Age-Related Macular Degeneration Screening—What Is Next?

Abstract

Age-related macular degeneration is one of the leading causes of blindness worldwide in patients older than 60 years. Because of the aging population, it is of great importance to diagnose the disease early to help the patients get access to treatment. In its incipient stages, patients are usually asymptomatic, but the advanced disease can progress very fast, and it severely affects visual acuity. Over time there have been proposals for screening programs for patients with AMD, but at the moment there are no programs in most countries. The disease burden is significant because there is no cure for late AMD, especially for dry AMD. The management for most of the early and intermediate patients affected consists of antioxidant supplements and risk factors reduction. Anti-VEGF injections are the mainstay treatment for patients with neovascular AMD, which can improve and stabilize vision but can’t cure the disease. In the last decade there have been some technological advancements in what concerns the diagnosis and also the management of the disease, so for patients with AMD there is now more hope than ever in every stage of the disease.

Introduction

Age-related macular degeneration (AMD) is a frequent, multifactorial, degenerative retinal disease that affects the elderly [1]. AMD impacts the macular region of the retina and results in a gradual loss of visual acuity [1] and it is the leading cause of blindness in developed nations, accounting for 8.7% of all cases of blindness worldwide [2]. According to projections, 288 million people will have age-related macular degeneration in 2040 [2].

AMD can be classified into early or late AMD, which is further classified into two types: dry or wet AMD, based on the choroidal neovascularization presence or its absence [1].

The AMD diagnosis was made in the past based on clinical inspection or evaluation of colored fundus images [3]. Because technology improved a lot, now there are a lot of imaging techniques available, spectral-domain optical coherence tomography (OCT) is the most useful tool for AMD because it confirms the diagnosis, providing a lot of data concerning the drusen size, other lesion position, amount, and extension and disease activity [1].

The disease burden on patients with AMD is a cause for great concern both from a social and economic point of view. In a recent study, conducted in Spain by Ruiz-Moreno et al., the economic burden of AMD patients, reflected on the direct costs of treatment with mostly vascular endothelial growth factor inhibitors (anti-VEGF) and the disease diagnostic and monitoring and concluded that the disease has significant costs for the healthcare system and for the patients [4].

Risk factors

The risk of developing age-related macular degeneration is determined by environmental factors, and genetic and non-genetic factors [5,6]. It is found by multiple studies that increasing age has the most solid association with the disease, and almost all the patients with advanced AMD are older than 60 years [7,8]. In what concerns gender, AMD has a higher incidence in the female gender [1], though in the past there was not sufficient evidence for the association of the female sex with late AMD [7,9]. There is a racial variation suggested in the studies which showed a higher prevalence of early, intermediate and of geographic atrophy (GA) in white people with European ancestry, but no specific association with neovascular AMD, which was similar in prevalence for all ethnicities [2]. Genetics has a role in AMD pathogenesis. Some of the genes associated with AMD are specific coding variants of the complement pathway such as complement factors H and I, and also there is a causal role for the TIMP metallopeptidase inhibitor 3 [10]. Known environmental risk factors for developing AMD are cigarette smoking and sunlight exposure [5]. Another association with AMD incidence is represented by cholesterol levels, obesity, and general diet [11]. Increased antioxidant intake can reduce the risk of developing late-stage disease [12,13].

Pathogenesis

Drusen are cellular debris deposits that consist in lipids, proteins and carbohydrates located between the retinal pigment epithelium (RPE) and Bruch’s membrane. These are one of the first AMD characteristic lesions which can be seen clinically with ophthalmoscopy and appear as small white or yellow deposits at the macular level. Depending on their size, drusen can be classified as small, medium, or large [1].

Basal laminar deposits are located between the basement membrane of the RPE and its plasma membrane. Their consistency is mainly based on collagen and proteins. These result in a thickened inner collagenous layer of the Bruch membrane [1]. These correlate with retinal pigment epithelium abnormalities [14].

Basal linear deposits are located between the basement membrane of the RPE and the inner collagenous part of Bruch membrane. They consist in membranous materials and create localized aggregations and soft drusen [14].

The Bruch’s membrane thickens and loses some of its permeability as a result of drusen deposition in tandem with other structural and biochemical changes related to AMD pathogenesis, such as chronic activation of the complement cascade and inflammation. RPE pigmentary anomalies, such as hypo- or hyperpigmentation, are caused by neurodegenerative changes within the photoreceptor-RPE complex [15].

Subretinal drusenoid deposits (SDD) or reticular pseudo drusen resemble to soft drusen structurally but are located above the RPE. These lesions are associated with a reduction in the choroid capillaries’ vascular flow area and vessel volume and also with a reduced choroidal thickness [16].

Geographic atrophy (GA) is one of the late AMD subtypes. The patient presents with a loss of RPE cells, photoreceptors, and choriocapillaris in the concerned area [17]. The atrophic lesions are classified according to a new consensus based on the OCT findings as incomplete or complete outer retinal atrophy (iORA or cORA) and incomplete or complete retinal pigment epithelium + outer retinal atrophy (iRORA or cRORA) [18].

