Early Outcome of Aflibercept 8 mg for Neovascular AMD in the Real-world Setting

Abstract

Purpose To investigate the early outcome of intravitreal aflibercept 8 mg in treating neovascular age-related macular degeneration (nAMD) of eyes that either received prior intravitreal anti-VEGF treatment or were treatment-naïve – under real-world conditions.

Methods This is a retrospective study of a total of 83 eyes with nAMD treated with aflibercept 8 mg. 61 of these eyes completed an initial loading phase of 3 monthly intravitreal injections (IVI) and were included in further analyses. Outcome parameters included best-corrected visual acuity (BCVA, logMAR), presence of intraretinal (IRF) and subretinal fluid (SRF), extent of pigment epithelium detachment (PED height, µm) and central subfield retinal thickness (CSRT, µm) in spectral domain optical coherence tomography (SD-OCT). Patients were assessed at the beginning of aflibercept 8 mg treatment and 4 weeks after completing the loading phase. The McNemar test was used to test for changes in distribution of eyes showing IRF and SRF, and the paired t test was performed to test for changes in BCVA, PED height and CSRT.

Results 51 eyes had been previously treated with intravitreal anti-VEGF, while 10 eyes were treatment-naïve. Mean BCVA after completed loading phase was 0.49 ± 0.31 logMAR and did not show changes to baseline BCVA, which was 0.49 ± 0.32 logMAR (p = 0.89). A significant reduction was observed in all disease activity parameters in SD-OCT. The proportion of eyes showing IRF and SRF decreased from 54.1% to 26.2% (p < 0.001) and 65.6% to 24.6% (p < 0.001), respectively. A reduction in PED height from 227.7 ± 114.6 µm at baseline to 191.9 ± 111.4 µm (p < 0.001) and a decrease in CSRT from 375.2 ± 126.2 µm to 308.8 ± 93.7 (p < 0.001) were recorded. Of all eyes (n = 83) that had received at least one IVI aflibercept 8 mg, 5 eyes (6.0%) developed symptomatic, non-infectious intraocular inflammation (IOI), which resolved completely with topical treatment.

Conclusion Our results demonstrate good effectiveness of intravitreal aflibercept 8 mg in the real-world setting, with significant reduction in disease activity parameters in SD-OCT after the loading phase with 3 monthly injections – while BCVA remained stable. While the PULSAR trial only included treatment-naïve eyes, our study underlines the value of aflibercept 8 mg, even for pre-treated eyes that responded insufficiently to previous anti-VEGF treatment. Further studies are needed to evaluate long-term outcome in real-world setting, in order to verify the extended IVI intervals shown in the PULSAR trial.

Zusammenfassung

Zielsetzung Untersuchung des kurzfristigen Effekts von Aflibercept 8 mg intravitreal zur Behandlung der neovaskulären altersbedingten Makuladegeneration (nAMD) bei vorbehandelten und therapienaiven Augen unter Real-World-Bedingungen.

Methoden Dies ist eine retrospektive Studie mit insgesamt 83 Augen, die aufgrund einer nAMD mit intravitrealen Injektionen (IVOM) Aflibercept 8 mg behandelt wurden. 63 Augen vervollständigten einen initialen Upload von 3 monatlichen IVOMs und wurden in weitere Analysen eingeschlossen. Die Outcome-Parameter beinhalteten den bestkorrigierten Visus (BCVA, logMAR), das Vorhandensein von intraretinaler (IRF) und subretinaler Flüssigkeit (SRF), die Höhe der Pigmentepithelabhebung (PED-Höhe, µm) und die zentrale Netzhautdicke (CSRT, µm) in der Spectral Domain-optischen Kohärenz-Tomografie (SD-OCT). Die Patienten wurden zu Beginn des Uploads sowie 4 Wochen nach Upload untersucht. Ein McNemar-Test wurde zur Untersuchung auf Veränderungen des Anteils and Augen verwendet, die IRF und SRF zeigten. Gepaarte t-Tests wurden zur Analyse von BCVA, PED-Höhe und CSRT verwendet.

Ergebnisse 51 Augen waren mit anderen Anti-VEGF-Präparaten vorbehandelt, während 10 Augen therapienaiv waren. Der durchschnittliche Visus nach erfolgtem Upload betrug 0,49 ± 0,31 logMAR und zeigte keine Veränderung zum Baseline-Visus von 0,49 ± 0,32 logMAR (p = 0,89). Es zeigte sich eine signifikante Reduktion aller in der SD-OCT untersuchten Aktivitätsparameter. Der Anteil an Augen mit IRF verringerte sich von 54,1% auf 26,2% (p < 0,001) und der Anteil an Augen mit SRF von 65,6% auf 24,6% (p < 0,001). Zudem war eine Reduktion der PED-Höhe von 227,7 ± 114,6 µm auf 191,9 ± 111,4 µm (p < 0.001) sowie der CSRT von 375,2 ± 126,2 µm auf 308,8 ± 93,7 (p < 0,001) festzustellen. Von allen Augen (n = 83), die bis zur Datenerfassung mindestens eine IVOM Aflibercept 8 mg erhalten hatten, entwickelten 5 Augen (6.0%) eine symptomatische, nicht-infektiöse intraokuläre Inflammation (IOI), die durch topische Therapie komplett regredient war.

Schlussfolgerungen Aflibercept 8 mg zeigt unter Real-World-Bedingungen nach initialem Upload eine gute Wirksamkeit mit stabilem Visus bei der Behandlung der nAMD, sowohl bei vorbehandelten als auch Therapie-naiven Augen. Weitere Studien sind notwendig, um das langfristige Outcome sowie insbesondere verlängerte IVOM-Intervalle unter realen Bedingungen zu untersuchen.

