Rapid Effects on Anatomical and Functional Outcomes following Upload with Faricimab in Treatment-Naïve nAMD: German Single-Center Study 45DRY

Abstract

Introduction

The bispecific antibody faricimab inhibits vascular endothelial growth factor A (VEGF-A) and angiopoietin-2 (Ang-2), two key mediators in the pathophysiology of neovascular age-related macular degeneration (nAMD), which is characterized by choroidal neovascularization, vascular instability, leakage, and intra/subretinal fluid with consequent vision loss. The aim of our study was to assess the effectiveness of faricimab at week 16 versus baseline in patients with treatment-naïve nAMD managed under real-world conditions in Germany.

Methods

45DRY was a retrospective, single-center analysis of consecutive patients with treatment-naïve nAMD initiating faricimab (four monthly injections at weeks 0, 4, 8, and 12). Outcomes were evaluated at baseline and week 16 (± 14 days). The primary endpoint was change from baseline in central retinal thickness (CRT); secondary endpoints were change in best-corrected visual acuity (BCVA) and presence/absence of intraretinal fluid (IRF), subretinal fluid (SRF), combined IRF + SRF, and subretinal pigment epithelium (sub-RPE) fluid. Analyses were descriptive.

Results

A total of 45 eyes from 45 patients were included (mean age 82.1 years; 51.1% female). Mean CRT decreased by −140.8 μm from 343.1 µm at baseline to 202.4 µm at week 16 [95% confidence interval (CI) −178.2 to −103.3]. Mean BCVA improved by −0.15 logMAR at week 16. Among eyes with retinal fluid at baseline, IRF resolved in 100% (24/24), SRF in 94.1% (32/34), combined IRF + SRF in 100% (17/17), and sub-RPE fluid in 90% (9/10). Maintenance of fluid-free status was universal in eyes negative at baseline for IRF (21/21), SRF (11/11), and combined IRF + SRF (28/28); sub-RPE fluid remained absent in 100% (35/35).

Conclusions

In routine clinical practice, a four-dose faricimab upload per label at the time of study conduct led to a rapid and substantial anatomical drying effect at week 16 with improvement in BCVA and marked CRT reduction. These real-world findings confirm that faricimab rapidly and effectively reduces retinal fluid, in line with post hoc analyses of the pivotal TENAYA and LUCERNE trials.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40123-026-01385-2.

Plain Language Summary

Neovascular age-related macular degeneration is a common eye disease in older adults and a major cause of vision loss. The disease occurs when abnormal blood vessels grow beneath the retina and leak fluid, which damages the central part of the retina that is needed for reading and recognizing faces. This study examined how well faricimab, a medicine that targets two biological pathways involved in blood vessel growth and leakage, works during the initial treatment phase in everyday clinical practice. Doctors in a German outpatient eye clinic treated patients who had not received previous eye injections with four monthly injections of faricimab and compared eye scans and vision tests before treatment and 16 weeks later. The study showed a strong reduction in retinal swelling, indicating rapid control of disease activity. Fluid within or beneath the retina disappeared in almost all eyes that showed fluid at the start of treatment, and eyes without fluid at baseline remained fluid-free. Vision improved by about two lines on a standard eye chart, even though many patients had cataracts or corneal changes that can limit visual improvement. These findings show that, in routine clinical care, early treatment with faricimab can quickly dry the retina and lead to meaningful improvements in vision. The results support the effectiveness of faricimab during the initial treatment phase for patients with neovascular age-related macular degeneration and help to expand real-world evidence beyond clinical trials.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40123-026-01385-2.

Key Summary Points

Why carry out this study?

Neovascular age-related macular degeneration (nAMD) is a leading cause of visual impairment and blindness in the elderly.

The aim of our retrospective single-center study was to investigate the rapid effectiveness of faricimab at the end of the upload phase (after four injections) in patients with treatment-naive nAMD in real-world clinical practice in Germany.

What was learned from the study?

Mean central retinal thickness (CRT) decreased by −140.8 μm from 343.1 µm at baseline to 202.4 µm at week 16 (primary endpoint), best-corrected visual acuity (BCVA) improved, and retinal fluid largely resolved [absence of intraretinal fluid (IRF) achieved in 100% of baseline-positive eyes; absence of subretinal fluid (SRF) in 94.1%; absence of IRF + SRF in 100%; absence of subretinal pigment epithelium (sub-RPE) fluid in 90%].

At the early post-upload assessment at week 16, faricimab resulted in a CRT reduction and an improvement in BCVA, both clinically meaningful, accompanied by substantial retinal fluid resolution in patients with treatment-naïve nAMD in real-world clinical practice.

