Fluid as a critical biomarker in neovascular age-related macular degeneration management: literature review and consensus recommendations

Current guidelines on the management of patients with neovascular age-related macular degeneration (nAMD) lack clear recommendations on the interpretation of fluid as seen on optical coherence tomography (OCT) imaging and the incorporation of this information into an ongoing disease treatment strategy. Our objective was to review current guidelines and scientific evidence on the role of fluid as a biomarker in the management of nAMD, and develop a clinically oriented, practical algorithm for diagnosis and management based on a consensus of expert European retinal specialists. PubMed was searched for articles published since 2006 relating to the role of fluid in nAMD. A total of 654 publications were screened for relevance and 66 publications were included for review. Of these, 14 were treatment guidelines, consensus statements and systematic reviews or meta-analyses, in which OCT was consistently recommended as an important tool in the initial diagnosis and ongoing management of nAMD. However, few guidelines distinguished between types of fluid when providing recommendations. A total of 52 publications reported primary evidence from clinical trials, studies, and chart reviews. Observations from these were sometimes inconsistent, but trends were observed with regard to features reported as being predictive of visual outcomes. Based on these findings, diagnostic recommendations and a treatment algorithm based on a treat-and-extend (T&E) regimen were developed. These provide guidance on the diagnosis of nAMD as well as a simple treatment pathway based on the T&E regimen, with treatment decisions made according to the observations of fluid as a critical biomarker for disease activity.


Introduction
Since the widespread introduction of optical coherence tomography (OCT) for the visualisation of the back of the eye in patients with eye diseases such as neovascular agerelated macular degeneration (nAMD), the evaluation of lesion morphology using OCT has become a key part of the clinical decision-making pathway [1]. Markers for disease activity based on OCT, including intraretinal and subretinal as well as subretinal pigment epithelium (RPE) fluid, are crucial for guiding management and treatment frequency of nAMD patients.
Recent advances in OCT technology have led to increases in speed and resolution that permit the detection of small structural changes to the retinal layers [2]. However, the interpretation of OCT images can be complex and challenging. Although this is an area of considerable scientific interest and extensive literature exists which attempts to evaluate the influence of different types of fluid on outcomes in nAMD, current guidelines may be lacking or open to misinterpretation when it comes to translating the diagnostic findings from an OCT into an ongoing disease treatment strategy. Clear treatment recommendations that consider both clinical and real-world considerations are therefore required.
The objective of this consensus article is to review the current guidelines and scientific evidence on the role of fluid as a biomarker in the management of nAMD and provide clinically useful recommendations based on a consensus of expert European retinal specialists. Furthermore, limitations of current literature and areas of further research are also highlighted.

Methods
A preliminary review of the literature on the role of fluid in the management of nAMD was performed by Novartis in preparation for a roundtable discussion with European retinal specialists (consensus panel, consisting of LK, MP, RDM, FGH, MRM, MN, FR, RS, SJT, JZV and SAZ), held in Zurich, Switzerland (19 July 2019). During this initial meeting, the available scientific evidenceand the lack of it-were discussed, resulting in the proposal from the consensus panel to develop simplified treatment recommendations in nAMD. The literature review was subsequently repeated with revised search parameters and the updated results were subject to further review by the consensus panel during the development of the treatment recommendations, ensuring scientific rigour and unbiased interpretation. Novartis was not involved in the interpretation of the literature search results or the development of the treatment recommendations.
The repeated literature search of PubMed was performed according to the predefined search parameters shown in Table 1, with other relevant publications included from information sources such as recent congress presentations and educational resources. The resulting publications were screened by title and abstract for relevance and according to the following exclusion criteria: case reports and studies with fewer than 50 patients; opinion pieces other than expert consensus recommendations and guidelines; non-English language publications; and publication date prior to 2006. The scientific evidence that was retrieved by the search was tabulated and graded according to recent European guidance [3].
The evidence was discussed by the consensus panel and used, along with their expert opinion and experience, to inform the development of a consensus management algorithm for patients with nAMD based primarily on observations of fluid from OCT monitoring. The nAMD-specific terminology used within this article follows recent consensus nomenclature for reporting nAMD data [4]. The term intraretinal fluid (IRF) is used throughout the document to standardise the different terms used to describe the presence of fluid within the retina including intraretinal cystoid oedema, intraretinal cysts, cystoid oedema, cystoid macular oedema and retinal fluid.