Keywords

retina - age-related macular degeneration - AMD - Aflibercept - real-world

Schlüsselwörter

Netzhaut - Aflibercept 8 mg - AMD - Anti-VEGF

Introduction

Age-related macular degeneration (AMD) is one of the main causes of visual impairment and blindness in adults in developed countries, being accountable for approximately 54% of cases of blindness among Caucasian ethnicity in the United States [1]. Late-stage AMD can develop as either geographic atrophy (GA) or neovascular AMD (nAMD). There has been successful treatment of nAMD for almost two decades, preserving central vision for many more years than before. The intravitreally applied medications mainly target vascular endothelial growth factor (VEGF), which has been found to be one of the main underlying causes of the formation of neovascular membranes in AMD. Despite the success of anti-VEGF intravitreal injections (IVI s), the high frequence of required IVI s – often on a monthly schedule – cause personal burden to the patient and puts him to risk for several complications, such as infectious endophthalmitis and retinal detachment. Moreover, socioeconomic costs are high and will even grow in the future, given the estimations of worldwide rising numbers of patients affected by nAMD [2]. On this account there has been a great effort to develop new therapeutics that would allow to achieve the same functional and anatomical effectiveness as other anti-VEGF drugs, but with fewer injections.

Aflibercept is not only targeting VEGF-A, but also other members of the VEGF-family, namely VEGF-B and Placenta Growth Factor (PGF) [3]. While VEGF-A is seen to be most associated with angiogenesis, other growth factors like PGF have been shown to contribute to the formation of CNV [4], [5]. By targeting multiple growth factors, aflibercept has a potential added benefit in the suppression of MNV formation. The drug has shown to be highly effective and safe in the use of treating nAMD [6]. With its new pharmaceutical formulation aflibercept 8 mg, the need of reduction of injection frequency has been addressed. Aflibercept 8 mg has been approved for the treatment of nAMD and diabetic macular edema (DME) in the US and EU. CANDELA was a phase 2 clinical trial comparing aflibercept 8 mg to aflibercept 2 mg using the same treatment regimen. While the safety profile was comparable in both arms, a tendency towards greater BCVA gains and better anatomical outcomes was seen in the aflibercept 8 mg arm. However, these findings did not reach statistical significance [7]. Phase 3 pivotal study PULSAR compared treatment-naïve nAMD patients randomized to either aflibercept 2 mg every 8 weeks (2q8) or aflibercept 8 mg with extended dosing intervals of initially every 12 (8q12) or 16 (8q16) weeks. Treatment regimens of patients in aflibercept 8 mg arms could be shortened after 16 weeks and/or extended after 52 weeks based on individual disease activity. The study demonstrated non-inferiority of visual gains of 8q12 and 8q16 arms, while after 2 years injection intervals had been extended to injections every 12 weeks or even longer in 88% of patients receiving aflibercept 8 mg. Morphologic response, namely reduction of central subfield retinal thickness (CSRT), was comparable in the aflibercept 2 mg and both 8 mg arms with very small fluctuations during the course of the 2 treatment years. Overall safety profile was comparable to the aflibercept 2 mg group [8]. The approval studies have demonstrated a good efficacy of aflibercept 8 mg.

Approval studies are usually designed as randomized controlled trials (RCT). RCTs undergo a complex randomization and blinding process in order to avoid potential bias and interfering factors. RCTs are considered to provide the highest evidence of treatment effects and aside from some exceptions, are usually required by regulatory authorities to approve a new drug. However, in some aspects the conditions established in an RCT differ substantially from the circumstances in daily practice. For instance, inclusion criteria of patients are very strict, mostly excluding patient with specific comorbidities or pre-treated patients. In a way, this leads to an artificial study population, which can compromise the transferability to the patient population seen in daily practice. After RCTs have led to the approval of a new drug, to close this gap it is therefore crucial to further investigate the efficacy and safety of new therapies in real-world settings.

To date, few studies have explored the efficacy of aflibercept 8 mg under real-world conditions. Hosoda et al. published a retrospective study with 23 treatment-naïve eyes showing significant improvement of BCVA and CSRT-reduction after 3 monthly IVI s. Dry macula was seen in all eyes after loading phase [9]. Matsumoto et al. published a retrospective study of 35 eyes consisting of 18 treatment-naïve and 17 pre-treated eyes. After one single IVI aflibercept 8 mg, significant improvement of BCVA and CSRT was seen and a dry macula was achieved in 57.1% [10].

In terms of safety, there are a few more studies reporting intraocular inflammation (IOI) following aflibercept 8 mg. Hosoda et al. reported one case of retinal pigment epithelial tear (RIP), but no cases of IOI [9]. Matsumoto et al. reported IOI associated with retinal vasculitis in 3 eyes (8.6%) after administering aflibercept 8 mg [10]. Sambhara et al. published a retrospective study of 36 pre-treated and 4 therapy-naïve eyes without any case of IOI within the follow-up period of 6 months [11]. Binder et al. published a case series of 41 eyes receiving aflibercept 8 mg, of which 12.0% of patients developed IOI. However, this study not only included nAMD patients, but also patients treated for diabetic macular edema (DME) [12]. Moreover, a case report about IOI associated with retinal vasculitis has been published by Hashiya et al. [13] and a multi-center cluster of 8 cases of IOI by Hoffmann et al. [14]. In summary, the reported frequency rates of IOI differed greatly between the mentioned real-world studies, ranging from 0.0% to 12.0%.