Introduction

Neovascular (or “wet”) age-related macular degeneration (nAMD) remains one of the leading causes of significant visual impairment and blindness among the elderly in developed countries [1]. The hallmark of nAMD is the uncontrolled proliferation of pathologic neovessels from the choroid into the macula, predominantly driven by dysregulated angiogenic signaling—most notably vascular endothelial growth factor A (VEGF-A) [2]. These fragile neovessels are prone to leakage and hemorrhage, resulting in intra- and subretinal fluid accumulation and, if untreated, irreversible loss of central visual acuity [3]. While VEGF-A is a primary mediator of angiogenesis, angiopoietin-2 (Ang-2) has emerged as a critical regulator of vascular instability and inflammatory vascular leakage. In pathologic retinal conditions such as nAMD, Ang-2 is often upregulated, where it functions to antagonize the stabilizing angiopoetin-1 (Ang-1)/Tie-2 signaling axis, thereby promoting vascular destabilization, leakage, and neovascular sprouting [4].

Early anatomical response, commonly understood as the rapid resolution of retinal fluid, has emerged as a key prognostic marker in the management of nAMD [5–7]. Retinal thickness and retinal fluid are anatomical measures of disease activity and have been closely linked to visual outcomes, with persistent intraretinal fluid associated with poorer long-term vision [5]. Consequently, achieving and maintaining dry macula early in the treatment course is considered an important therapeutic goal. Rapid macular drying in the first months of therapy has been linked to more favorable long-term visual prognoses, since persistent IRF is a known risk factor for visual decline [5, 8]. Moreover, when patients perceive visible improvement early, adherence is more likely, as patient motivation is reinforced by early benefit [9–11].

Faricimab (Vabysmo^®, Roche Registration GmbH, Germany), the first bispecific antibody targeting both VEGF-A and angiopoietin-2, has demonstrated robust anatomic efficacy across pivotal phase III trials in nAMD, diabetic macular edema (DME), and retinal vein occlusion (RVO) [12–17]. In post hoc analyses of the pivotal TENAYA/LUCERNE trials in nAMD, Cheung et al. reported a faster time to first absence of retinal fluid with faricimab compared with aflibercept 2 mg, supporting its potential to induce early and sustained drying [7]. Similarly, findings from the PROOF study with anti-VEGF therapy indicated that patients achieving earlier fluid resolution experienced greater visual acuity gains and maintained superior outcomes over time [6]. Early anatomical response may also promote better adherence, as early anatomical success reinforces patient confidence and engagement with ongoing therapy [9–11]. Collectively, these data suggest that early fluid resolution may contribute to durable functional and adherence benefits in nAMD management.

Numerous real-world studies have demonstrated the safety and efficacy of faricimab outside controlled trial settings [18, 19]. Most available or ongoing German studies have reported outcomes in heterogeneous nAMD populations, including both treatment-naïve and previously treated eyes [20], or in switch-only cohorts [21–26]. However, for Germany, data from cohorts composed exclusively of patients with treatment-naïve nAMD remain limited [27].

The aim of our retrospective, single-center study was to complement the existing body of real-world evidence for patients with treatment-naïve nAMD with data from a German outpatient ophthalmology practice. Special emphasis was placed on the fast effectiveness of faricimab assessed at the week-16 visit following completion of the “upload” phase (initial monthly loading phase of four consecutive intravitreal injections) using Vabysmo^® 120 mg/mL solution for injection at a dose of 6 mg (0.05 mL) administered intravitreally every 4 weeks (monthly) for the first four doses. The regimen aligned with faricimab’s label for nAMD valid in the European Union (EU) at the time of its approval in 2022 and preceded the 2024 EU-label update introducing a three-dose upload regimen. Anatomical outcomes included central retinal thickness (CRT) and the presence or absence of intraretinal and subretinal fluid compartments, while functional change was evaluated in terms of best-corrected visual acuity (BCVA).

Methods

Study Design and Participants

45DRY was a retrospective, single-center study conducted at Augenzentrum Nordwest, Rostock, Germany. Patient data from routine clinical practice between 1 February 2023 and 1 February 2025 were analyzed.

The inclusion criteria for this study were: (1) patients with treatment-naive nAMD who received faricimab as first-line treatment between 1 February 2023 and 1 February 2025. (2) Optical coherence tomography (OCT) available at baseline and at week 16 (±14) days; baseline was defined as the date when a patient initiated treatment with faricimab in one eye. As only one eye per patient was treated in this retrospective research project, there was no need to define an index date at eye level. (3) Provision of written consent for the use of health-related data for this research. No exclusion criteria were applied.