Results
The literature review was performed on September 25, 2019 ( Fig. 1). After screening of 654 publications and excluding those that were not relevant or were outside the scope of the review, a total of 66 publications were included. Of these, 14 publications were treatment guidelines, consensus statements and systematic reviews or meta-analyses, while 52 publications reported primary evidence from clinical trials, studies, and chart reviews.

Treatment guidelines
Six treatment guidelines from institutions in Europe and the USA were retrieved by the search. In the Royal College of Ophthalmologists' (RCOphth) guidance on the use of ranibizumab in nAMD from 2009, new subretinal fluid (SRF) with or without haemorrhage is included as one criteria for treatment initiation, while their definition of disease activity for continuation of treatment includes IRF, SRF, sub-RPE fluid and haemorrhage [5]. In later guidelines on AMD from 2013, the RCOphth provided similar recommendations relating to fluid [6]. These latter guidelines have since been archived following the publication of the National Institute for Health and Care Excellence (NICE) guidelines on age-related macular degeneration (AMD) in January 2018. In these, NICE states that OCT should be offered to individuals with suspected active nAMD, or for ongoing monitoring of patients with active nAMD. No specific guidance is given with regard to fluid and treatment or management of the condition [7].
Few guidelines distinguished between types of fluid when providing recommendations, with the same retreatment approach generally recommended regardless of the type and location of fluid observed. One of the few that made a distinction between fluid types was the 2014 EURETINA guideline on nAMD, which advised that IRF, SRF and RPE detachments are important signs of neovascular activity independent of central retinal thickness (CRT), and that a 'zero tolerance' approach to OCT criteria is justified given the rapid progression of exudative features and progressive loss of vision when initiation of treatment is delayed in nAMD. However, longstanding persistent IRF should be considered a sign of irreversible retinal damage which should not prompt continued retreatment. Performing OCT was recommended as the most useful tool for evaluating morphological changes since it provides the most accurate reflection of the recurrence of disease activity. Qualitative morphology-based OCT data were considered to be more sensitive than current quantitative measurements such as CRT for detecting choroidal neovascularisation (now termed macular neovascularisation [MNV]) activity [8].
In the American Academy of Ophthalmology preferred practice pattern for AMD from 2015, there is no specific mention of how to interpret retinal fluid in diagnosis or follow-up, other than a statement that as-needed treatment should be based on the presence or absence of SRF or IRF [9]. Finally, recommendations on outcome measures for macular degeneration provided by the International Consortium for Health Outcomes Measurement and a group of experts in 2016 advised that the presence of IRF, SRF or haemorrhage attributable to neovascular lesion activity (as determined by the treating ophthalmologist) should be assessed at each clinic visit [10].

Consensus statements
A number of expert consensus statements have provided guidance on the management of nAMD including recommendations relating to fluid and other anatomical parameters visualised using OCT. These are broadly consistent but differ in the detail of interpreting the various morphological features.
In 2011, a group of 22 European experts provided consensus recommendations for anti-vascular endothelial growth factor (VEGF) management of nAMD based on morphological criteria. Suggested retreatment criteria under a pro-re-nata (PRN) regimen included IRF, SRF, diffuse foveal thickening and expanding serous pigment epithelium detachment (PED). Criteria for delaying treatment included the absence of the above criteria, stable serous PED and stable IRF that has not responded to three intravitreal injections [11]. Notably, at the time of these recommendations, retreatment criteria were based on the assessment of a single transfoveal OCT image [11]. A committee of UKbased retinal experts published a consensus paper defining response to anti-VEGF therapy in nAMD in 2015. They noted that there is often little correlation between morphological and functional responses to anti-VEGF treatments, and so recommended a combination of morphology and function as the means of determining treatment response, with the morphology component defined as IRF, SRF and retinal thickening [12].
In a 2017 expert round-table consensus on the treatment of nAMD with aflibercept in the second year of therapy, fluid was a recommended consideration when making the decision to maintain a fixed regimen or move to a treat-andextend (T&E) dosing schedule. The criteria for not extending the treatment interval included persistent macular fluid with stable vision, recurrent fluid, and decrease in vision in the presence of fluid. Extension of intervals between treatments was recommended for eyes with no macular fluid and stable vision [13].
A recent Greek consensus statement on the management of nAMD recognised the importance of morphological signs of disease activity observed using OCT, which the authors note correspond to early signs of recurrence prior to measurable loss of VA. The main anatomic parameters to be taken into consideration according to their recommendations were CRT, SRF, IRF, anatomy of the outer retinal layers and PED [14].