Therefore, only few studies with small sample size have investigated real-world efficacy of aflibercept 8 mg and safety profile varies greatly. To contribute to the few reported experiences under real-world conditions, we evaluated all patients who were treated with aflibercept 8 mg for nAMD in our clinic until early September 2024. We investigated short-team outcome of disease activity parameters after initial loading phase with 3 monthly injections as well as overall safety.

Materials and Methods

This is a retrospective, single-center study. The study was approved by the ethics committee of Westphalia-Lippe, Germany. Research and data management were executed in conformity with the views of the Declaration of Helsinki and its amendments.

Study protocol

All patients who received their first IVI aflibercept 8 mg for nAMD at our clinic between February of 2024 and early September of 2024 were retrospectively reviewed and included in this study. No exclusion criteria were applied at initial selection of study population. Patients were either switched to aflibercept 8 mg from other anti-VEGF medications including aflibercept 2 mg, faricimab, ranibizumab and bevacizumab due to insufficient effect or received aflibercept 8 mg as their first IVI drug after being newly diagnosed with nAMD. Insufficient effect was defined by persisting disease activity parameters 4 weeks after last IVI treatment, such as new bleeding, increase of sub-Retinal Pigment Epithelium (RPE)-fluid or persisting IRF and/or SRF. There was no specific selection process on why treatment-naïve patients received aflibercept 8 mg as first IVI agent. The usual treatment strategy was to start with a loading phase of 3 monthly injections and a first follow-up visit 4 weeks later. In individual cases, eyes did not receive a loading phase but only single injections right from the start. These eyes were not included in further calculations. Furthermore, patients who did not meet follow-up visit 4 weeks after loading phase were not included in further calculations. Reasons included personal decision to discontinue aflibercept 8 mg treatment by the patient, missing follow-up OCTs, discontinuation due to adverse events and discontinuation due to poor visual acuity below decimal visus of 0.05, which is mandatory for IVI treatment. At follow-up, based on the individual treatment response, a decision was made whether aflibercept 8 mg treatment could be continued. If at follow-up no treatment response was seen, aflibercept 8 mg was discontinued.

In the case of treatment continuation, the earliest possible next IVI aflibercept 8 mg could be administered 8 weeks after finishing loading phase, based on the approved time interval by the EMA. If at first follow-up after loading phase no treatment response was seen, aflibercept 8 mg treatment was classified as failed and patients had to be switched to a different anti-VEGF drug. In each case, the first interval until next given treatment was investigated up to 16 weeks after finished loading phase. Further IVI s after completing loading phase were handled in a “pro re nata” approach, meaning that single IVI s were given as soon as new disease activity was seen. If patients missed one or more visits during the observation period of 16 weeks after loading phase, these patients were excluded from re-treatment interval analyses.

Outcome measures

Outcome measures included functional and morphological parameters. Best-corrected visual acuity (BCVA) was measured using Decimal charts and were then converted to logMAR for statistical analyses. Morphological parameters included several disease activity measures in spectral domain optical coherence tomography (SD-OCT, Spectralis HRA+OCT, Heidelberg Engineering, Heidelberg, Germany). These included presence of intraretinal fluid (IRF) and subretinal fluid (SRF), which was assessed as either “yes” or “no”. If neither IRF nor SRF was detectable, macula was labeled “dry”. Achieving a dry macula is considered good treatment response and controlled disease activity. However, in some pre-treated, therapy-recalcitrant eyes, dry macula is sometimes not achievable and the primary goal is to contain disease activity as much as possible. Central subfield retinal thickness (CSRT, µm) was measured as mean thickness between Bruchʼs membrane (BM) and Internal limiting membrane (ILM) within the 1 mm foveal diameter. The extent of pigment epithelium detachment (PED height, µm) was measured as the greatest distance between BM and pigment epithelium (PE) apex within the SD-OCT scan. Each OCT was manually checked for accuracy of fovea centering and retinal layer segmentation by a single trained and certified Reading Center Grader. If automatic fovea detection and/or automatic segmentation of BM and/or ILM were incorrect, they were appropriately adjusted. Given the high proportion of pre-treated eyes with advanced disease stage, most of OCTs had to be manually adjusted. All data was collected at the beginning of aflibercept 8 mg treatment (baseline), 4 weeks after completing loading phase (follow-up) and up to 16 weeks after loading phase to investigate the re-treatment interval.

Statistical analysis

In terms of outcome parameters, only eyes which completed full loading phase of 3 monthly injections with aflibercept 8 mg and a follow-up visit after 4 weeks were included for further analysis. In terms of adverse effects, the whole cohort of 83 eyes was investigated to be able to draw reasonable conclusions about overall safety.

Statistical analyses were calculated with software R (Version 4.2.3, R Core Team 2023. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria). Statistical tests were two-tailed. McNemar test was used to test for significant changes in distribution of eyes showing IRF and SRF. Paired t-test was performed to test for significant changes in BCVA, PED height and CSRT. P values lower than 0.05 were considered significant. P-values were adjusted by the method of Benjamini-Hochberg against multiple testing.

Results

Population

Between February of 2024 and early September of 2024, a total of 83 eyes of 79 patients were started with aflibercept 8 mg for nAMD in our clinic. Until the end of December 2024, a total of 329 injections with aflibercept 8 mg had been administered in this cohort. 7 patients (8.4%) decided to discontinue loading phase with aflibercept 8 mg because of personal reasons, 7 eyes (8.4%) were missing follow-up OCT, 4 eyes (4.8%) only received single injections instead of a full loading phase, 3 eyes (3.6%) were discontinued during loading phase due to adverse events and 1 eye (1.2%) was discontinued during loading phase due to poor visual acuity. Therefore, 22 eyes (26.5%) did not finish full loading phase of 3 monthly injections and a follow-up visit and could therefore not be included in further calculations.

