior uveitis and vitritis) after 6 months of treatment.

MNV-Type and Treatment Regimen Subgroup Analyses

When comparing the subgroup of eyes with a type 1 MNV at baseline (n = 49 out of 91 eyes, 53.8%) with the remaining study population (n = 42, 46.2%), no statistically significant differences were observed in terms of final visual acuity, CRT, CST, drying, treatment interval, and total injection number (p > 0.05).

In total, 59 out of 91 eyes (64.8%) completed a strict loading phase of 3 afl8 injections. The mean time between the first and second injection was 4.4 ± 0.8 (median 4, IQR 4–4.6) weeks, whereas between the second and third injection it was 5.5 ± 3.2 (median 4.3, IQR 4–6) weeks. The mean time between the first and third injection was 9.9 ± 3.5 weeks (median 8.9, IQR 8–10). At 1 year of follow-up, no differences were found in terms of visual acuity, CRT, CST, PED height, drying, and interval between the subgroup of eyes receiving a clean loading dose and those who did not (p > 0.05).

During the maintenance phase, an “observe and plan” treatment regimen was adopted in 29 eyes (31.9%), while the remaining 62 eyes (68.1%) were treated according to a T&E protocol. At baseline, the T&E subgroup had a higher mean CST compared with the observe and plan subgroup (409.0 versus 349.9 µm, respectively, p < 0.001). At 1 year of follow-up, no differences were found in terms of visual acuity, CRT, CST, drying, final treatment interval, and number of injections received in 1 year (p > 0.05).

Discussion

In this multicenter, real-world analysis from the Swiss Retina Research Network, afl8 demonstrated functional and anatomic benefits in treatment-naïve nAMD eyes over 1 year, with extended treatment intervals and a favorable safety profile. These findings align with and expand upon the results of pivotal phase 3 trials by validating the performance of the high-dose formulation under routine clinical conditions.

In the PULSAR randomized controlled trial, mean VA gains at week 48 were +6.7 and +6.2 letters for afl8 administered every 12 or 16 weeks, respectively, while in our cohort, mean VA change at 12 months was + 1.0 ± 13.2 letters [12]. Although the overall visual gain was small in our cohort, this difference can be largely explained by a ceiling effect, as baseline VA in our population (68.4 letters) exceeded that of PULSAR (59.6 ± 13.3 letters). It is indeed well known that eyes entering anti-VEGF treatment with higher vision have limited potential for further improvement [22]. Nonetheless, it should be noted that visual acuity was maintained throughout follow-up, with no significant change from baseline. At 12 months, the median VA remained high at 75 letters, exceeding the average values reported in pivotal trials, although no direct comparison between study populations can be made. A further consideration is that, in routine clinical practice, VA is not always assessed using standardized protocols, introducing potential measurement variability that may have attenuated the observed functional changes.

In our cohort, visual acuity showed a slight improvement up to 6 months, followed by a modest decline toward month 12. This pattern mirrors real-world observations that, despite stable anatomic findings, structural disease progression such as macular atrophy or fibrosis can limit long-term visual recovery. Notably, in our study, the development of macular atrophy was negatively correlated with visual gain at 12 months, suggesting that retinal tissue loss may partially explain the late visual trend. Among eyes without baseline atrophy, ten developed new atrophic changes during follow-up, most commonly within type 2 and type 3 MNV. This aligns with findings by Cho et al., who demonstrated that the development and progression of macular atrophy vary according to MNV subtype [23].

Consistent with PULSAR, our study documented robust anatomic responses with a mean decrease in CRT by 110 µm and in PED height by 87 µm over the first year. These changes are comparable to the mean CRT reductions reported in PULSAR at week 48, supporting a sustained anatomic effect of afl8 also under real-world conditions. Rapid drying was evident after the first injection, with two-thirds of eyes remaining fluid-free at 1 year. This rate is in line with pivotal trial trends and early real-world reports [13–17], supporting the durability potential of afl8 mg in routine clinical practice..

A key rationale for developing high-dose aflibercept was to achieve improved durability. In PULSAR, 83% of eyes treated with afl8 maintained ≥ 12-week intervals by week 48 [12]. In our real-world cohort, 56.1% of eyes achieved ≥ 12-week intervals, and 21.9% were extended ≥ 16 weeks at 12 months. Importantly, in our study, extension intervals were supported by anatomical evidence, as more than 80% of eyes extended beyond 16 weeks showed complete fluid resolution (Fig. 4), compared with approximately 67% in PULSAR at month 12.¹² This suggests a more anatomy-driven approach in daily practice, where OCT findings rather than fixed protocols determine interval length. As a result, real-world strategies may favor slightly shorter intervals to ensure disease stability and minimize undertreatment risk.

It remains unclear to what extent treatment regimen differences, such as adherence to a clean loading phase or the use of T&E versus O&P strategies, may influence real-world outcomes. To explore this aspect, subgroup analyses were performed showing no significant differences in 12-month visual or anatomic results between eyes with type 1 MNV and those with other lesion subtypes, or between eyes with a clean loading phase and those with minor deviations. Similarly, outcomes were comparable between T&E and O&P regimens, with no significant difference at 12 months. Although these exploratory findings seem to support the robustness and flexibility of afl8 across individualized treatment strategies, they should be interpreted with caution and confirmed in larger prospective cohorts.