This research was conducted in full compliance with the Guidelines for Good Pharmacoepidemiology Practice (GPP) published by the International Society of Pharmacoepidemiology (https://www.pharmacoepi.org/resources/policies/guidelines-08027), the laws and regulations in Germany as applicable, and the Declaration of Helsinki. The research plan and relevant supporting information was approved by the competent Ethics Committee of the University Rostock on 2 May 2025 (A 2025–0115), before the research was initiated. All patients provided consent for the use and anonymization of the data and explicitly consented to any secondary use of their data.

Study Objectives and Endpoints

The primary objective of this study was to evaluate the effectiveness of faricimab in reducing CRT at week 16. Secondary objectives were the proportion of eyes exhibiting intraretinal fluid (IRF), subretinal fluid (SRF), combined IRF and/or SRF, and subretinal pigment epithelium (sub-RPE) fluid, as well as the assessment of BCVA improvement at week 16.

The primary endpoint of this study was the change in CRT from baseline to week 16. CRT was measured in micrometers (μm), with assessments performed at baseline and week 16 using spectral-domain optical coherence tomography (SD-OCT). The date of each CRT evaluation was documented.

The secondary endpoints included both functional and anatomical parameters. Functional response was assessed by the change in BCVA from baseline to week 16, recorded in logarithm of the minimum angle of resolution (logMAR) and Snellen decimal equivalents. Dates of BCVA measurements were also captured.

Anatomical secondary endpoints comprised the proportion of eyes with presence or absence of IRF, SRF, combined IRF and/or SRF, and sub-RPE fluid at week 16. For each of these fluid compartments, the number of affected eyes was recorded at baseline and at week 16, based on OCT imaging, along with the respective dates of image acquisition.

Patient demographics and baseline ocular characteristics were summarized using descriptive statistics. The baseline variables included patient age at study entry, the number of treated female and male participants, and the number of patients presenting with cornea opacifications or lens opacifications (cataract). In addition, the number of patients with a history of intraocular surgery—excluding cataract extraction—was reported.

Imaging Acquisition and Analysis

All optical coherence tomography (OCT) imaging was performed using the REVO Fundus Camera (FC) spectral-domain OCT system (Optopol Technology Sp. z o.o., Zawiercie, Poland). CRT measurements obtained with the REVO FC OCT included the RPE layer, which should be taken into account when comparing CRT values across different studies and OCT platforms.

Imaging and data analysis were conducted using the integrated Spectral Optical Coherence Tomography (SOCT) software suite. CRT and retinal fluid compartments were evaluated using standard three-dimensional (3D) raster scans centered on the fovea. For each patient, the treated eye was evaluated at baseline (prior to first injection) and at week 16 (±14 days).

Detailed Early Treatment Diabetic Retinopathy Study (ETDRS) subfield analyses and lesion location classification (subfoveal, parafoveal, or extrafoveal) were not systematically recorded in routine practice and were therefore not available for this retrospective analysis. Macular neovascularization (MNV) subtype classification (type 1, type 2, or type 3) was not systematically documented and was not analyzed. OCT grading was based on routine site assessment of the scans; no central reading center or formal inter-rater reliability assessment was part of this retrospective study.

Statistical Analysis

All planned analyses were descriptive in nature and aimed to provide an overview of the functional and anatomical effects of faricimab in a treatment-naïve patient population with nAMD. No inferential statistical testing was performed because the study was prespecified as an exploratory descriptive real-world analysis; confidence intervals were considered sufficient to describe the precision and direction of the observed effects.

The focus of this research was on summarizing and presenting the observed data; therefore, no formal sample size calculation was conducted owing to the exploratory character of the study. However, reference data from the VOYAGER study indicated a mean change in central subfield thickness (CST) of −85.1 µm (standard deviation [SD] 105.6) at month 6 in patients with treatment-naïve nAMD [28]. On the basis of this estimate, a sample size of 45 eyes was considered sufficient to achieve a precision (defined as the half-width of the 95% confidence interval [CI]) of 30.9 µm. Thus, a total sample size of ~45 patients was deemed adequate to address the exploratory objectives of this study.

As this was a retrospective, observational study, missing values were expected and were not imputed unless otherwise specified. Given that only one eye per patient was treated, patient-level and eye-level data were identical throughout the analysis.

Demographic variables, including patient age and sex, as well as the number of treated eyes, were summarized using descriptive statistics. Baseline ocular characteristics, such as the presence of vision-impairing corneal or lens opacities and any prior intraocular surgery (excluding cataract surgery), were also summarized descriptively.