However, 61 eyes (73,5%) of 59 patients were included in the study with 24 eyes (39.3%) of male patients and 37 eyes (60.6%) of female patients. Mean age was 82.0 ± 7.3 years (58 – 95 years). 51 eyes (83.6%) had had prior intravitreal anti-VEGF treatment while 10 eyes (16.4%) were treatment-naïve. Of the eyes that had prior treatment, 34 eyes (66.7%) were switched from aflibercept 2 mg, 13 eyes (25.5%) from faricimab, 2 eyes (3.9%) from ranibizumab and 2 eyes (3.9%) from bevacizumab. Pre-treated eyes on average received 39 (3 – 141) injections with 2.4 (1 – 5) different substances before starting aflibercept 8 mg. Switching to aflibercept 8 mg was mainly due to poor response to previous anti-VEGF drugs, in some cases the primary goal was to extend treatment intervals.

Outcome measures

An overview of all tested biomarkers is shown in [Table 1]. At baseline, mean BCVA was 0.49 ± 0.32 logMAR in all eyes. At follow-up, functional outcome did not change. 4 weeks after loading phase mean BCVA was 0.49 ± 0.31 logMAR (p = 0.89, [Fig. 1]). Dividing the cohort into pre-treated and treatment-naïve eyes did not change this observation. BCVA in pre-treated eyes remained stable (p = 0.75) as well as in treatment-naïve eyes (p = 0.75).

Table 1 Biomarker development from baseline to follow-up. Development of all biomarkers from baseline to follow-up; all eyes (“all”) and sub-group analyses of pre-treated and treatment-naïve eyes.

	Baseline	Follow-up	p value
BCVA (logMAR)
All (n = 61)	0.49 ± 0.32	0.49 ± 0.31	0.89
Pre-treated (n = 51)	0.48 ± 0.32	0.47 ± 0.30	0.75
Treatment-naïve (n = 10)	0.55 ± 0.38	0.58 ± 0.37	0.75
Intraretinal fluid (%)
All (n = 61)	54.1	26.2	< 0.001
Pre-treated (n = 51)	52.9	27.5	< 0.001
Treatment-naïve (n = 10)	60.0	20.0	0.16
Subretinal fluid (%)
All (n = 61)	65.6	24.6	< 0.001
Pre-treated (n = 51)	62.7	27.5	< 0.001
Treatment-naïve (n = 10)	80.0	10.0	< 0.05
IRF or SRF (%)
All (n = 61)	98.4	45.9	< 0.001
Pre-treated (n = 51)	100.0	50.1	< 0.001
Treatment-naïve (n = 10)	90.0	20.0	< 0.05
PED height (µm)
All (n = 52)	227.7 ± 114.6	191.9 ± 111.4	< 0.001
Pre-treated (n = 46)	223.9 ± 114.2	193.5 ± 115.3	< 0.001
Treatment-naïve (n = 6)	257.3 ± 124.4	180.3 ± 82.0	< 0.05
CSRT (µm)
All (n = 60)	375.2 ± 126.2	308.8 ± 93.7	< 0.001
Pre-treated (n = 50)	371.6 ± 120.7	315.5 ± 94.0	< 0.001
Treatment-naïve (n = 10)	392.8 ± 156.9	271.3 ± 86.9	< 0.005

Fig. 1 Development of BCVA of all eyes (n = 61).

Before starting aflibercept 8 mg, 33 out of 61 eyes (54.1%) showed IRF. After 3 monthly injections, there was only 16 eyes (26.2%) left with IRF (p < 0.001). Looking at only pre-treated eyes, the proportion of eyes showing IRF was reduced from 52.9% to 27.5% (p < 0.001). There was a trend towards a decrease of the proportion of treatment-naïve eyes showing IRF from 60.0% to 20.0%. However, most likely due to small sample size, this effect was not significant (p = 0.16). A similar effect was seen regarding subretinal fluid. While at baseline 40 eyes (65.6%) showed SRF, at follow-up only 15 eyes (24.6%) still had SRF (p < 0.001). The proportion of eyes showing SRF was significantly reduced in pre-treated eyes (p < 0.001) as well as treatment-naïve eyes (p < 0.05), from 62.7% to 27.5% and from 80.0% to 10.0%, respectively. When looking at both IRF and SRF combined, at baseline 60 out of 61 eyes (98.4%) showed either IRF or SRF. There was only one treatment-naïve eye that showed neither IRF nor SRF, but ophthalmoscopy revealed macular bleeding and OCT showed new unsharp SHRM as signs of neovascular activity, which resolved after loading phase. At follow-up, only 28 eyes (45.9%) showed IRF or SRF, meaning that 54.1% were declared “dry” after loading phase with aflibercept 8 mg (p < 0.001). The proportion of eyes showing IRF or SRF was significantly reduced in pre-treated eyes as well as in treatment-naïve eyes, from 100.0% to 50.1% (p < 0.001) and from 90.0% to 20.0% (p < 0.05), respectively.

Out of all 61 eyes, 52 eyes (85.2%) showed a PED. In these eyes, mean PED height before aflibercept 8 mg loading phase was 227.7 ± 114.6 µm and decreased to 191.9 ± 111.4 µm at follow-up (p < 0.001, [Fig. 2]). PED height in pre-treated eyes was reduced from 223.9 ± 114.2 µm to 193.5 ± 115.3 µm (p < 0.001) and PED height in treatment-naïve eyes was reduced from 257.3 ± 124.4 µm to 180.3 ± 82.0 µm (p < 0.05). At last, a significant reduction of CSRT was seen. Mean CSRT of all eyes decreased from 375.2 ± 126.2 µm at baseline to 308.8 ± 93.7 µm at follow up (p < 0.001, [Fig. 2]). CSRT in pre-treated eyes was reduced from 371.6 ± 120.7 µm to 315.5 ± 94.0 µm (p < 0.001) and CSRT in treatment-naïve eyes was reduced from 392.8 ± 156.9 µm to 271.3 ± 86.9 µm (p < 0.005).