Systematic reviews
Four systematic reviews were included in the literature review. The earliest of these was a systematic review on OCT for diagnosis, monitoring and guiding treatment for nAMD by Mowatt and colleagues from 2014, which concluded that strategies involving OCT alone for diagnosis and/or monitoring were unlikely to be cost-effective, while those that also included fluorescein angiography (FA) and other imaging techniques were more likely to be considered cost-effective. However, many of the studies included in this review used older, time-domain OCT technology which may have compromised the specificity of the technique in terms of detecting active nAMD. For the purposes of this review, nAMD was considered to be active or inactive, with no specific discussion relating to fluid [15]. In contrast, a review by Schmid-Erfurth and Waldstein from 2016 provided detailed information on imaging biomarkers in nAMD. The authors concluded that CRT is an inferior prognostic biomarker for guiding retreatment compared with localisation of fluid in different compartments, including IRF and SRF. IRF at baseline is negatively associated with VA, while SRF at baseline (i.e., in naive patients) is associated with superior visual benefits and a lower rate of progression towards atrophy. The finding of SRF is associated with all lesion types and is typically the first exudative sign in Type 1 lesions. RPE detachment was identified as unresponsive to therapy and responsible for visual decline [16]. A later systematic review on OCT in the management of AMD by the same group provided a detailed discussion of morphological features indicative of disease activity, but was accompanied by no clear guidance for treatment [17].
A systematic review of the evidence on using morphological predictors to modify treatment protocols in nAMD was performed by Ashraf et al. [18], finding that a good response in terms of reduction in SRF at 12 weeks predicted good visual outcomes, but that patients with PED and IRF achieved smaller visual gains and their treatment intervals should be extended with caution.

