14 Age-Related Macular Degeneration
Age-related macular degeneration (AMD) is a chronic, progressive disease of the macula, and it is the leading cause of severe irreversible central vision loss and legal blindness in individuals 65 years of age or older in the United States and in other Western countries. 1 , 2 , 3 , 4 , 5 , 6 A meta-analysis reported that, in Caucasians aged 40 years and older, the prevalence of early AMD was 6.8%, and the prevalence of late AMD was 1.5%, with prevalence increasing with patient age. 7 Overall, AMD is estimated to affect more than 8 million people in the United States, with advanced AMD affecting 1.75 million individuals. 6
AMD is the leading cause of irreversible central vision loss in people older than 65 years in the United States.
AMD is classified into two well-defined clinical forms. The more common form is non-neovascular AMD (non-NVAMD), also known as “dry” or “nonexudative” AMD. 8 , 9 The second form is NVAMD, also referred to as “wet” or “exudative” AMD. Differentiating dry versus wet disease has important implications for both prognosis and treatment. Findings of non-NVAMD include abnormalities of the retinal pigment epithelium (RPE) and drusen. Early dry AMD is associated with minimal visual acuity loss; however, it may progress to late dry AMD, otherwise referred to as geographic atrophy, where there is loss of the RPE. Late dry AMD accounts for nearly 20% of legal blindness related to AMD. 9 , 10 NVAMD is defined by the appearance of choroidal neovascularization (CNV) and can be associated with severe vision loss with subsequent subretinal fibrosis or disciform scarring if left untreated. While CNV occurs in 10 to 15% of all patients with AMD, it accounts for more than 80% of cases of severe central vision loss associated with AMD. 11
In the past two decades, significant advancements have been made in the diagnosis of AMD, especially the development of optical coherence tomography (OCT). Historically, clinical biomicroscopy with a contact lens and imaging in the form of fluorescein angiography were used for the evaluation of CNV; however, current OCT scanners are able to provide rapid, noninvasive cross-sectional images of the retina and even choroid, rendering contact lens examinations nearly obsolete. With the evolution of technology from time domain to spectral domain and now swept-source OCT, physicians are able to obtain progressively higher resolution OCT images, which provide detailed information about the pathological changes in AMD. 12
Prior to 2004, treatment options for exudative AMD included laser photocoagulation, verteporfin (Visudyne, Valeant) in photodynamic therapy (PDT), submacular surgery, and macular translocation. Unfortunately, these therapies had minimal success in maintaining vision, and most patients experienced progressive vision loss. In 2004, the treatment of exudative AMD was revolutionized by the advent of anti–vascular endothelial growth factor (anti-VEGF) therapy. Instead of preventing vision loss, the prognosis of exudative AMD was redefined, with the potential to improve vision, and in most cases preserve vision. 9
This chapter reviews the current knowledge of AMD, critical to understanding present management and potential future treatment strategies.
14.2 Non-Neovascular Age-Related Macular Degeneration
14.2.1 Clinical Features
The features of non-NVAMD include drusen (singular, “druse”), areas of increased pigmentation or hyperpigmentation in the outer retina associated with drusen, and areas of depigmentation or hypopigmentation of the RPE. The later stages of dry AMD may include geographic and nongeographic areas of RPE atrophy. All of these features are termed “non-neovascular,” as they do not involve manifestations associated with the growth of CNV (a feature of the late stage of AMD), discussed in the second section of this chapter. 10
The pathogenesis of early AMD is characterized by thickening of Bruch’s membrane due to lipid and protein accumulation, leading to the formation of sub-RPE deposits referred to as “drusen.” 10 Drusen appear as yellowish white lesions at the level of the RPE and can appear clinically in various shapes and sizes. The size of drusen can be classified as small (<63 µm), intermediate (63–125 µm), or large (> 125 µm). 13 , 14 Drusen are then further described by their appearance as either “hard” or “soft.” If a significant number of drusen coalesce, they are described as “confluent.” 6 , 10
Drusen are composed of lectin binding saccharides, including mannose, glucosamines, galactose, and sialic acid. Lectins are glycoproteins that can bind reversibly to specific saccharide sequences. Multiple proteins are present in drusen including amyloid P component, apolipoprotein E, immunoglobulins, and complement. 15 , 16 In addition, drusen contain molecules associated with both humoral and cellular immunity. 15 , 16 Hard drusen are typically small (< 63 µm) and have well-defined borders (Fig. 14-1). The presence of at least one small druse in the macula is nearly ubiquitous and is not believed to be associated with an increased risk of the development of vision loss from geographic atrophy or the neovascular form of AMD. 3 , 17 , 18 On fluorescein angiography, hard drusen may show hyperfluorescence during the early transit phase as underlying choroidal fluorescence shows through the tiny “window defects” created by the drusen. This fluorescence fades as the choroidal fluorescence fades in the late-phase frames.