Continuous variables were summarized using mean, median, SD, first quartile, third quartile, minimum, and maximum. Categorical variables were summarized by counts and percentages. Where applicable, 95% confidence intervals were calculated; for proportions, exact Clopper–Pearson 95% CIs were used.

For the primary endpoint, CRT values (in µm) were summarized at baseline and at week 16 using descriptive statistics [mean, SD, median, range, and interquartile range (IQR)]. The change from baseline to week 16 was additionally reported with a 95% CI.

Analysis of the secondary endpoints followed the same approach. BCVA was assessed at baseline and week 16 and expressed in logMAR and Snellen decimal equivalents. Changes from baseline were reported only for patients with available data at both time points.

The number and proportion of eyes showing a transition from presence to absence of SRF/IRF/sub-RPE fluid from baseline to week 16 were also summarized descriptively. For the combined endpoint, IRF + SRF “presence” was defined as the detection of IRF and SRF, whereas “absence” was defined as absence of either IRF or SRF.

As this was a retrospective analysis of real-world data (RWD), patient visits did not follow a fixed schedule. To enable evaluation of outcomes at week 16, analysis time points were defined using windowing rules. The observation closest to day 113 (i.e., 16 weeks after the index date) and within a ±14-day window (day 99 to day 127) was used for analysis. If two observations were equidistant from day 113, the later of the two was used.

Windowing was applied before assessing the extent of missing data. If no data were available within the defined time window for a given eye, the time point was considered missing and not imputed.

No separate statistical analysis plan (SAP) was created, as the study was exploratory in nature with a limited set of predefined endpoints and analyses.

Statistical analyses were performed using R version 4.4.1.

Results

Participants

A total of 45 treatment-naïve eyes from 45 patients with nAMD were included in the final analysis. All patients completed the upload phase, receiving four consecutive monthly intravitreal injections of faricimab, and had valid OCT and BCVA assessments at both baseline and week 16 (±14 days).

The mean patient age at baseline was 82.1 years, with a range from 70–93 years. Slightly more than half of the treated patients were female (51.1%). A total of 40.0% of patients exhibited lens opacification, whereas corneal opacification was documented in 13.3%. In addition, 4.4% of patients had undergone previous intraocular surgery other than cataract extraction. Baseline demographic and ocular characteristics are presented in Table 1.

Table 1.

Demographic and ocular characteristics

Variable	Total (N = 45)
Age, years
Mean (SD)	82.1 (5.5)
Median	83.0
Range	70–93
Gender, n (%)
Male	22 (48.9%)
Female	23 (51.1%)
Lens opacification, n (%)
No	27 (60.0%)
Yes	18 (40.0%)
Cornea opacification, n (%)
No	39 (86.7%)
Yes	6 (13.3%)
Intraocular surgery, n (%)
No	43 (95.6%)
Yes	2 (4.4%)

N number of subjects in the analysis set, SD standard deviation

Primary Endpoint: Change in CRT

The primary endpoint was the change from baseline in CRT at week 16. Faricimab treatment resulted in a substantial and clinically meaningful anatomical response. The mean CRT decreased by −140.8 μm from 343.1 µm at baseline to 202.4 µm at week 16 (95% CI −178.2 to −103.3), reflecting effective retinal fluid resolution during the initial treatment phase. The CRT reductions are visualized in Fig. 1. See Supplementary Table S1 in the Electronic Supplementary Materials for detailed descriptive statistics of CRT at baseline, at week 16, and the change from baseline.

Fig. 1.

Central retinal thickness (CRT) (n = 45). SD standard deviation

Secondary Endpoint: Change in BCVA

Consistent with CRT improvements, patients also experienced visual benefits. At week 16, the mean change in BCVA was −0.15 logMAR (IQR: 0.20) from 0.58 at baseline to 0.43 at week 16. Conversion of the observed logMAR values to ETDRS letters indicates a clinically meaningful gain of approximately ten ETDRS letters corresponding to two lines on the ETDRS chart [29, 30]. Detailed descriptive statistics of BCVA in logMAR at baseline, at week 16, and the change from baseline, are presented in Table 2. See Supplementary Table S2 in the Electronic Supplementary Materials for detailed descriptive statistics of BCVA in Snellen decimal equivalents.

Table 2.