Primary evidence
The 52 primary publications of clinical trials and studies retrieved by the literature search were reviewed for relevant detail on the role or impact of fluid in nAMD. Of these, one publication was the primary output of a randomised controlled trial (RCT) [19], 21 publications were post-hoc analyses, exploratory analyses and prospective cohort studies related to several medium and large RCTs (ABC trial [20], PIER [21], CATT [22][23][24][25][26][27], MONT BLANC [28], EXCITE [29,30], GEFAL [31], VIEW 1 and 2 [32][33][34][35][36] and HARBOR [37][38][39]), 5 publications described prospective, nonrandomised studies, and the remaining 25 publications were retrospective chart reviews and case series. Table 2 provides a summary of the studies and their findings, while Table 3 compares features of several of the key RCTs of anti-VEGF therapy in nAMD, including the retreatment criteria applied to the flexible treatment arms or phases of these trials.
Several studies reported that the presence of baseline SRF predicts a good response to anti-VEGF treatment, resulting in favourable visual outcomes [25,26,37,[40][41][42]. Evidence also suggests that small amounts (defined by the FLUID study as less than 200 µm) of residual stable SRF can be tolerated without impact on VA [19,43]. However, one study reported that recurrent SRF is predictive of a poor functional prognosis [44]. A substantial number of studies reported consistent findings indicating that the presence of IRF (at baseline or recurring) is predictive of a poor prognosis [22,23,25,27,31,41,42,[44][45][46][47][48][49][50][51][52]. New or growing sub-RPE fluid is reported in one study as being an early sign of conversion to nAMD [53], while another study found an increase in sub-RPE fluid to be a marker for progressive disease activity which warrants treatment [32]. While this could be considered a useful predictive biomarker if observed over time, it should be noted that the presence of sub-RPE fluid in a single OCT scan without the presence of SRF and IRF is not necessarily indicative of disease progression.
Several publications found the type or spatial localisation of fluid to have limited prognostic value in terms of predicting response to anti-VEGF therapy [33,38,54,55]. In the VIEW studies, a post-hoc analysis reported that BCVA • SRF volume modestly correlated with contrast sensitivity (P = 0.004) • Increased retinal thickness at the foveal centre correlated with decreased VA (P < 0.001) 4 Mariani et al. [73] Consecutive case series Treatment-naive nAMD beginning treatment with ranibizumab 99 ns • Loss of VA after the initial 3 months was associated with the presence of PED at baseline (P = 0.01) 4 Silva et al. [79] Single-centre, prospective, observational, longitudinal 2-year study with 1-year extension nAMD in the non-study eye and early age-related maculopathy in the fellow (study) eye 62 62 • OCT findings did not show any correlation with an increased risk of conversion to nAMD 4 Kolb et al. [55] Retrospective, interventional case series Treated with anti-VEGF and followed for ≥12 months ns 75 • CRT, integrity of the ellipsoid zone or presence of IRF or SRF was not predictive of VA outcome after 12 months of treatment • Eyes with thicker CRT at baseline had a greater reduction in CRT which was associated with better VA outcomes 4 Padnick-Silver et al. [53] Prospective, observational, non-randomised study nAMD in the non-study eye and non-exudative macular degeneration in the fellow (study) eye 79 79 • In 13/15 patients who developed nAMD in the study eye over 2 years, disease progression was identified on OCT before FA or examination showed changes • New or growing sub-RPE fluid and small areas of SRF or IRF were early signs of conversion to nAMD 4 Wickremasinghe et al. [50] Prospective, consecutive, nonrandomised study    Vogel et al. [42] Retrospective, consecutive chart review nAMD treated with anti-VEGF with follow-up of  Fluid as a critical biomarker in neovascular age-related macular degeneration management change from baseline to week 52 was independent of the presence or absence of fluid at week 12 [34]. However, contrasting evidence from the PIER study suggests that an absence of fluid on OCT is predictive of greater BCVA gains with anti-VEGF treatment [21]. A significant number of studies found either abnormally thick or abnormally thin retinas to be associated with poor outcomes [20,22,25,27,31,49,52,56,57]. Reduced retinal thickness has been associated with decreased retinal sensitivity [58], and an increased total volume of subretinal tissue has been correlated with decreased VA or contrast sensitivity [20,57]. In a retrospective study of patients initially treated with ranibizumab and then switched to aflibercept, subfoveal thickening and increased retinal central subfield thickness were reported to be predictive of poor prognosis in non-treatment naive patients [49]. In contrast, however, two studies reported that change in retinal thickness is not predictive of treatment outcomes [54,59].
Finally, a small number of publications commented on correlations between fluid and required anti-VEGF injection frequency. Two publications reported that a thicker retina at baseline was associated with greater injection requirements [37,60], while another stated that the presence of SRF was predictive of the need for a higher injection frequency [28]. A post-hoc analysis of the VIEW studies reported that the absence of retinal fluid at 1 year was predictive of the ability to achieve extended treatment intervals of at least 12 weeks [36].

Algorithm for the management of nAMD
Based on the available scientific evidence described above and the experience of the consensus panel, an algorithm for the most optimal management of patients with nAMD based on fluid observed using OCT and other imaging technologies is recommended, irrespective of country guidance and resource constraints, as shown in Fig. 2.

Aim to iniƟate treatment as soon as possible following diagnosis (usually within 1 week according to most European recommendaƟons)
Signs of disease acƟvity present?