The lesions encountered in neovascular AMD may start at the choroidal level or in the outer retina. In neovascular AMD, RPE cells secrete VEGF [19] which promotes endothelial cell proliferation and migration which further generates angiogenesis [20]. Macular neovascularization (MNV) can be: type 1 MNV– the new vessels originate from the choroid capillaries and extend into the sub-RPE space, type 2 MNV – the membrane extends from the choroid through Bruch’s membrane and RPE into the subretinal space, and type 3 MNV starts from the retinal circulation and after it extends into the outer retina and choroid [21]. Polypoidal choroidal vasculopathy (PCV) is located in the sub-RPE space, resembling to type 1 MNV and it is a vascular network of branching vessels with aneurysmal dilations, called polyps which are likely to bleed [21].

Recently with ultra-widefield imaging technology it has been found that peripheral lesions are also frequently noticed in the eyes of patients with AMD, and it is claimed that the disease does not only affect the central retina, but also the peripheral part [22].

Diagnosis and classification

The usual clinical evaluation includes the following: visual acuity testing, fundus examination for detecting characteristic lesions (drusen, macular neovascularization, hemorrhages, etc.). In addition, retinal imaging such as OCT is mandatory, and some clinicians also perform fluorescein angiography. Fundus color photography is an old and good method of evaluation and also of progression monitoring [3]. Fundus autofluorescence images are very good for geographic atrophy diagnosis, measurement, and progression monitoring and this imaging method is widely used by clinicians and also in studies [23].

The signs of the disease are reduced visual acuity and metamorphopsia when tested with the Amsler grid. Symptoms vary with the stage of the disease. Early AMD patients can be asymptomatic or present only mildly reduced reading ability or mild central image distortion. Late AMD patients can present with severe image distortion, central vision loss, scotoma, reduced contrast sensitivity, difficulty performing daily tasks such as reading or driving, and even visual hallucinations [24].

Amsler grid is a square-shaped grid that can be used for self-monitoring in order to detect and monitor some of the frequent symptoms of the disease, metamorphopsia or scotoma [25]. Patients can be trained to use the test at home and when changes appear they should be acknowledged to undergo a retinal examination in an ophthalmology service, and by doing so, early diagnosis is promoted [25].

To diagnose AMD, the clinician must differentiate normal macular aging changes from degenerative abnormalities. In agreement with the Beckman classification [26] small drusen (<63 microns) represent normal aging modifications, medium drusen (>63 microns and <125 microns) is encountered in early AMD, large drusen (>125 microns) or retinal pigmentary epithelium abnormalities are found in intermediate AMD. Late AMD is the main cause of vision loss and it has two forms: dry AMD or geographic atrophy (GA) and wet AMD or neovascular form [26].

Multimodality

Color fundus photography (CFP) closely resembles biomicroscopic inspection and illustrates a wide variety of fundus abnormalities, including pigmentary abnormalities and various subtypes of macular drusen. The following criteria were used to categorize AMD: drusen size, consistency, location, quantity, and area of involvement. These plans also included measurements of GA size, position, and area as well as the location and quantitative area of hyper- or hypopigmentation [26,27]. With the development of more sophisticated retinal imaging techniques, CFP has lost favor as the main technique for the detection and monitoring of AMD.

Methods like multicolor confocal laser ophthalmoscopy (cSLO) (Figure 2-A), near-infrared reflectance (NIR) (Figure 2-D), fluorescein angiography (FA), and optical coherence tomography angiography (OCTA), retro mode imaging (RMI) (Figure 2-C) provide supplementary and confirmatory information, while OCT and fundus autofluorescence (Figure 2-B) have surfaced as important adjuncts for the tracking of AMD.

The retina, retinal pigmentary epithelium, and choroid can be examined in high-quality cross-sectional and en-face images using SD-OCT (Figure 1), which displays a segmentation in profoundness and has a precision that is comparable to that of light microscope histology [28].

OCTA uses motion contrast to identify blood flow and gathers volumetric information in three dimensions of the retina and choroid. The result is a high-resolution image with adequate depth segmentation of the vascular layers of the retina and choroid. All phases of AMD appear to be accompanied by changes in the choriocapillaris. While identifying retinal and choroidal neovascularization has been the main application of OCTA in AMD, dry AMD has also been studied using this technology [29,30].

FA is an old invasive method that has been the gold standard for neovascular AMD assessment, which implies dye injection, a long time dedicated to image acquisition and a reduced picture resolution. In a recent study conducted by Guliano et al concluded that CFP combined with SD-OCT has about the same diagnostic accuracy as CFP combined with FA [31].

FAF offers high-contrast retinal pictures that are particularly helpful for identifying atrophy-prone regions. Due to the loss of RPE cells, which contain intrinsic fluorophores like lipofuscin, atrophic lesions manifest as areas of reduced autofluorescence. Loss of retinal sensitivity is linked to the lack of signal. FAF is currently the anatomic parameter measured for tracking GA progression in most of the studies [29,32]. The most widely used FAF imaging technique at the moment uses a confocal SLO with a blue light excitation frequency filter of 488 nm and an emission filter of 500 to 521 nm [33].