Soft drusen are defined as having a greatest linear dimension of more than 63 µm and borders without sharp edges (Fig. 14-2). Soft drusen are believed to be a hallmark of AMD. Development of CNV and geographic atrophy of the RPE appear to be more frequently associated with soft drusen than with hard drusen alone. 14 , 18 , 19 , 20 , 21 , 22 Angiographically, soft drusen may exhibit hypo- or hyperfluorescence. Some investigators have suggested that the intensity of staining may be due to variable polarity of the drusen components. 23 Alternatively, the variability in fluorescence may be due to the degree of pigment epithelial detachment (PED), allowing fluorescein to pool or variability in the degree of overlying RPE depigmentation which allows for more or less transmission of underlying fluorescence.
On spectral domain OCT (SD-OCT), small and intermediate drusen can be seen as discrete areas of RPE elevation with variable reflectivity underlying the RPE (Fig. 14-3a). Drusen can be associated with changes in the overlying outer retina, specifically with evidence of disruption of the outer segment (OS) ellipsoid zone (previously known as the inner segment/outer segment junction zone). 12
CNV appears to be more frequently associated with soft drusen than with hard drusen alone.
Drusen larger than 125 µm have greater clinical significance than smaller drusen and are typically described as soft in appearance. Larger drusen may become confluent and evolve into drusenoid RPE detachments. 6 On OCT, larger drusen or drusenoid pigment epithelium detachments (PEDs) create RPE elevations that are often dome shaped, with a hypo- or medium-reflective material between the RPE and underlying Bruch’s membrane (Fig. 14-3b). 12 On fluorescein angiography, a drusenoid PED fluoresces faintly during the early transit phase and does not progress to bright hyperfluorescence in late frames. Reticulated pigment clumping is frequently seen overlying the detachment.
In AMD, there are significant morphological changes at the level of the RPE in addition to the presence of drusen. There are two variations of sub-RPE deposits, also referred to as basal deposits, which are separated by the RPE basement membrane. “Basal laminar deposits” refer to a layer of abnormal basement membrane material on the inner aspect of the RPE basement membrane. “Basal linear deposits” are formed by a layer of membranous debris on the external aspect of the RPE basement membrane and lie within the inner collagenous zone of Bruch’s membrane. 6
Abnormalities of the Retinal Pigment Epithelium
Focal hyperpigmentation seen clinically corresponds to clumps of pigmented cells at the level of the RPE or pigment that has migrated to the level of the photoreceptor nuclei in the outer retina (Fig. 14-4). 20 , 24 On OCT, pigment clumping and migration can be seen as focal areas of hyperreflectivity. These areas are often seen in the outer nuclear layer or near areas of the RPE overlying drusen. 12 It has been shown that the presence of focal pigment clumping correlates with an increased risk of developing CNV or atrophy of the RPE. 19 , 25 , 26
The presence of focal pigment clumping correlates with an increased risk of developing CNV or geographic atrophy.