Descriptive statistics of best-corrected visual acuity (BCVA) in logMAR

Parameter	Visit	Value	Statistic	Total (N = 45)
BCVA (logMAR)	Baseline	Absolute value	Minimum	0.0969
			Q1	0.3010
			Median	0.39794
			Mean	0.58217
			95% CI	0.43245; 0.73189
			SD	0.49836
			Q3	0.6990
			IQR	0.3979
			Maximum	2.3010
			Range	2.2041
	Week 16	Absolute value	Minimum	0.0000
			Q1	0.2007
			Median	0.30103
			Mean	0.43091
			95% CI	0.29886; 0.56296
			SD	0.43952
			Q3	0.6021
			IQR	0.4014
			Maximum	2.3010
			Range	2.3010
		Change from baseline	Minimum	−1.0000
			Q1	−0.2041
			Median	−0.10037
			Mean	−0.15126
			95% CI	−0.21964; −0.08289
			SD	0.22758
			Q3	0.0000
			IQR	0.2041
			Maximum	0.3010
			Range	1.3010

BCVA values of hand motion below the chart (HBW) were imputed with 0.005

BCVA best-corrected visual acuity, CI confidence interval, N number of subjects in the analysis set, Q1 25th percentile, Q3 75th percentile, SD standard deviation, IQR interquartile range, BCVA(logMAR) −log(BCVA), BCVA (logMAR) change BCVA (logMAR) at 16 weeks − BCVA (logMAR) at baseline, Range max–min

Additional Secondary Endpoints: Retinal Fluid Dynamics

Detailed OCT-based evaluation of retinal fluid compartments revealed high rates of fluid resolution:

IRF: At baseline, IRF was present in 24 of 45 eyes (53.3%); at week 16, IRF had resolved in all of these 24 eyes, i.e., the proportion of eyes achieving absence of IRF was 100% (24/24). Overall, 21 of 45 eyes were IRF-free at baseline (21/45; 46.7%) and all remained IRF-free at week 16, i.e., the proportion of IRF-free eyes at baseline maintaining absence of IRF at week 16 was 100% (21/21). At the level of the overall cohort, all 45 eyes demonstrated absence of IRF at week 16, corresponding to 100% (45/45). The IRF status at baseline and at week 16 is shown in Fig. 2.

Fig. 2.

Intraretinal fluid (IRF) status at baseline and at week 16 (n = 45)

SRF: At baseline, SRF was present in 34 of 45 eyes (75.6%); at week 16, SRF had resolved in all but 2 of these eyes, i.e., the proportion of eyes achieving absence of SRF was 94.1% (32/34). In addition, 11 of 45 eyes were SRF-free at baseline (11/45; 24.4%) and all of these 11 eyes remained SRF-free at week 16, i.e., the proportion of SRF-free eyes at baseline maintaining absence of SRF at week 16 was 100% (11/11). At the level of the overall cohort, 43 of 45 eyes demonstrated absence of SRF at week 16, corresponding to 95.6% (43/45). The SRF status at baseline and at week 16 is shown in Fig. 3.

Fig. 3.

Subretinal fluid (SRF) status at baseline and at week 16 (n = 45)

IRF and SRF were present in 17 of 45 eyes (37.8%) at baseline and resolved in all of these 17 eyes by week 16, i.e., the proportion of eyes achieving absence of IRF and SRF was 100% (17/17). Overall, 28 of 45 eyes were IRF- and SRF-free at baseline (28/45; 62.2%) and all maintained absence of IRF and SRF at week 16, i.e., the proportion of IRF- and SRF-free eyes at baseline maintaining absence of IRF and SRF at week 16 was 100% (28/28). At the level of the overall cohort, all 45 eyes demonstrated absence of IRF and SRF at week 16, corresponding to 100% (45/45). The status of IRF and SRF at baseline and at week 16 is shown in Fig. 4.

Fig. 4.

Combined intraretinal fluid (IRF) + subretinal fluid (SRF) status at baseline and at week 16 (n = 45). For the combined endpoint, IRF + SRF “presence” was defined as the detection of IRF and SRF, whereas “absence” was defined as absence of either IRF or SRF

Sub-RPE fluid was present in 10 of 45 eyes (22.2%) at baseline and resolved in all but 1 of these eyes by week 16, i.e., the proportion of eyes achieving absence of sub-RPE fluid was 90% (9/10). Overall, 35 of 45 eyes (77.8%) without sub-RPE fluid at baseline continued to be sub-RPE fluid-free at week 16, i.e., the proportion of eyes maintaining absence of sub-RPE fluid was 100% (35/35). At the level of the overall cohort, 44 of 45 eyes demonstrated absence of sub-RPE fluid at week 16, corresponding to 97.8% (44/45). The sub-RPE fluid status at baseline and at week 16 is shown in Fig. 5.