If 3 consecuƟve visits occur with insufficient anatomical treatment response
If nAMD diagnosis confirmed to be correct ConƟnue to treat If there is sƟll insufficient anatomical response aŌer at least 2-3 addiƟonal injecƟons, consider switching to an alternaƟve anƟ-VEGF agent When switching to a different agent, this can be done with or without an iniƟaƟon phase based on the type of lesion, level of disease acƟvity and paƟent characterisƟcs, according to the judgment of the clinician Check diagnosis, performing addiƟonal imaging if needed and considering potenƟal misdiagnoses (see Table 4) One or more parameters present?

Yes No
Consider differenƟal diagnoses (see Table 4) Other imaging modaliƟes as required in order to confirm diagnosis based on observaƟons

Diagnosis and diagnostic techniques
The consensus panel agreed that morphological parameters observed on OCT are the most important criteria in routine clinical practice for the diagnosis of nAMD. The whole stack of images should be used, to give as full a picture as possible. The characteristic features considered to be indicative of nAMD are SRF, IRF and fibrovascular PED (Fig. 2a). OCT can permit differentiation of the type of MNV and location of fluid, but at baseline many cases also require additional information from other imaging modalities in order to confirm the diagnosis. If available, OCTangiography (OCT-A) is considered to be valuable in order to visualise the neovascular complex. If OCT-A is not available, FA can be used to visualise leakage from the lesion, but is otherwise no longer judged to be a mandatory technique for nAMD diagnosis in all AMD cases. However, clinical signs visualised using biomicroscopy or fundus photography/examination are still considered useful to support the diagnosis. In cases where PCV or type 3 MNV is suspected, ICGA and OCT-A are recommended to confirm this diagnosis. In addition to the morphological and clinical signs of nAMD, patient age over 50 years is an important criterion for a diagnosis of nAMD. A range of conditions have the potential to masquerade as nAMD. Table 4 lists a number of these potential misdiagnoses or pitfalls. Of these, the most common are adultonset foveomacular vitelliform dystrophy and central serous chorioretinopathy. When examining a patient with putative nAMD, the clinician should be aware of and exclude these common differential diagnoses.