Atrophic abnormalities of the RPE may be another manifestation of AMD. With atrophy of the RPE, the overlying outer layers of the retina and the underlying choriocapillaris may also deteriorate and become atrophic. “Geographic atrophy” is an area of retina that is sharply demarcated with loss of the underlying RPE and is usually initially located extrafoveally. This atrophy allows for increased visualization of the underlying choroidal vessels (Fig. 14-5a). 27 The fovea itself is typically spared until late in the course of the disease. Before geographic atrophy involves the fovea, vision is limited by the size of the functioning fovea so that only a portion of a word or an image may be perceived. As geographic atrophy extends into the fovea, patients will experience severe vision loss. 27 Vision loss in geographic atrophy is usually perceived by the patient to be gradual, even when central fixation is involved. 20
OCT is able to successfully identify areas of geographic atrophy visualized as loss of the RPE and the overlying photoreceptors, leading to retinal thinning in the region of geographic atrophy (Fig. 14-5b). There is choroidal hyperreflectivity due to thinning of the overlying retina and RPE. 12 At the edges of the geographic atrophy, the external limiting membrane (ELM) and OS ellipsoid zone may be abnormal and taper off. This abnormality at the borders of retinal thinning suggests that there is photoreceptor loss beyond the margins of the geographic atrophy. OCT offers a mechanism to objectively quantify and follow geographic atrophy over time. 12
Fundus autofluorescence (FAF) is another important imaging modality for identifying and monitoring progression of geographic atrophy. Normally, RPE cells have natural autofluorescence due to the accumulation of lipofuscin. The loss of RPE autofluorescence represents disruption of the RPE–photoreceptor complex, likely due to the death of photoreceptor cells and/or RPE cells. Thus, areas of geographic atrophy appear as discrete hypoautofluorescent lesions. There are various patterns seen on FAF imaging in patients with geographic atrophy, suggesting multiple different phenotypes of geographic atrophy. 28 Comparison of OCT and FAF images in patients with geographic atrophy shows that loss of FAF signal correlates with changes in reflectivity of the choriocapillaris on OCT images (Fig. 14-6). 27 Similar to OCT, FAF can be used to monitor for progression of GA over time.
14.2.2 Differential Diagnosis
Several other maculopathies have features that resemble AMD but have distinctly different etiologies. The need to recognize and differentiate them from AMD is paramount, as prognosis and treatment may vary greatly from that of AMD.
Reticular pseudodrusen form a yellowish interlacing network about 250 µm in diameter that resembles soft, confluent drusen. These pseudodrusen typically present first in the outer macula, but can then develop more peripherally. They are best identified on red-free imaging and do not fluoresce on fluorescein angiography. Clinically, pseudodrusen can appear similar to soft confluent drusen, but can be easily distinguished from drusen on SD-OCT. On SD-OCT, the deposits are seen above the RPE, which is distinctly different from the sub-RPE space in which drusen are found. 29 , 30 When evaluated histologically, reticular pseudodrusen have a similar composition to that of soft drusen. 31 The risk of progression to late AMD is much lower with reticular pseudodrusen than with soft drusen. 6
Basal Laminar Drusen (Cuticular Drusen)
Basal laminar drusen, also known as “cuticular drusen,” are multiple, small, discrete, round, yellow sub-RPE lesions (Fig. 14-7a). The clinical term “basal laminar drusen” should not be confused with the histopathological terms “basal laminar deposits” and “basal linear deposits” (defined in subsection “Drusen”). 24 On OCT, cuticular drusen cause a saw-tooth elevation of the RPE with some disruption of the OS ellipsoid zone and ELM. On fluorescein angiography, cuticular drusen demonstrate early hyperfluorescence, creating a “starry night” appearance that fades in the later phase frames (Fig. 14-7b). Historically, basal laminar drusen were thought to be secondary to thickening of the RPE; however, it is now thought that, histopathologically, they are unlikely different from drusen. 12 , 32 Eyes that have cuticular drusen can form a pseudovitelliform macular detachment. The natural course of these pseudovitelliform lesions may be progressive collapse with the development of marked geographic atrophy or clearing with minimal atrophy. Nevertheless, patients with basal laminar drusen are at risk for developing CNV and any suggestion of CNV on OCT or fluorescein angiography should be considered for treatment.
Pattern dystrophies are a heterogenous group of disorders characterized by bilateral, symmetric pigmentary disturbance of the central macula, with an onset in midlife and usually mild reductions in visual acuity. Inheritance of most of these appears to be autosomal dominant and the prognosis for retention of useful vision into late adulthood appears to be good. Clinically, various patterns of yellow-orange lesions that appear at the level of the RPE have been described. They are usually bilateral and located in the central macula (Fig. 14-8). Occasionally, they can be unilateral or eccentric. The best known of these patterns includes adult-onset foveomacular pigment epithelial dystrophy, butterfly-shaped pigment dystrophy, and reticular pigment dystrophy. Differentiation among them may be difficult because the same dystrophy may have variable phenotypic expression within a family. 33 , 34
Ophthalmoscopic clues that suggest a macular dystrophy rather than AMD are the absence of typical soft drusen as well as a difference in the character of the pigmentation, which is deep to the retina, gray-greenish in appearance, and often reticular with a macular dystrophy. The pigmentation is sometimes best appreciated on transillumination, as direct illumination may reveal only subtle changes. The “pattern” may also be highlighted on fluorescein angiography with blockage of fluorescence by the pigment and hyperfluorescence surrounding the pigment. Clinicopathologic correlation demonstrates a thick layer of periodic acid–Schiff (PAS) positive material between an atrophic RPE and Bruch’s membrane, with large pigmented cells filled with lipofuscin and extracellular pigment superficial to Bruch’s membrane. Like basal laminar drusen, these disorders may be associated with a vitelliform lesion that mimics CNV in its fluorescent pattern on angiography. Again, it is important to differentiate this vitelliform lesion from CNV, given the former does not need treatment, while CNV generally requires treatment. 35
14.2.3 Management and Course
Severe vision loss in AMD results from fovea-involving atrophy or complications of CNV. With the development of anti-VEGF agents, severe vision loss associated with CNV has been significantly reduced especially with early intervention. The main predictor of visual acuity outcome following anti-VEGF treatment is the visual acuity at the time of initiation of treatment; therefore, early detection and prompt initiation of treatment are essential.