Fig. 2 Development of CSRT (n = 60) and PED height (n = 52).

After finishing loading phase, out of all 61 eyes, 6 eyes missed at least one follow-up visit within the following 16 weeks. However, 55 eyes were investigated on how much time passed until another IVI treatment was necessary. At first follow-up visit 4 weeks after loading phase, in 13 eyes (23.6%) there was no sufficient disease control to wait until next follow-up visit 4 weeks later. Because in-label treatment with aflibercept 8 mg does not allow treatment intervals below 8 weeks after finishing loading phase, most of these eyes (10 eyes, 18.2%) were switched again to different anti-VEGF drugs. However, in 3 cases, an additional IVI aflibercept 8 mg was given at first follow-up visit, before reaching 8-week interval. The decision for this off-label use was made after careful consideration together with the patient and based on the therapy-recalcitrant course with no treatment response to prior anti-VEGF therapeutics. In total 42 eyes (76.3%) showed increasing MNV activity and need for another IVI only after at least 8 weeks, with 17 eyes (30.9%) even reaching a time span without MNV activity of at least 12 weeks and 6 eyes (10.9%) of even 16 weeks or longer.

Severe adverse events

Out of all the 83 eyes that had received at least one IVI with aflibercept 8 mg, 5 eyes (6.0%) showed intraocular inflammation (IOI) few days after injection. Until end of December 2024, a total of 329 IVI s aflibercept 8 mg had been administered in this cohort. This means IOI occurred in 1.5% of IVI s. 4 eyes developing IOI were previously treated with other anti-VEGF medications and were switched to aflibercept 8 mg, while one eye was treatment-naïve. 2 eyes had had lower-dose aflibercept 2 mg before at least once without occurrence of intraocular inflammation. On average 2.6 (2 – 4) days had passed after last IVI with aflibercept 8 mg until patients experienced decline of visual acuity (4 eyes) and/or mild pain (2 eyes) and/or new floaters (1 eye). 2 cases occurred after the first IVI of aflibercept 8 mg, while the other cases presented after 3, 5 and 6 IVI s. On slit-lamp examination, patients showed mild to moderate vitreous body inflammation, sometimes accompanied by anterior chamber cells and keratic precipitates. None of the cases showed hypopyon, retinal hemorrhages, retinal infiltrates or signs of retinal vasculitis. Depending on the degree of inflammation, patients were treated with varying frequency of topical corticosteroidal drops, sometimes additionally mydriatic drops or parabulbar corticosteroidal injections. Eventually, all of the cases resolved with topical treatment and did not need surgical intervention as bacterial endophthalmitis was ruled out.

Moreover, one patient (1.2%) showed a massive decrease of platelet count in the blood from 227.000/µl at the beginning of aflibercept 8 mg treatment to only 28.000/µl after 3 injections (normal platelet count ranges from 150.000 – 300.000 platelets/µl). Even though a causal relationship to the medication was questionable, aflibercept 8 mg was discontinued in this patient. Bone marrow puncture was initiated to rule out underlying hematologic disease and revealed toxic damage as the likeliest cause of bone marrow injury. Platelet count spontaneously increased to normal range again within the next months and was at 207.000/µl 9 months later.

Discussion

In order to be able to face rising patient numbers in the future, but also to lower IVI-associated risks like bacterial endophthalmitis, reduce personal burden and socioeconomic costs, new anti-VEGF treatment options are being developed. The goal is to achieve same or even better functional and anatomical outcomes in contrast to previous anti-VEGF drugs, but with extended IVI intervals. Aflibercept 8 mg has recently been approved for the treatment of nAMD in the US and EU. Pivotal trial PULSAR was able to show that after 2 years of treatment, 88% of the treatment-naïve patients needed IVI intervals of every 12 weeks or even longer with non-inferiority of visual gains and comparable morphologic response, namely CSRT-reduction. Also, safety profile was comparable to that of aflibercept 2 mg with 1.3% of eyes developing IOI in aflibercept 8 mg groups and 2.1% in aflibercept 2 mg arm [8].

To this date, only limited data is available about effectiveness and safety profile of aflibercept 8 mg in real-world setting. Hosoda et al. reported significant improvement of BCVA and CSRT-reduction after 3 monthly IVI s in 23 treatment-naïve eyes, dry macula was achieved in all eyes after loading phase [9]. Matsumoto et al. reported significant improvement of BCVA and CSRT in 35 eyes consisting of 18 treatment-naïve and 17 pre-treated eyes after one single IVI. Dry macula was achieved in 57.1% [10].