Treatment
The agreement of the consensus panel was that, regardless of the anti-VEGF agent used, T&E is the recommended regimen for the management of nAMD because it provides comparable clinical outcomes to fixed monthly or bimonthly injections with a reduction in injection burden compared with fixed dosing [61,62]. T&E also provides a reduction in the number of clinic visits compared with PRN with monthly monitoring, provided that a one-step visit (with follow-up and injection on the same day) is possible. The potential for development of atrophy with intensive anti-VEGF therapy was considered to be less of a concern than the likelihood of visual acuity loss resulting from undertreatment, since a link between the number of injections and the risk of developing atrophy has never been proven. On the contrary, current evidence suggests that anti-VEGF therapy is not a significant risk factor for the development of macular atrophy. For example, a post-hoc analysis of data from the HARBOR study reported no association of number of ranibizumab injections with macular atrophy development, and no significant association between regimen (monthly vs PRN treatment) and macular atrophy development [63]. Another post-hoc analysis of data from the same trial using Classification of Atrophy Meetings (CAM) group atrophy criteria found no differences in the incidence or progression rates of new macular atrophy among study arms, anti-VEGF doses, or treatment regimens [64]. However, there is evidence that neovascularisation type may be associated with the development of atrophy, with patients with type 1 MNV at baseline less likely to develop atrophy than eyes with other forms of MNV [65]. In contrast, patients with type 3 MNV and subretinal drusenoid deposits at baseline have a high risk of atrophy development [66]. Treatment with anti-VEGF therapy should be initiated as soon as possible once the diagnosis of nAMD is made. Guidelines and institutional guidance vary throughout Europe, with most recommendations advising that treatment should take place within 1 week of referral. The NICE guidelines mandate treatment within 14 days of referral, but specify that referral should take place within 1 working day of diagnosis [7]. Treatment should begin with an initiation phase before the clinician considers extending the treatment interval (Fig. 2b). This usually consists of three injections given at monthly intervals, but in some circumstances (as seen in real-world datasets [67]) could entail just two injections depending on the response of the individual patient. After that, the patient is evaluated for extension criteria, and the treatment interval can be increased by 2-4 weeks at a time.
Treatment should be given proactively at each visit-a key aspect of T&E treatment design. The decision on whether the treatment interval should be extended, retained, or reduced is also made at each visit, and is based on disease activity as assessed using OCT. The signs of disease activity that should trigger a reduction in treatment interval include new haemorrhage beneath or within the retina, new or persistent IRF, new or increased SRF, increased size of PED, or the presence of subretinal hyperreflective material which would indicate the presence of a neovascular membrane. If one or more of these signs are present, the treatment interval should be reduced by 2-4 weeks, to a minimum of 4 weeks. However, in a minority of cases with recurrent disease activity, the clinician may feel that an extension or reduction of 1 week might be more appropriate.
If disease activity is observed at three consecutive visits, with no sign of anatomical and/or functional improvement, the clinician should consider whether the initial diagnosis of nAMD was correct, using additional imaging modalities to provide more information if necessary. It might be possible that the patient is not a non-responder but has instead been misdiagnosed for nAMD. If further investigation confirms the original diagnosis, then the consensus panel recommends that treatment should be continued for at least 2 to 3 additional injections at the minimum interval permitted by the product label before a switch to an alternative anti-VEGF therapy is considered. If a patient is switched to a different anti-VEGF therapy due to lack of efficacy, this should be done with a new initiation phase. However, a simulated switching study has suggested that continuation of initial therapy will, in many cases, result in a gradual improvement or stabilisation similar to that commonly reported following a therapy switch in published anti-VEGF switching studies [68].
If there is no evidence of disease activity at the treatment visit, the clinician may consider extending the treatment interval by 2-4 weeks [69], to a maximum of 16 weeks (or potentially more with longer-acting anti-VEGF agents), however, there will be a higher risk of recurrence [67]. If a patient reaches stability at a particular treatment interval, this should be maintained over the long term if feasible. If the treatment interval is alternately being extended and reduced at each visit, the clinician can consider that the shorter of the two intervals is the more appropriate one for the patient and maintain this interval for a period of time before re-evaluating the patient's treatment needs in due course. There is currently no evidence to support stopping anti-VEGF treatment in patients with stable disease, as disease activity will very likely recur, but the clinician may consider that continued long-term anti-VEGF therapy may not be appropriate or beneficial in patients with low vision who have fibrosis, extensive subretinal haemorrhage, subfoveal disruption of the external limiting membrane or the ellipsoid zone or central atrophy of the RPE.