Fig. 5.

Subretinal pigment epithelium (sub-RPE) fluid status at baseline and at week 16 (n = 45)

The structural response to faricimab in a representative patient with treatment-naïve nAMD with presence of retinal fluid at baseline is shown in Fig. 6a,b.

Fig. 6.

A, B Structural response to faricimab in a representative patient with treatment-naïve nAMD with presence of retinal fluid at baseline. Spectral-domain optical coherence tomography (SD-OCT) horizontal B-scans of the right eye obtained at baseline (A) and at week 16 (B). At baseline, OCT demonstrates intraretinal fluid (IRF), subretinal fluid (SRF), and subretinal pigment epithelium (sub-RPE) fluid involving the foveal center, consistent with active disease. At week 16, following intravitreal faricimab administered according to the study protocol, OCT shows complete resolution of IRF, SRF, and sub-RPE fluid with restoration of the foveal contour. Best-corrected visual acuity (BCVA) improved from 0.5 at baseline to 0.63 at week 16. This case is representative of the study population presenting with retinal fluid at baseline. BCVA best-corrected visual acuity, IRF intraretinal fluid, OCT optical coherence tomography, SD-OCT spectral-domain optical coherence tomography, SRF subretinal fluid, sub-RPE subretinal pigment epithelium

The descriptive statistics of retinal fluid status at baseline and week 16 are presented in Table 3, underscoring the rapid anatomical efficacy of faricimab across all retinal compartments. Supplementary Table S3 in the Electronic Supplementary Material summarizes mean BCVA, mean CRT, and the number and percentage of eyes with presence of retinal fluid at baseline and at week 16 (N = 45), whereas Supplementary Table S4 presents the corresponding data for eyes with absence of retinal fluid at the same time points.

Table 3.

Descriptive statistics of optical coherence tomography (OCT) results (N = 45)

	IRF	SRF	IRF + SRF	Sub-RPE fluid
Baseline
Yes (%)	24 (53.3%) (95% CI 37.9; 68.3)	34 (75.6%) (95% CI 60.5; 87.1)	17 (37.8%) (95% CI 23.8; 53.5)	10 (22.2%) (95% CI 11.2; 37.1)
No (%)	21 (46.7%) (95% CI 31.7; 62.1)	11 (24.4%) (95% CI 12.9; 39.5)	28 (62.2%) (95% CI 46.5; 76.2)	35 (77.8%) (95% CI 62.9; 88.8)
Week 16
Yes (%)	0 (0.0%) (95% CI 0.0; 7.9)	2 (4.4%) (95% CI 0.5; 15.1)	0 (0.0%) (95% CI 0.0; 7.9)	1 (2.2%) (95% CI 0.1; 11.8)
No (%)	45 (100.0%) (95% CI 92.1; 100.0)	43 (95.6%) (95% CI 84.9; 99.5)	45 (100.0%) (95% CI 92.1; 100.0)	44 (97.8%) (95% CI 88.2; 99.9)

CI confidence interval, IRF intraretinal fluid, N number of subjects in the analysis set, SRF subretinal fluid, sub-RPE subretinal pigment epithelium

Discussion

In this real-world single-center cohort of patients with treatment-naïve nAMD, completion of a four-dose faricimab upload per label at the time of study conduct was associated with an excellent anatomical and functional response at week 16, including a marked reduction in CRT by −140.8 µm from baseline (primary endpoint), clinically meaningful improvement in BCVA, near-complete resolution of retinal fluid compartments in eyes affected at baseline, and maintenance of fluid-free status in eyes with absence of retinal fluid at baseline.

Although many real-world reports use central subfield thickness (CST) rather than central retinal thickness (CRT), both parameters are derived from the same OCT data and segmentation algorithms. Consequently, CRT and CST values tend to show parallel trends, allowing for broadly comparable interpretation of anatomic outcomes in clinical practice [5, 31]. Differences in OCT devices, segmentation algorithms, and the use of CRT rather than CST may nevertheless contribute to variability in the magnitude of thickness change reported across studies. Accordingly, our week-16 outcomes align with findings from early real-world cohorts in patients with treatment-naïve nAMD that describe CRT or CST reductions, retinal fluid reductions, and BCVA gains over the initial faricimab dosing phase under routine conditions:

In the multicenter TRUCKEE study, treatment-naïve eyes (n = 39) showed a mean CST reduction of −84.5 µm after one faricimab injection and −80.1 µm after three injections (n = 13; p = 0.204) [32]. Beyond categorical grading, the TRUCKEE deep-learning subanalysis quantified retinal fluid volumes and showed a measurable reduction after a single faricimab injection (mean −60.7 nL from a baseline of 155.6 nL) with sustained fluid control over subsequent injections and progressively longer treatment intervals—reflecting a rapid and durable anatomical response under real-world conditions [33]. Furthermore, we observed substantial BCVA gains at week 16, despite the fact that approximately 40% of our cohort had lens opacities and 13% had corneal opacities. This is notable since such media opacities would tend to blunt the measurable visual gain. Under these conditions, the observed visual acuity gains are likely conservative, as media opacities would be expected to attenuate measurable functional improvement. In the TRUCKEE study, treatment-naïve eyes demonstrated a mean BCVA gain of +4.9 ETDRS letters after one injection (n = 39; p = 0.076) and +8.1 letters after three injections (n = 13; p = 0.437) [32]. The fact that our cohort achieved a gain of approximately ten ETDRS letters despite less favorable optical media underscores the robustness of the functional effect in our real-world setting. These findings support the paradigm—consistent with prior evidence that elimination of intraretinal fluid is associated with greater visual gains [5, 8]—that early anatomical drying can drive functional improvement, even in eyes with suboptimal optical conditions. Our findings also align with the pooled post hoc analysis of the TENAYA/LUCERNE head-to-head dosing phase (weeks 0–12), in which about 77% of faricimab-treated patients had absence of both IRF and SRF at week 12, compared with 67% on aflibercept 2 mg [7]. The dual resolution of IRF and SRF is clinically significant, given that persistent IRF, in particular, is associated with worse visual outcomes [5, 8, 34, 35].

Extending the comparison of our results to findings from other real-world cohorts further supports the consistency of early treatment outcomes with faricimab in treatment-naïve nAMD: In a prospective single-center study, Grimaldi et al. observed that 93.3% of treatment-naïve eyes achieved a dry macula within a median of 8 weeks after a four-dose loading phase. Significant CST reductions and sustained anatomical and functional gains were documented, and follow-up at weeks 12 (end of loading) and 16 (1 month post-loading) found 63.7% of eyes remained dry, highlighting early dryness and its potential relation to prolonged therapeutic effect [36]. In a Korean single-center series of 69 treatment-naïve nAMD eyes, Han et al. reported highly significant visual gains (mean logMAR improvement from 0.64 ± 0.41 to 0.47 ± 0.39 at 3 months) and high rates of fluid resolution after three loading injections of faricimab (SRF resolution in ~84% of eyes and IRF resolution in ~91%). The mean reduction in central retinal thickness was ~158 µm (424.1 → 266.3 µm) [37]. In comparison, the −140.8 µm CRT reduction observed in our cohort appears particularly pronounced, given the unselected, routine-practice setting. Consistent with these data, our analysis also showed near-universal fluid resolution in eyes with fluid at baseline (IRF 100%; SRF 94.1%), underscoring the strong early drying effect achieved with faricimab. The Italian FARIT study, a multicenter real-world cohort that included a treatment-naive nAMD subgroup, likewise documented a fast anatomical response after the loading phase, with interval decisions driven by BCVA stabilization/improvement and fluid resolution [38].

Taken together with prior international evidence, our results help to expand the currently scarce real-world data for faricimab from Germany for patients with treatment-naïve nAMD, supporting a more comprehensive understanding of early faricimab effectiveness under routine conditions. A retrospective single-center cohort from the University Eye Hospital Tuebingen by Kunzmann et al., including both treatment-naïve and previously treated nAMD eyes receiving up to three faricimab injections, showed early improvements in CRT and anatomical parameters, together with signals toward visual acuity gain, supporting faricimab’s early drying effect in a real-world setting [20]. In a retrospective single-center university cohort from Munich, Hafner et al. reported significant early OCT biomarker improvements (including reductions of structural disease activity markers) after a four-dose faricimab upload in treatment-naïve nAMD, accompanied by BCVA gains, reinforcing the notion of rapid anatomic response under routine care [27]. These German real-world data (Kunzmann et al.; Hafner et al.)—demonstrating early anatomic improvements within the first three to four faricimab injections—align with our week-16 findings and underscore the clinical value of evaluating the drying effect at the end of the upload phase [20, 27]. The assessment at week 16 is more than just a time point after the end of the upload; it represents the window during which the intrinsic potency of the therapy is most visible, before confounding factors such as interval adjustments, dropout, or regimen changes intervene. A strong early response—evidenced by a −140.8 µm CRT reduction, near-complete fluid resolution, and above-average BCVA gain in spite of optical limitations—suggests a favorable prognosis. Moreover, maintaining patient adherence—often reinforced by visible early improvements—is critical for sustaining long-term visual outcomes in nAMD [9].