Higher-end visible spectrum light frequencies, which are barely absorbed by media opacities, neurosensory retinal layers, and macular luteal pigments, are used in NIR imaging. Atrophic areas look brighter than regions without atrophy in contrast to blue-light FAF imaging [29]. With this imaging modality, reticular pseudo-drusen can be seen particularly clearly [34]. Another advantage is that patient discomfort is reduced compared to other imaging techniques.

Retro mode is an imaging technique that uses a scanning laser ophthalmoscope with a deviated aperture. It uses infrared laser radiation, that penetrates the deeper retinal layers and the choroid [35,36]. The result is a pseudo-three-dimensional image that highlights different retinal lesions such as drusen, [35,36] subretinal drusenoid deposits [37,38] or pigment epithelial detachments [39].

Machine learning and artificial intelligence in AMD

In what concerns recent technology advancements in ophthalmology, deep learning techniques, machine learning and computer-aided diagnostic systems can be used in order to diagnose and even stage the AMD, in concordance with current classifications used in practice based on the OCT findings and biomarkers detection [40,41]. Disease staging is important since every stage of AMD necessitates a different therapeutic approach, with the aim of preventing visual impairment. The machine learning systems can even predict progression of the disease or visual acuity loss [42]. These diagnostic and monitoring methods will need to be validated in time, but the benefits of standardized procedures using telemedicine in AMD patients would be very helpful for a lot of patients [40].

Screening for AMD

For a long time, it has been questioned if a screening programme for AMD would be cost effective, and it has been tried several times in different countries. In one screening programme, [43] the objectives and perspectives must were defined from the beginning, such as the natural disease progression, health-related quality of life relating to the disease, clinical presentation rates, the effectiveness of the screening test, treatment effectiveness and also the costs of blindness [43].

Figure 1. SD-OCT section though the fovea showing drusen which apear as small elevations of the retinal pigment epitelium.

Figure 1. SD-OCT section though the fovea showing drusen which apear as small elevations of the retinal pigment epitelium.

Figure 2. Early AMD with multimodal imaging showing drusen on A. Color scanning laser ophthalmoscopy B. Barely visible on blue-FAF C. Small elevations on retromode imaging and on D. Near Infrared imaging showing spotted areas.

Figure 2. Early AMD with multimodal imaging showing drusen on A. Color scanning laser ophthalmoscopy B. Barely visible on blue-FAF C. Small elevations on retromode imaging and on D. Near Infrared imaging showing spotted areas.

A recent investigation conducted in Japan [44] that aimed to generate a screening program model for AMD by using a Markov model of the best-case-scenario analysis. By screening patients starting at the age of 40 years and ending at age of 74 years there would be a 40% decline in the number of patients who lose their sight to AMD [44]. The study concluded that a screening for AMD reduces blindness but is not cost-effective [44].

Another option would be implementing AMD screening as part of the diabetic retinopathy (DR) screening and be performed at the same time. Chan et al. concluded that including intermediate AMD screening into the routine DR screening for patients with diabetes would be cost-effective, and this would mean using the images for the DR screening and grading them also for AMD, and also further treating the diagnosed patients with oral supplements [45].

Telehealth strategies in ophthalmology were tested for AMD screening, and probably in the future, there will be some solutions regarding including a tool kit with a fundus camera and an OCT connected to a network, which would help a lot of patients in receiving the treatment they need [46].

Garcia-Layana et al. proposed a screening tool for AMD self-evaluation with a questionnaire that can be useful to evaluate the risk of advanced disease and does not necessitate an eye examination [47].

Prevention and treatment

The AREDS study [12] proved that a formulation of oral supplements containing antioxidants (vitamin C, vitamin E, beta-carotene, and copper) and zinc reduced the AMD progression risk to the advanced disease by around 25%. In the AREDS 2 study, beta-carotene was replaced with lutein and zeaxanthin, which were considered superior for a couple of reasons, one of which is reducing the risk of lung cancer in past smokers [13].

Anti-VEGF agents like bevacizumab, ranibizumab, [48] aflibercept [49] or brolucizumab [50] have proved to be effective for intra- and subretinal fluid resolution and visual acuity improvement in patients with neovascular AMD and are now standard in most of the countries [51].

At the moment there are a series of on-going clinical trials for wet AMD as well as for dry AMD. Anti-VEGF gene therapy for neovascular AMD [52] as well as multiple trials for non-neovascular AMD, including complement factor inhibitors such as Pegcetacoplan, Avacincaptad, Eculizumab and cardiolipin stabilizer Elamiprimide, which is meant to reduce the reactive oxygen species production in the mitochondria [53]. Other therapeutic options that look promising include cellular regeneration therapy, RPE transplantation, and photoreceptor transplantation [54].

Conclusion

Now more than ever patients with AMD have some kind of therapeutic option for every stage of the disease, and a lot of new treatments are to be soon available that hopefully will cure the disease, not only slow the progression. Probably when ocular telehealth strategies will be implemented more and better in developed and developing countries, especially addressing lower-income areas, screening for AMD will become widely available and will help many patients, preventing blindness and saving years of sight.