Given the infrequency of clinical examination, it is important for patients with dry AMD to perform self-monitoring. Ideally, the patient with non-NVAMD should perform monocular checks of vision to detect metamorphopsia or scotomas on a daily basis using an Amsler grid. Upon detecting new visual symptoms, patients should present promptly for examination.
Development of CNV can often be asymptomatic; therefore, it is recommended that patients with non-NVAMD be examined every 6 to 12 months, perhaps even more often for patients at higher risk for conversion from dry to wet AMD. Some retinal specialists monitor dry AMD patients with OCT images to ensure detection of early exudation. If there is evidence of neovascular abnormalities such as intraretinal fluid, subretinal fluid, or hemorrhage on clinical examination or OCT, then fluorescein angiography may be indicated.
Recently, advances have been made in computerized home monitoring systems to detect development and progression of AMD. The Home Monitoring of Eye (HOME) study recruited participants, aged 55 to 90 years, at significant risk of developing NVAMD. The patients had either bilateral large drusen or large drusen in one eye in addition to advanced AMD in the other eye. This was an unmasked, multicenter, randomized trial comparing the ForeseeHome (Notal Vision) device in addition to standard of care alone. Standard 36 care consists of instructions provided specifically from each investigator for self-monitoring, including use of Amsler grids. The ForeseeHome device uses preferential hyperacuity perimetry and telemonitoring to detect changes in visual function associated with development of CNV due to AMD. 36 The study found that participants at risk of development of CNV benefited from the home monitoring strategy for earlier detection. Those patients in the device arm had a 5-day median time of alerts or symptoms to examination, whereas those patients in the standard care arm had a median time to examination of 7.5 days. During the study period, 82 patients progressed to CNV. Of those, 51 were detected in the device arm, while only 31 were detected in the standard care arm. The device arm also had significantly better mean visual acuity at the time of CNV diagnosis compared to the standard care arm. Early detection increases the likelihood of better visual acuity results with intravitreal anti-VEGF therapy. 36
With the goal of detecting CNV as early as possible, patients with AMD should perform monocular checks of vision for metamorphopsia or scotoma using an Amsler grid on a daily basis.
14.2.4 Risk Factors for Progression to Late Age-Related Macular Degeneration
Several genetic and nongenetic risk factors for AMD have been identified. Some of these risk factors are modifiable, while others are not. Antioxidant and mineral supplementations have been demonstrated to lower the risk of conversion from nonexudative to exudative AMD.