Limitations: This is a retrospective, descriptive, and single-center analysis, with a short observation window limited to two time points (baseline and week 16) and a qualitative assessment of fluid (presence/absence) rather than volumetric quantification—constraints typical of practice-based datasets and acknowledged in observational reporting guidance [39, 40]. This pragmatic approach reflects clinical practice but may overestimate complete fluid resolution compared with quantitative fluid volume analysis. The absence of interim time-course analyses during upload precludes inferences about intra-upload dynamics; therefore, our inferences are restricted to the week-16 outcome. Selection of the later observation when two visits were equidistant from the target time point was intended to reflect maximal exposure to the loading regimen under real-world conditions.

The analysis was restricted to CRT, and the absence of detailed ETDRS subfield assessments and lesion location data may limit the ability to fully capture and interpret perifoveal or extrafoveal anatomical changes. The observed variability in CRT change was higher than anticipated based on the assumptions used for sample size justification, resulting in wider-than-expected confidence intervals and reduced precision of the estimates. The lack of systematic MNV subtype classification (type 1, type 2, or type 3) may limit the interpretation of anatomical outcomes, particularly with respect to lesion-specific morphological characteristics. The treatment regimen applied in this study reflects the EU label at the time of study conduct and precedes the 2024 update introducing a three-dose upload regimen, which may limit direct applicability of the findings to current clinical practice.

Further, the generalizability of our findings is limited, as this was a single-center study, and results may not be fully transferable to other patient populations or clinical settings.

Implications and outlook: Considering the marked CRT reduction, BCVA improvement, and drying effect observed in this cohort, the present findings in patients with treatment-naïve nAMD receiving faricimab align well with both pivotal and real-world evidence [7, 19, 20, 27, 32, 36–38]. A notable strength of this study is the complete follow-up, with all included patients (45/45) having valid assessments at both time points, thereby minimizing the risk of attrition bias. Longer-term follow-up of the 45DRY cohort should assess the durability of anatomic control, functional outcomes, and real-world treatment interval trajectories. The international VOYAGER program is designed to generate multi-year, multi-country real-world evidence in patients with treatment-naïve nAMD, including data from sites in Germany [28].

Although adherence was not assessed in our study, nonadherence and nonpersistence to intravitreal therapy are common across macular diseases and are influenced by treatment burden; consequently, regimens that achieve early, durable drying and permit interval extension are plausibly supportive of persistence in practice, a hypothesis consistent with the broader adherence literature but requiring prospective evaluation of faricimab in patients with treatment-naïve nAMD [41, 42].

Conclusions

Within the constraints of a retrospective, practice-based design, completion of a four-dose faricimab upload in patients with treatment-naïve nAMD was associated at week 16 with (i) a reduction in CRT from baseline, (ii) an improvement in BCVA, and (iii) marked drying of retinal fluid compartments, with sustained retinal fluid-free status in eyes exhibiting absence of retinal fluids at baseline. These findings demonstrate the rapid effectiveness of faricimab in real-world clinical practice; however, they do not permit inference about intra-upload dynamics or longer-term durability. Future analyses of this cohort and forthcoming data from VOYAGER should clarify durability of anatomic control, visual outcomes, and real-world interval strategies with faricimab in patients with treatment-naïve nAMD [28].

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

All authors contributed to the study conception and design. Material preparation and data collection and analysis were performed by Maximilan Haase. The first draft of the manuscript was written by Maximilian Haase, and Maximilian Haase and Esther Wittenborn commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

Sponsorship for this study and the Rapid Service Fee were funded by Roche Pharma AG, 79639 Grenzach-Wyhlen, Germany.

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

Maximilian Haase has received speaker honoraria and research support from Roche Pharma AG. Esther Wittenborn is an employee of Roche Pharma AG.

Ethical Approval

This research was conducted in full compliance with the Guidelines for Good Pharmacoepidemiology Practice (GPP) published by the International Society of Pharmacoepidemiology (https://www.pharmacoepi.org/resources/policies/guidelines-08027), the laws and regulations in Germany as applicable, and the Declaration of Helsinki. The research plan and relevant supporting information was approved by the competent Ethics Committee of the University Rostock on 2 May 2025 (A 2025–0115), before the research was initiated. All patients provided consent for the use and anonymization of the data and explicitly consented to any secondary use of their data.