Discussion and conclusion
The aim of this consensus article is to consider the evidence and guidance currently available in the scientific literature on the role of fluid in the management of nAMD and provide recommendations as to how it might be integrated into everyday clinical practice based on the opinion of a panel of expert retinal specialists. Our understanding of the role of fluid in nAMD is still evolving and in some instances the observations reported in the scientific literature are conflicting and confusing. The treatment recommendations provided here are based on our best interpretation of the available data at this time. The resulting algorithm for the diagnosis and management of nAMD provides clear guidance on recommended diagnostic tools and what they can be used to identify, as well as a simple treatment pathway based on the T&E regimen. It aims to provide the best possible visual outcomes for patients whilst acknowledging the restrictions that are inevitably encountered in real-world clinical practice. Treatment decisions are made according to observations of fluid as a biomarker for disease activity in nAMD. This publication is not an exhaustive review of the T&E regimen, which varies in detail between publications, but provides a recommended version of the T&E regimen based on the combined clinical experience of the consensus panel, and guided by fluid.
The detection of fluid on OCT is generally used to imply the presence of a VEGF-related leak that the clinician could expect to respond to anti-VEGF therapy. However, in some cases, the fluid spaces seen on OCT may actually be structural changes such as outer retinal tubulation that are not responsive to anti-VEGF treatment [70]. Where this is suspected, strategies to confirm that fluid is VEGF-driven include monitoring patients shortly after treatment (e.g., 2 weeks after injection) to check for a short-lived treatment response, checking for leakage from the lesion using FA, and assessing whether the putative fluid worsens with an extended treatment interval.
There are a number of limitations associated with this review and consensus. The scientific evidence reviewed here is limited in that the literature search retrieved only one level 1 evidence trial (the FLUID study) that specifically aimed to evaluate the impact of fluid in the management of nAMD [19]. Even this study had limitations in terms of determining the effect of treating fluid versus leaving it untreated since patients were treated at every visit in both treatment arms. Interestingly, both the arms where SRF was more tolerated and the arm where it was treated more aggressively had relatively high and nearly identical injection frequencies (means of 15.8 and 17 injections over 2 years). The remainder of the evidence comes from a number of RCTs in which the effect of fluid on treatment outcomes was an observational, secondary or exploratory outcome or the subject of a post-hoc analysis, or in the form of lower level evidence from prospective but uncontrolled trials and retrospective chart reviews. For the purposes of this review, all publications that met the literature review inclusion criteria have been considered, regardless of the level of evidence.
The imaging technologies used in the studies included here have not remained constant over the time span of the literature review. OCT has evolved from time domain to spectral domain modalities, meaning that the observations reported by the earliest publications returned by the literature search are not directly comparable to the more recent publications. Several of the larger RCTs used these older imaging techniques, which may detract from the relevance of their findings to current practice. Other measured parameters may have also changed over time.
A final limitation of this work is that for the purposes of providing clear guidance that can be used on a day-to-day basis by the practising clinician, this review and consensus focusses only on the role of fluid in nAMD. In addition to fluid, a range of other morphological features visible on OCT such as external limiting membrane, ellipsoid zone and RPE disruption, and the presence of PED and hyperreflective material have been associated with poor visual acuity outcomes [23, 24, 27-29, 31, 35, 40, 42, 45, 49-52, 71-75]. A relationship has also been observed between visual acuity outcomes and the type of neovascularisation. At baseline, type 1 MNV is a predictor of better visual acuity following anti-VEGF treatment compared with other lesion types, and eyes with this type of lesion often have SRF [76]. These associated morphological findings are also important features with prognostic value which can co-exist with fluid. However, given the difficulties in visualising, identifying and consistently assessing some of these other features, we are of the opinion that fluid is the most practical and useful biomarker of VEGF upregulation and MNV activity in nAMD.
In conclusion, gaps exist in the scientific literature on the role of fluid in the management of patients with nAMD. The limitations described here highlight the real need for appropriately designed and executed studies to provide a standardised and detailed understanding of the appearance of different specific fluid manifestations and their consequences on clinical outcomes. However, it is quite clear that the primary treatment goal is to eliminate fluid as effectively as possible. Future research into this important area could provide valuable insights to direct optimal treatment to achieve this. In the meantime, following expert consideration of the evidence available, we recommend that patients with nAMD receive anti-VEGF therapy according to a T&E regimen with treatment intervals determined according to fluid-based disease activity parameters observed using OCT. funding for research from: Allergan, Bayer, Novartis and Sifi. FR is a consultant for Allergan, Alimera, Bayer, MS&D, Novartis and Roche. RS is a consultant for Allergan, Alimera, Bayer, Novartis and Théa. SJT is a consultant for Bayer and Novartis, received grant support from Bayer, Novartis and Heidelberg Engineering, and is involved in research for Allergan, Roche, Bayer, Novartis and Boehringer-Ingelheim. JZV is a consultant for Alcon, Alimera Sciences, Allergan, Bayer, Novartis and Roche, has received grants from Allergan and Novartis and is a lecturer for Bausch and Lomb, DORC, Topcon and Zeiss. SAZ is a consultant for Allergan, Bayer HealthCare, Novartis and Roche and has received grant support from Bayer HealthCare and Novartis.
Publisher's note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons. org/licenses/by/4.0/.