Genetic Risk of Age-Related Macular Degeneration
Several genes have been identified as risk factors for development of AMD. Patients with specific single-nucleotide polymorphisms (SNPs) in the complement factor H region (CFH) and age-related maculopathy susceptibility protein 2 (ARMS2) genes have been shown to have higher association with geographic atrophy and exudative AMD than with early AMD. 37 , 38 In 2005, a strong association between AMD and CFH was identified. 39 This association has been confirmed in multiple ethnic groups around the world. Complement factor H is a serum glycoprotein that serves as a natural inhibitor of complement factor 3 (C3) convertase and blocks the activation of the alternative complement pathway in normal cells. Identification of susceptibility genes associated with the complement pathway confirms the hypothesis that inflammatory pathways play an important role in the pathogenesis of AMD. 9 , 39
In addition to the inflammatory pathway, it is believed that oxidative stress also plays a strong role in AMD progression. Certain genetic mutations that predispose patients to oxidative cellular injury likely have an important role in AMD predisposition. The exact function of ARMS2 is not yet known; however, the protein was originally found in the outer membrane of mitochondria. Mitochondria help control oxidative stress via their role in oxidative phosphorylation in the retina. 9
The Beaver Dam Eye Study found that the estimated population-attributable risk fraction, when at least one CFH risk allele was present, was 9.6 and 53.2% for early and late AMD, respectively. When at least one ARMS2 risk allele was present, that risk was 5.0 and 43.0%, respectively. 37
Besides CFH and ARMS2, a meta-analysis evaluating 14 case–control studies regarding the HTRA1 promoter polymorphism identified an important association with AMD. Because the HRTA1 promoter is close to the ARMS2 gene, it is possible that both together influence AMD susceptibility. 9 Multiple other genes have been studied and are being currently investigated for their association with AMD, but their role is not felt to be as strong as CFH and ARMS2. 39
Genetic testing has now become commercially available for patients with AMD, but it has not been widely adopted into clinical practice. A recent publication has suggested that those patients with certain genetic makeups may have a differential response to AREDS supplementation, and in some cases may in fact have poorer outcomes with standard AREDS formulations. 40 , 41 However, this is controversial 40 , 42 and, to date, no other investigators have found a similar link. As more information becomes available, there may be increased interest in genetic testing of AMD.
Nongenetic Risk Factors for Development of Age-Related Macular Degeneration
A meta-analysis was published in 2010 that evaluated 45 studies looking at 16 risk factors that were felt to be associated with development of advanced AMD. Of the 16 risk factors identified in this systematic review and meta-analysis, age, smoking, cataract surgery, and family history were strongly and consistently associated with late AMD. 43 Most studies have shown no difference between male and female genders for risk of development of AMD; however, at the 10-year follow-up from the original AREDS study, female gender was identified as a risk for progression to NVAMD. 44
The largest modifiable risk factor for AMD is smoking. Jonasson and colleagues showed that of 2,196 patients who smoked and were without evidence of AMD at baseline examination, 14.9% developed AMD. The odds ratio (OR) of developing AMD in a current smoker was 2.07, while it was 1.36 for former smokers. 1 The Beaver Dam Eye Study also confirmed that smoking is associated with an increased risk of progression of minimal to moderate early AMD, as well as from severe early to late AMD. 45 Many other studies have also validated this association. 39 , 46 , 47
Besides smoking, other systemic risk factors including cardiovascular disease, hypertension, and obesity have been associated with increased risk of developing late AMD. Some systemic risk factors have stronger evidence than others. 39 , 48 , 49 Pooled data in a meta-analysis showed a statistically significant increased risk for those with cardiovascular disease and hypertension. 43
There has been a question as to the role of ultraviolet (UV) light exposure in AMD. UV light exposure is thought to possibly contribute to the development of AMD by producing reactive oxygen species and free radicals in the outer retina. The hypothesis of phototoxicity being associated with AMD development is based on experimental and animal studies. Some studies have found an association while others have not. Currently, large epidemiological studies have not shown convincing evidence of the association of UV light and AMD. 39 , 50 , 51 , 52
Some studies have reported evidence that cataract surgery is associated with progression of AMD. In 2003, a study was published that showed that exudative AMD or geographic atrophy developed in 6 to 7% of pseudophakic patients, but in only 0.7% of phakic patients. 53 The AREDS study, however, did not find any clear association between cataract surgery and the risk of progression of AMD. 5
Modifying Risk for Progression of Age-Related Macular Degeneration
Various studies from the 1980s and 1990s suggested a role of antioxidant status and zinc levels with the risk of AMD. 10 , 54 , 55 , 56 Since those studies were published, two large, randomized, controlled trials have shown the benefit of vitamin and mineral supplementation in the prevention of progression to advanced AMD. The first was the Age-Related Eye Disease Study (AREDS), which enrolled 4,575 patients into four different categories based on size and extent of drusen, RPE abnormalities in each eye, the presence of advanced AMD in the fellow eye, and visual acuity. 5
Those with category 1 disease had fewer than 5 small (<63 µm) drusen and had good visual acuity of 20/32 or better in both eyes. Patients with category 2 disease had multiple small drusen, at least one intermediate druse (63–124 µm), and/or pigment changes with vision of 20/32 or better in both eyes. Category 3 eyes had absence of advanced AMD in both eyes, one eye with at least 20/32 vision, at least one large druse (>125 µm), significant intermediate drusen, and/or geographic atrophy outside of the center of the macula. The final group, category 4, had visual acuity of 20/32 or better, and no advanced AMD in the study eye, but the fellow eye had advanced AMD or vision less than 20/32 and AMD changes that accounted for the reduced visual acuity. 5
Patients were randomized to take daily oral tablets with (1) antioxidants (500 mg vitamin C, 400 IU vitamin E, and 15 mg beta carotene), (2) 80 mg zinc oxide and 2 mg cupric oxide, (3) antioxidants plus zinc, or (4) placebo. 5 The AREDS demonstrated that patients with categories 3 and 4 AMD benefited from antioxidant and mineral supplementation. The risk reduction for those patients taking antioxidants and zinc together was 25%. 5 As a result of the study, the AREDS formula was recommended for all patients with category 3 or 4 AMD.