While Hosoda et al. reported no cases of IOI and just one case of retinal pigment epithelial tear (RIP) in their study [9], Matsumoto et al. observed 3 eyes (8.6%) developing non-infectious IOI associated with retinal vasculitis. All cases resolved with topical corticosteroid treatment [10]. Sambhara et al. didnʼt observe IOI within the follow-up period of 6 months in 36 pre-treated and 4 therapy-naïve eyes either [11]. In contrast, a case series by Binder et al. demonstrated 5 of 41 eyes (12.0%) developing IOI after aflibercept 8 mg. It is important to note that in this study, not only nAMD patients, but also patients treated for diabetic macular edema (DME) were included [12]. Further, Hashiya et al. published a case report about a therapy-recalcitrant patient who responded well to aflibercept 8 mg anatomically, but developed IOI associated with retinal vasculitis after a few weeks. Even though the patient was successfully treated with corticosteroids, visual impairment remained [13]. Another recently published case report of retinal vasculitis following 11 days after aflibercept 8 mg describes a patient who was treated with topical, intravitreal and systemic corticosteroids [15]. Hoffmann et al. published a case series of 8 eyes developing IOI after intravitreal aflibercept 8 mg. None of these eyes showed signs of retinal vasculitis and resolved completely with topical corticosteroid treatment and non-steroidal anti-inflammatory drugs [14]. In summary, the few studies available have shown short-term anatomical and functional outcomes following aflibercept 8 mg to be satisfying. However, the study-designs are heterogenous and sample sizes are small. Moreover, the reported frequency rates of IOI differ greatly between the mentioned real-world studies, ranging from 0.0% to 12.0%. IOIs reported have described to be associated with and without retinal vasculitis.

The purpose of our study was therefore to investigate the short-term effectiveness and safety profile of aflibercept 8 mg in our clinic. Medical records of a total of 83 eyes of 79 patients were reviewed. 61 eyes of 59 patients finished 3 monthly injections with aflibercept 8 mg and a follow-up visit 4 weeks later and were therefore available for analysis of treatment outcome. A significant reduction of all disease activity parameters including IRF, SRF, PED height and CSRT was seen. Dry macula was achieved in 54.1% of eyes. This finding is in line with results by Matsumoto et al., who observed dry macula in 57.1% of cases [10]. Comparability is likely due to the fact that both studies included pre-treated as well as treatment-naïve eyes. This is in contrast to the real-world study published by Hosoda et al., which only included treatment-naïve eyes and dry macula was achieved in 100.0% of eyes after loading phase [9]. The difference between pre-treated and treatment-naïve eyes is supported by sub-group analyses in our study. The decrease of PED height and CSRT reduction seemed to be greater in treatment-naïve eyes than in pre-treated eyes, and treatment-naïve eyes were considered “dry” in 80.0% of cases. However, this observation has be interpreted cautiously as the treatment-naïve subgroup (n = 10) was significantly smaller that pre-treated eyes (n = 51) in this study.

While the CANDELA and PULSAR trials, as well as real-world studies by Hosoda et al. and Matsumoto et al. observed significant functional gains [7], [8], [9], [10], BCVA showed no improvement in our cohort. This finding is most likely due to the greater proportion of pre-treated, often therapy-recalcitrant eyes and a small sample size of treatment-naïve eyes in our study. Moreover, this finding was not surprising. A stable BCVA is often a satisfactory result during anti-VEGF treatment and indicates a good control of disease activity. In the context of longer observation periods, even under anti-VEGF treatment BVCA actually gradually decreases over time [16]. Hence, we see the stable BCVA in our cohort in concordance with good effectiveness of the drug. After finishing loading phase, a large portion of the eyes (76.3%) only needed another IVI treatment after 8 weeks or longer and were therefore eligible for further use of the treatment given the minimum interval of 2 months required in the EMA approval of the drug. 17 eyes (30.9%) even reached an interval without MNV activity of at least 12 weeks. In 13 eyes (23.6%) treatment with aflibercept 8 mg was not sufficient and most of the eyes were therefore switched to other drugs. In the PULSAR trial, after 2 years of treatment 88% of patients reached at least 12-week dosing intervals [8]. Considering the PULSAR study only included treatment-naïve patients, the discrepancy to our study, which mainly included pre-treated patients, can be explained.

However, it has to be pointed out that the CSRT reduction seen in our study, as well as demonstrated in PULSAR trial, did not show superiority to eyes treated with aflibercept 2 mg [8]. A tendency towards superior anatomical and functional outcome of aflibercept 8 mg over aflibercept 2 mg was first described in the CANDELA study, although this difference did not reach statistical significance [7].

In terms of adverse events, out of the full cohort of 83 eyes that had received at least one IVI aflibercept 8 mg, 5 eyes (6.0%) developed symptomatic, non-infectious IOI without any signs of retinal vasculitis. In all of the 5 eyes with IOI in this study, we stopped treatment with aflibercept 8 mg immediately and switched to a different therapeutic to avoid another inflammatory response. Although we were able to successfully treat every case without any long-lasting complications, a precise strategy how to proceed with the further anti-VEGF treatment of these cases or even avoid these adverse events in the first place is lacking. While patients will very likely need further IVI s, it remains unclear if they are still eligible to receive aflibercept 8 mg. Two of the eyes had received lower-dose aflibercept 2 mg IVI s in the past without developing IOI. 3 of the 5 eyes developed IOI only after 3 or more IVI s aflibercept 8 mg. This is also in line with findings by Hoffmann et al., who reported 8 cases of IOI after receiving aflibercept 8 mg, with a majority of cases having had prior aflibercept 2 mg IVI s and 2 cases having had prior aflibercept 8 mg IVI s without IOI events [14]. Also one of the 3 cases of IOI reported by Matsumoto et al. had received prior aflibercept 2 mg IVI s without inflammatory response [10]. Therefore, based on our findings, we cannot state that either prior uneventful aflibercept 2 mg treatment or prior uneventful aflibercept 8 mg secures future aflibercept 8 mg treatment without adverse inflammatory events. Compared to the PULSAR trial which showed a rate of 1.3% developing IOI, the rate was higher in our study. Other real-world studies have reported IOI to a varying degree, ranging from 0.0% to 12.0% of the observed cohorts [9] [10] [11], [14]. While Matsumoto et al. and single case reports have reported retinal vasculitis in association with IOI [10], [13], [15], we did not see any signs of retinal vasculitis in the cases of IOI in our cohort.