In terms of safety, afl8 demonstrated a reassuring profile in our cohort. Two ocular adverse events (2.2%) were recorded (i.e., one RPE tear and one sterile intraocular inflammation) with no occlusive vasculitis or endophthalmitis. A total of 18 eyes (19.8%) discontinued treatment within the first year, most commonly switching to another anti-VEGF agent. This relatively high switching rate likely reflects the availability of faricimab during the study period, prompting early changes, especially among “frequent flyer” eyes requiring shorter intervals despite afl8. Such behavior typifies competitive real-world markets and should be considered when interpreting 1-year retention rates.

This study has several limitations. The retrospective design and lack of a 2 mg comparator limit causal inference. Heterogeneity in visual acuity assessment, with centers using different refraction standards, may have also introduced variability despite standardized conversions. Minor differences in retreatment criteria and follow-up intervals also reflect real-world heterogeneity. Approximately one-fifth of eyes lacked 12-month data, introducing potential attrition bias. Lastly, the modest sample size restricted statistical power for detailed subgroup analyses by MNV type and treatment regimen, which should therefore be interpreted as exploratory.

Conclusions

Aflibercept 8 mg demonstrated sustained anatomic efficacy, visual maintenance, and durability in a real-world Swiss cohort of treatment-naïve nAMD eyes. Compared with PULSAR, our results confirm a solid control of disease activity, with a high proportion of eyes remaining dry at extended intervals, underscoring the effectiveness of anatomy-guided regimen individualization in daily clinical practice.

Author Contributions

Conception and design: Gabriela Grimaldi, Nicolò Bartolomeo, Aude Ambresin, and Moreno Menghini. Data collection: Gabriela Grimaldi, Nicolò Bartolomeo, Justus G. Garweg, Isabel B. Pfister, Christin Schild, Arianna Peyla, Anne Tillmann, Eva Cristina De Oliveira Figueiredo, Tahm Spitznagel, Alice M. Kitay, Felix Gabathuler, Malaika Mihal Kurz-Levin, Sandrine Zweifel, Chiara M. Eandi, Gábor M. Somfai, Marion R. Munk, Andreas Ebneter, Aude Ambresin, and Moreno Menghini. Analysis and interpretation: Gabriela Grimaldi, Nicolò Bartolomeo, Isabel B. Pfister, Aude Ambresin, and Moreno Menghini. All the authors participated in drafting and critically revising the final version of the manuscript before submission.

Funding

No funding was received for the conduct of this study or for the preparation of this manuscript. The Rapid Service Fee was funded by the authors.

Data Availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Conflict of Interest

Gabriela Grimaldi: consultant for Apellis, Bayer, and Roche. Lecturer for AbbVie, Apellis, Bayer, and Roche. Grant support from Bayer and Roche. Justus G. Garweg: advisor and speaker for several pharmaceutical companies. Contributes to several international industry-sponsored clinical studies in the fields of retinal disease and uveitis (AbbVie, Bayer, Novartis, Roche, Amgen, MS Pharma). This manuscript is independent of these activities. Sandrine Zweifel: consultant and advisor for Alcon, Allergan, Apellis, Bayer, Endogena, Novartis, Roche, and Zeiss, and grant support from Bayer, Novartis, and Roche, outside the submitted work. Andreas Ebneter: Advisor for Bayer Schweiz AG, Apellis Inc. and Roche Schweiz AG. He is a former employee of F. Hoffmann-La Roche Ltd. Gábor M. Somfai: consultant for Apellis, Allergan, Bayer, Carl Zeiss, and Roche. Chiara M. Eandi: advisory board participant and lecturer for AbbVie, Apellis, Astellas, Bayer, and Roche. Marion R. Munk: consultant for Abbvie, Allergan, Acucela, Aviceda Therapeutics, Alimera, Bayer, Lumithera, OD-OS, Isarna Therapeutics, Roche, Novartis, Oculis, Ocuterra, genesight Therapeutics, Kubota, Böhringer-Ingelheim, Eyepoint, Ocular Therapeutix, Apellis, Astellas, Zeiss, RetinAI, and Dandelion. Aude Ambresin: consultant for Bayer, Novartis, Roche, Apellis, and Earlysight; lecturer fees from Optovue, AbbVie, Novartis, Bayer, Roche, Apellis, Astellas, and RetinAi. Moreno Menghini: consultant for Alcon, Apellis, Astellas, Bayer, and Roche. Lecturer for Bayer and Roche. Grant support from Bayer and Roche. Nicolò Bartolomeo, Isabel B. Pfister, Christin Schild, Arianna Peyla, Anne Tillmann, Eva Cristina De Oliveira Figueiredo, Tahm Spitznagel, Alice M. Kitay, Felix Gabathuler, and Malaika Mihal Kurz-Levin have nothing to disclose.

Ethical Approval

The study was approved by all participating ethics committees, led by the ethics committee of the canton of Berne (registration no. 2024-01026). Coded patient data were included on the basis of general or specific informed consent, in line with ICH-GCP E6, the Declaration of Helsinki, and Swiss law.