A follow-up study, AREDS2, evaluated the safety and efficacy of removing beta-carotene while adding lutein and zeaxanthin and/or omega-3 long-chain polyunsaturated fatty acid (LCPUFA) supplementation to the original AREDS formulation to reduce the risk of progressing to advanced AMD. For the primary analysis, AREDS2 enrolled 4,203 participants, ages 50 to 85 years, at 82 clinical sites across the United States. Of the 4,203 participants, 3,036 agreed to a secondary randomization, which aimed to evaluate the effect of eliminating beta-carotene and reducing the zinc level from the original AREDS formulation. 57
Both lutein and zeaxanthin are significant components making up macular pigment; therefore, it was felt that these compounds might be of benefit in reducing risk of progression to advanced AMD. While planning the original AREDS, it was believed that lutein may have a potential benefit in reducing risk of progression; however, at the time of the first study, lutein was not commercially available. 58 The reason for removing beta-carotene from the AREDS formula stemmed from a reported increased risk of lung cancer associated with beta-carotene nutritional supplementation in smokers. 14
In the primary analysis, the AREDS2 showed no positive or negative effect of adding lutein, zeaxanthin, omega-3 LCPUFAs, or the combination on the progression to advanced AMD compared to the original formulation. Removing beta-carotene also did not have any statistically significant effect on the progression to advanced AMD when lutein and zeaxanthin were added. 57 With regard to the comparison of 80 to 25 mg of zinc, there was not enough evidence to make a recommendation regarding ideal dose. While the primary analysis did not show an independent benefit of lutein and zeaxanthin, some secondary analyses have shown there might be some positive effect. Therefore, given all of these findings and the increased risk of lung cancer in smokers with beta-carotene supplementation, beta-carotene was removed from the AREDS2 formulation and lutein and zeaxanthin were added.
The AREDS2 formulation includes:
500 mg vitamin C
400 IU vitamin E
10 mg lutein
2 mg zeaxanthin
80 mg zinc
2 mg cupric oxide/copper
(Note: No beta carotene or omega-3 fatty acids.)
The original AREDS formula should not be recommended for patients who smoke.
Grading Risk of Progression of Age-Related Macular Degeneration
As we have gained significant knowledge about risk factors for progression to late AMD, some investigators have tried to create models to help identify individuals at risk for progression. Some of these risk models are more complicated than others. In 2013, a large committee convened to create a new classification system of AMD. They determined risk of progression to late AMD based on two basic characteristics seen on clinical examination: if there was one or more large drusen (>125 µm) and/or any hyperpigmented or hypopigmented abnormalities associated with drusen. For each risk factor in each eye, a risk score of 1 is associated. There is a maximum total score of 4 for a person, 2 per eye. The 5-year risk of advanced AMD is 100-fold larger between those patients with a score of 0 and those with a score of 4. The risk of progression to advanced AMD with a score of 1 is 12%, 25% for those between 2 and 3, and 50% for those between 3 and 4. These percentages are then further modified based on other systemic and external risk factors for progression of AMD. For example, the risk for progression varies based on smoking status. 44 , 59 , 60
14.3 Neovascular Age-Related Macular Degeneration
Characterized by the growth of fibrovascular tissue from the choroidal circulation, NVAMD is the leading cause of severe vision loss in people older than 65 years in the United States and Western world. Although it accounts for only 8% of patients with late stage of AMD, it is responsible for 85% of the severe visual loss caused by AMD. 4 The vision loss is caused by the resultant subretinal fibrosis from the CNV and is irreversible.