Other anti-VEGF substances like ranibizumab, bevacizumab and aflibercept 2 mg have also shown to cause non-infectious IOI with variable incidence rates. Brolucizumab has been associated with cases of IOI accompanied with retinal vasculitis, typically developing with a delay of a few weeks after injection [17]. The underlying mechanism of these adverse reactions are unclear, but there are some hypotheses trying to explain. It may be possible that some patients are either inherently possessing antibodies against anti-VEGF drugs or may develop antibodies after exposure to anti-VEGF drugs, both resulting in an auto-inflammatory response [18]. Other hypotheses include anti-VEGF-induced endothelial cell damage, immunogenic alterations of anti-VEGF drugs by interactions with silicone oil released from syringes, contamination with endotoxins during manufacturing process and improper handling of IVI preparation [10], [17]. As for non-infectious IOIs following IVI s of other anti-VEGF drugs, the precise cause of IOIs seen in our study remains unclear.

One patient (1.2%) developed a massive decrease of platelet count in the blood following 3 IVI s aflibercept 8 mg. Bone marrow puncture revealed toxic damage as the most likely cause of bone marrow injury. After discontinuing aflibercept 8 mg, the platelet count spontaneously rose to normal ranges again. Toxic damage of bone marrow can have several causes, including drugs like chemotherapeutics, immunosuppressants, non-steroidal anti-inflammatory drugs (NSAID) and some other medications [19], [20]. However, anti-VEGF drugs have not been reported to cause bone marrow damage yet. Considerations regarding potential systemic effects, in particular coagulopathies, arise from observations made in cancer therapy, where systemically administered anti-VEGF medications are associated with thromboembolic and haemorrhagic events [21]. Moreover, several studies demonstrated lowered serum VEGF levels after intravitreal anti-VEGF injections [22], [23], [24]. Possibly specific risk groups such as elderly people with previous arteriothrombotic events or diabetic patients might be of greater risk for systemic adverse effects after anti-VEGF IVI s [25], [26]. However, the impact on the general population remains unclear [26]. Even though uptake of anti-VEGF into platelets can be demonstrated and might contribute to systemic adverse events [27], to our knowledge a reduction of blood platelet count in particular has not been found associated with anti-VEGF IVI s yet. Moreover, apart from medications, bone marrow suppression can be caused by several other diseases and environmental factors, including infection and auto-inflammatory disease [28], [29]. A causal relationship between intravitreal aflibercept 8 mg and low platelet count in the reported case is therefore uncertain.

The presented study has several limitations, such as its retrospective character and the single-center design. In this study, only short-term outcome of aflibercept 8 mg loading phase and the first interval up to 16 weeks, until another treatment was necessary, was assessed. These observations serve an important purpose of reporting first efficacy of the new therapeutic. However, expansion of IVI intervals, as demonstrated in PULSAR study, was not investigated. Retreatment behaviour beyond this relatively short follow up is still unknown in real-world setting and longer observation periods are needed to determine whether intervals can be extended. Another important limitation is the relatively small sample size of only 10 treatment-naïve eyes, which limits the ability of drawing conclusions in this subgroup.

In conclusion, aflibercept 8 mg has shown good short-term effectiveness in treating nAMD. Future studies with longer follow-up duration are necessary to determine long-term outcomes and necessary frequency of IVI s, as well as safety profile in real-world setting.

Conclusion Box

Already known:

• Pivotal trial PULSAR has shown promising results with extended injection intervals in a large part of newly diagnosed nAMD eyes treated with aflibercept 8 mg while achieving similar visual gains compared to eyes treated with aflibercept 2 mg.

• Limited data is available confirming effectiveness of aflibercept 8 mg in nAMD under real-world conditions, especially in pre-treated nAMD patients.

Newly described:

• This study indicates good effectiveness of intravitreal aflibercept 8 mg in real-world setting. A significant reduction of disease activity parameters in SD-OCT after loading phase with 3 monthly injections is shown, while BCVA remains stable. This finding applies to both pre-treated as well as therapy-naïve eyes.

Conflict of Interest

Grün M has received a speaker honorarium from Roche Pharma. Faatz H has received consulting honoraria from Novartis and Hexal/Sandoz. Lommatzsch A has received speaker honoraria from Roche Pharma and Bayer. Rothaus K declares to have no known competing financial interests.