14.3.1 Clinical Features
CNV should be suspected in anyone complaining of metamorphopsia or scotoma, especially if the person is older than 65 years and is known to have non-NVAMD. 61 Slit-lamp biomicroscopy may reveal retinal elevation from underlying fluid or blood, pigment epithelial elevation from sub-RPE fluid, blood or fibrovascular tissue, intraretinal or subretinal lipid, or cystic edema of the sensory retina overlying the CNV membrane (Fig. 14-9a). Other causes of subretinal hemorrhage, such as macroaneurysms, lacquer cracks in pathologic myopia, or choroidal tumors, should be ruled out. The fundus may appear normal in spite of the presence of CNV, especially if the only sign present is subretinal fluid, which may be very subtle by ophthalmoscopy.
Fluorescein angiography is the gold standard in the diagnosis of exudative AMD to identify and characterize CNV. Early-, mid-, and late-phase frames of the macula enable determination of the CNV boundaries (well demarcated or poorly demarcated), pattern (classic or occult), and location (subfoveal or extrafoveal).
CNV may have a variety of appearances on fluorescein angiography. The basic patterns recognized are classic and occult (Fig. 14-10). While these classifications are no longer used as frequently as they were in the past, prior to the use of anti-VEGF agents, they can provide some important prognostic information. Classic CNV is apparent on the early transit-phase frames as a bright, well-demarcated area of hyperfluorescence with significant leakage apparent in the later phase frames as the dye pools in the subretinal space and obscures the boundary of this area of hyperfluorescence. Lacy vessels, sometimes visualized in the early-phase frames, are not required for a lesion to be considered classic. In fact, this appearance is present in only a minority of the classic CNV lesions associated with AMD and may be seen in a lesion judged to have a fluorescent pattern of occult CNV, as both classic and occult CNV consist of new blood vessels.
The term occult CNV refers to two types of patterns seen in eyes with AMD-related CNV. The first pattern, termed a fibrovascular pigment epithelial detachment (FVPED), is best appreciated with stereoscopic views at approximately 1 to 2 minutes after dye injection; it appears as an irregular elevation of the RPE stippled with hyperfluorescent dots. Staining or leakage is evident in the late-phase frames as fluorescein collects within the fibrous tissue or pools in the subretinal space overlying the FVPED. The exact boundaries of FVPEDs can be determined only when fluorescence sharply outlines the elevated RPE. The amount of elevation is dependent on the quality of the stereophotograph and the thickness of the fibrovascular tissue. Unfortunately, the boundary of a FVPED is usually not well demarcated. The second pattern, late leakage of an undetermined source, refers to late choroidal-based leakage in which there is no clearly identifiable classic CNV or FVPED in the early or midphase of the angiogram to account for an area of leakage in the late phase. Often, this pattern of occult CNV can appear as speckled hyperfluorescence overlying pooling of dye in the subretinal space.
The terminology of classic and occult CNV is currently less commonly used due to the transition away from laser treatment and toward using intravitreal anti-VEGF therapy. With the increased use of OCT, a new classification scheme has been created. SD-OCT is able to identify CNV and its associated intraretinal fluid, subretinal fluid, and PEDs (Fig. 14-9b). 12 Type 1 CNV originates in the choroid and then breaks through Bruch’s membrane and extends under the RPE. If the neovascularization comes from the choroidal circulation and then extends through the RPE and into the subretinal space, then it is referred to as type 2 CNV. Often, CNV is a combination of both type 1 and type 2 disease. If the disease is predominantly type 1, it is referred to as minimally classic. If it is predominantly type 2, then it is predominantly classic. 32 , 62 Type 3 neovascularization refers to retinal angiomatous proliferation (RAP). This is characterized by retinal neovascularization that emanates from the deep capillary plexus of the retina. There are multiple stages of progression of RAP, with the final stage including choroidal and retinal neovascularization that consolidates to become a retinal–choroidal anastomosis. This is associated with leakage within and below the retina and involves development of a PED. 63 , 64