• References

• 1 Congdon N, OʼColmain B, Klaver CC. Eye Diseases Prevalence Research Group. Causes and Prevalence of Visual Impairment Among Adults in the United States. Arch Ophthalmol 2004; 122: 477-485
• 2 Wong WL, Su X, Li X. et al. Global prevalence of age-related macular degeneration and disease burden projection for 2020 and 2040: a systematic review and meta-analysis. Lancet Glob Health 2014; 2: e106
• 3 Ciombor KK, Berlin J. Aflibercept–a Decoy VEGF Receptor. Curr Oncol Rep 2014; 16: 368
• 4 Bhisitkul RB. Vascular endothelial growth factor biology: clinical implications for ocular treatments. Br J Ophthalmol 2006; 90: 1542-1547
• 5 Rakic J-M, Lambert V, Devy L. et al. Placental Growth Factor, a Member of the VEGF Family, Contributes to the Development of Choroidal Neovascularization. Invest Ophthalmol Vis Sci 2003; 44: 3186-3193
• 6 Heier JS, Brown DM, Chong V. et al. Intravitreal Aflibercept (VEGF Trap-Eye) in Wet Age-related Macular Degeneration. Ophthalmology 2012; 119: 2537-2548
• 7 Wykoff CC, Brown DM, Reed K. et al. Effect of High-Dose Intravitreal Aflibercept, 8 mg, in Patients With Neovascular Age-Related Macular Degeneration: The Phase 2 CANDELA Randomized Clinical Trial. JAMA Ophthalmol 2023; 141: 834-842
• 8 Lanzetta P, Schulze A, Schmidt-Ott U. et al. Intravitreal Aflibercept 8 mg Injection in Patients With Neovascular Age-Related Macular Degeneration: 60-Week and 96-Week Results from the Phase 3 PULSAR Trial. Presented at the 23rd European Society of Retina Specialists (EURETINA) Congress. Accessed April 11, 2025 at: https://investor.regeneron.com/static-files/06016223-ff10-4fff-b9bb-3162f0ef27d9
• 9 Hosoda S, Sakurada Y, Fukuda Y. et al. Short-Term Outcomes of Three Consecutive Monthly Loading Administrations of Aflibercept 8 Mg for Treatment-Naïve Exudative Age-Related Macular Degeneration. Pharmaceuticals (Basel) 2025; 18: 438
• 10 Matsumoto H, Hoshino J, Numaga S. et al. Retinal vasculitis after intravitreal aflibercept 8 mg for neovascular age-related macular degeneration. Jpn J Ophthalmol 2024; 68: 531-537
• 11 Sambhara D, Vakharia P, Eichenbaum DA. Real-world efficacy and safety of 8 mg aflibercept in neovascular AMD: a case series. BMJ Open Ophthalmol 2025; 10: e002091
• 12 Binder KE, Bleidißel N, Charbel Issa P. et al. Noninfectious Intraocular Inflammation After Intravitreal Aflibercept. JAMA Ophthalmol 2025; 143: 499-506
• 13 Hashiya N, Maruko I, Miyaguchi Y. et al. Intraocular inflammation after intravitreal injection of aflibercept 8 mg for treatment-refractory neovascular age-related macular degeneration: a case report. BMC Ophthalmol 2025; 25: 42
• 14 Hoffmann L, Michels S, Eandi C. et al. Aflibercept high-dose (8 mg) related intraocular inflammation (IOI) – a case series. BMC Ophthalmol 2024; 24: 520
• 15 Sisk RA. Occlusive Retinal Vasculitis After Aflibercept 8 mg Injection for Wet Macular Degeneration. Retin Cases Brief Rep 2025;
• 16 Keenan TD, Vitale S, Agrón E. et al. Visual Acuity Outcomes after Anti–Vascular Endothelial Growth Factor Treatment for Neovascular Age-Related Macular Degeneration: Age-Related Eye Disease Study 2 Report Number 19. Ophthalmol Retina 2020; 4: 3-12
• 17 Anderson WJ, da Cruz NFS, Lima LH. et al. Mechanisms of sterile inflammation after intravitreal injection of antiangiogenic drugs: a narrative review. Int J Retina Vitreous 2021; 7: 37
• 18 Cox JT, Eliott D, Sobrin L. Inflammatory Complications of Intravitreal Anti-VEGF Injections. J Clin Med 2021; 10: 981
• 19 Galati G, Tafazoli S, Sabzevari O. et al. Idiosyncratic NSAID drug induced oxidative stress. Chem Biol Interact 2002; 142: 25-41
• 20 Kreipe HH. Medikamentös induzierte Knochenmarkveränderungen. Pathologie (Heidelb) 2022; 43: 256-262
• 21 Kamba T, McDonald DM. Mechanisms of adverse effects of anti-VEGF therapy for cancer. Br J Cancer 2007; 96: 1788-1795
• 22 Zehetner C, Kirchmair R, Huber S. et al. Plasma levels of vascular endothelial growth factor before and after intravitreal injection of bevacizumab, ranibizumab and pegaptanib in patients with age-related macular degeneration, and in patients with diabetic macular oedema. Br J Ophthalmol 2013; 97: 454-459
• 23 Carneiro ÂM, Costa R, Falcão MS. et al. Vascular endothelial growth factor plasma levels before and after treatment of neovascular age-related macular degeneration with bevacizumab or ranibizumab. Acta Ophthalmol 2012; 90: e25-e30
• 24 Matsuyama K, Ogata N, Matsuoka M. et al. Plasma levels of vascular endothelial growth factor and pigment epithelium-derived factor before and after intravitreal injection of bevacizumab. Br J Ophthalmol 2010; 94: 1215-1218
• 25 Bressler NM, Boyer DS, Williams DF. et al. Cerebrovascular accidents in patients treated for choroidal neovascularization with ranibizumab in randomized controlled trials. Retina 2012; 32: 1821-1828
• 26 Zarbin MA. Anti-VEGF Agents and the Risk of Arteriothrombotic Events. Asia Pac J Ophthalmol 2018; 7: 63-67
• 27 Sobolewska B, Fehrenbacher B, Münzer P. et al. Human Platelets Take up Anti-VEGF Agents. J Ophthalmol 2021; 2021: 8811672
• 28 Pascutti MF, Erkelens MN, Nolte MA. Impact of Viral Infections on Hematopoiesis: From Beneficial to Detrimental Effects on Bone Marrow Output. Front Immunol 2016; 7: 364
• 29 Zheng P, Chang X, Lu Q. et al. Cytopenia and autoimmune diseases: A vicious cycle fueled by mTOR dysregulation in hematopoietic stem cells. J Autoimmun 2013; 41: 182-187

Publication History

Received: 06 March 2025

Accepted: 08 July 2025

Accepted Manuscript online:
15 July 2025

Article published online:
01 September 2025