With the increased use of SD-OCT imaging, new features of AMD have been discovered that had not been identified previously. The presence of what have now been named outer retinal tubulations (ORTs) are seen in some patients with many diseases associated with retinal injury and photoreceptor disruption, but most commonly with exudative AMD. 65 ORTs represent photoreceptor rearrangement and degeneration. On SD-OCT, ORTs appear as round, hyporeflective structures that are surrounded by a hyperreflective lining, usually overlying areas of pigment epithelial alteration or subretinal fibrosis. The hyperreflective border of the tubulation is thought to be the inner segment–outer segment junction of the photoreceptors. It is important to distinguish ORTs from cystoid macular edema because of the implications for treatment with anti-VEGF therapy, as ORTs do not represent active exudation. 65 , 66
The end stage of exudative AMD is the development of a fibrotic disciform scar (Fig. 14-11). The intraretinal and subretinal fluid, as well as hemorrhage, associated with CNV can lead to disorganization of the retina. There is extensive loss of RPE and photoreceptors with development of a disciform scar. When these scars involve the fovea, there is significant vision loss. 67 Although a disciform scar represents end-stage disease, there can still be reactivation of CNV leading to further hemorrhage; therefore, these patients should be followed up at less frequent but regular intervals. 68
Pigment Epithelial Detachments
Various lesions in AMD will result in elevation of the pigment epithelial layer of the retina. A fibrovascular PED is a type of occult CNV described earlier (Fig. 14-12). A serous detachment of the RPE (Fig. 14-13) is distinguishable angiographically from a fibrovascular PED by the appearance of well-circumscribed, bright, uniform hyperfluorescence in the early phase with persistence of bright fluorescence in the later phases, with little, if any, leakage into the subretinal space. Although the clinical course is variable, most of these patients will show evidence of CNV within 1 to 2 years. Some serous PEDs manifest a notch at the edge that is believed to be a sign of occult CNV. 69
On OCT, PEDs are broad elevations of the RPE above Bruch’s membrane. Fibrovascular PEDs are associated with variable amounts of serous exudation and/or hemorrhage. Owing to this exudation, the slope of the PED may vary depending on its fluid content. On OCT, avascular serous PEDs appear as dome-shaped elevations of the RPE. Underlying the RPE is a homogenously hyporeflective space. Vascularized serous PEDs likely occur with growth of a CNV lesion in the sub-RPE space that has significant associated exudation. This exudation creates a serous fluid compartment that looks similar to that of avascular PEDs; however, some have small collections of solid material (fibrovascular proliferation) adherent to the outer surface of the RPE.
Hemorrhagic PEDs form when there is hemorrhage in the sub-RPE space from neovascularization. The hemorrhage obscures the view to Bruch’s membrane. 12 A hemorrhagic PED decreases the ability to detect CNV beneath the blood on fluorescein angiography as well. The appearance of a hemorrhagic PED sometimes may mimic a choroidal melanoma; however, the lack of low internal reflectivity on ultrasound examination will assist in this distinction.
A tear of the RPE can occur spontaneously in an eye with a serous PED or following treatment. 70 Studies have shown that the incidence of RPE tears after anti-VEGF therapy is similar to spontaneous RPE tears associated with the natural history of the disease. 71 On OCT, the discontinuity of the RPE can often be seen. If the torn area has no underlying fibrovascular tissue, one will see early bright, sharply demarcated hyperfluorescence on fluorescein angiography in the area denuded of RPE. An adjacent hypofluorescent area will be apparent where the scrolled RPE is now blocking underlying fluorescence (Fig. 14-14). If fibrovascular tissue persists in the area of torn RPE, the fluorescence may not be as bright in the denuded area.
With RPE tears, visual acuity may fall precipitously. Some studies have shown that the majority of patients will have continued vision loss with longer follow-up. The major factor that determines visual prognosis is whether or not there is loss of RPE in the foveal center. Unfortunately, RPE tears are often associated with progressive fibrovascular scarring, which also contributes to vision loss. 71 Thus, when using anti-VEGF agents, if a RPE rip occurs and there is persistent activity of the choroidal neovascular membranes, most still recommend continued treatment with anti-VEGF therapy to help prevent fibrovascular scarring. While there may be concern for enlargement of the RPE tear with further consolidation of the CNV with treatment, studies have shown benefit to treatment. 72 Tears of the RPE where visual acuity was preserved even though subfoveal RPE was absent have been reported, but these cases did not have evidence of associated CNV. 9 , 33 Coco and colleagues described a series of patients with RPE tears that did not involve the foveal center. They found that continued anti-VEGF therapy led to a statistically significant improvement in vision. 72 Studies have been published that show some patients regain vision with repopulation of the RPE after the tear. 73