Age-related macular degeneration (AMD) is now one of the major causes of central vision loss.

It involves the macular area and when it progresses and destroys the central fovea, quality of life is seriously compromised.

The ability to read, drive, recognize faces or watch television is impaired or lost.

It is a disease associated with aging and progressive tissue degeneration.

Most of these senior citizens had anticipated the opportunity to enjoy life at leisure doing their preferred activities and find profound limitations and are deeply disappointed in their expectations.

This was the dismal state of affairs until a few years ago when the introduction of new therapeutic agents changed dramatically the expected outcome.

The most important development has been the clinical demonstration that agents inhibiting vascular endothelial growth factor and, therefore, the formation and development of new vessels not only preserve visual acuity but also improve visual function.

This find was a true revolution in ophthalmology.

The retina like the central nervous system, particularly in old age, was not considered capable of regeneration and, therefore, new therapies were expected to offer only stabilization of the disease.

When laser photocoagulation was shown to be effective for the treatment of diabetic retinopathy this was a breakthrough but it offered only stabilization of the disease and maintenance of visual acuity present at time of treatment.

Now, new available treatments offer real improvement in visual acuity, implicating improved visual function and some degree of recovery of the retinal neuronal network.

AMD is, therefore, an active scientific area with new information becoming available almost everyday on the pathophysiology, clinical phenotypes, markers of disease progression, new treatments and novel treatment regimens.

It is a complex multifactorial disease where aging associates with genetic factors and inflammatory responses to local cell injury.

This means that the treatment of AMD must also address the various factors involved in disease progression and, most likely, will involve a combination of therapies after clear identification of the different AMD phenotypes.

This book reviews the new concepts on AMD and is, therefore, timely.

It is a concerted effort of Portuguese ophthalmologists which together with a few other European experts in the field cover the subject paying particular attention to daily practice management of AMD.

I am particularly happy to see this collaborative effort coming from my country, Portugal.

It shows that Portuguese ophthalmology is at the forefront of European and International ophthalmology.

It shows also that there is in Portugal a real spirit of collaboration between colleagues working together to improve the vision of their patients.

1. Introduction

Although age-related macular degeneration (AMD) is the third cause of blindness worldwide, and the first in industrialized countries^(1), epidemiological data on this disease remain scarce and partial. The very first studies were published in the 1980’s. Since then, a number of population-based studies have been conducted, first in Caucasian populations of the United States and other industrialized countries (Australia, Europe). Studies have more recently been extended to other ethnical subgroups of industrialized countries (African Americans, Latinos) and to other parts of the world (India, China, and Japan).

 

2. Prevalence of late AMD in Caucasians from industrialized countries

The prevalence of late AMD in the main epidemiological studies performed in Caucasians of industrialized countries are presented in Table 1.

Author	Study Years study conducted	Country Number of subjects, age	Prevalence of late AMD(%)
United States
Friedman, Baltimore Eye Study(28)	1985-1988, USA	N=2518, age >= 40 years	1.23
Klein, Beaver Dam Eye Study(29)	1988-1990, USA	N=4752, age >= 40 years	1.64
Bressler, Salisbury Eye Evaluation Project(30)	1993-1995, USA	N=1773, age >= 65 years	Exudative AMD : 1.7 Geographic atrophy: 1
Klein, Atherosclerosis Risk in Communities Study(31)	1993-1995, USA	N=8984 , age 48-72 years (one eye only)	0.
Klein, Cardiovascular Health Study(32)	1997-1998, USA	N=1998, age 69-97 years (one eye only)	1.3
Klein, Multi-Ethnic Study of Atherosclerosis(10)	2000-2002, US	N=2315, age 45-84 years	0.6
Australia
Mitchell, Blue Mountains Eye Study(33)	1992-1994, Australia	N=3632, age >= 50	2.06
Van Newkirk, Melbourne VIP(34)	1992-1996, Australia	N=4345, age >= 40 years	0.68
Europe
Vingerling, Rotterdam Study(35)	1990-1993, Netherlands	N=6774, age >= 55 years	1.65
Vingerling, Rotterdam Study(35)	1995-1997, France	N=2196, age >= 60 years	1.9
Jonasson, Reykjavik Eye Study(5)	1996, Iceland	N=1022, age >= 50 years	3.5
Topouzis, Thessaloniki Eye Study(37)	2000, Greece	N=1022, age >= 50 years	2.5
Bjornsson, Oslo Macular Study(6)	2002, Norway	N=459, age > 50 years	2.8
Augood, Eureye(4)	2001-2002,	N=4753, age >= 65 years	3.3
	7 european countries

Table 1. Prevalence of late AMD in Caucasians from industrialized countries

 

The included studies were restricted to those having classified AMD from retinal photographs and used the international classification⁽²⁾, in order to make comparisons between studies easier. In these studies, late AMD is defined by the presence of neovascular AMD and/or geographic atrophy. In the United States, the prevalence of late AMD ranged from 0.2% to 1.6% according to studies. Since the prevalence of late AMD increases sharply with age, most of the differences between studies were due to differences in the age distribution. For instance, the lowest prevalence rate (0.2%) was observed for the Atherosclerosis Risk in Communities (ARIC) Study. In this study, the age range of participants was 48-72 years, thus excluding the oldest subjects, in which the prevalence of late AMD is the highest. This study also performed photographs in only one eye of each participant, which may have led to undetected unilateral cases and thus underestimation of the prevalence of late AMD. This is also the case for the Cardiovascular Health Study, which also found relatively low prevalence (1.3% in subjects aged 69 to 97 years). Prevalence rates observed in Caucasians from Australia and Europe appear relatively similar to those observed in the United States (Table 1), when taking into account differing age distributions in the different studies. In European studies, the prevalence of late AMD ranged from 1.65% to 3.5%.

In order to better compare the prevalence rates among studies, age-specific prevalence rates need to be estimated. This has been done in a meta-analysis performed in 2004 by Friedman et al⁽³⁾. The authors concluded that prevalence rates were not different among Caucasian populations of industrialized countries, including the United States, Australia and the Netherlands. Similarly, in the EUREYE Study, which included 7 European countries (Norway, Estonia, Ireland, France, Italy, Greece, Spain), no significant differences were found among the participating countries^(4). Therefore, the prevalence of late AMD appears to be similar in the Caucasian populations from the United States, Australia and European countries, despite major geographical and lifestyle differences. Thus, the meta-analysis from Friedman et al. probably constitutes the most reliable available estimate of prevalence rates for these countries, since it bears on studies gathering altogether more than 25000 subjects⁽³⁾. In this meta-analysis, the prevalence rates increase sharply with age (Figure 1), from less than 0.5% in subjects 50 to 60 years, to 12% and 16% in men and women aged 80 years or more, respectively. While men tend to have higher prevalence rates than women in the younger age groups (less than 80 years), late AMD is more frequent in women than in men in the oldest age group (80 years or more). This may be due to the higher mortality in ageing men, which leads to a higher selection of the oldest subjects. Finally, in the late AMD groups, two clinical entities exist: neovascular AMD and geographic atrophy. In most studies, neovascular cases represent about 55% of all late AMD cases⁽³⁾. Nordic European countries (in particular Iceland and Norway) may be an exception, since the few studies performed in these countries suggest a higher rate of geographic atrophy^(5,6).

 

Figure 1. Prevalence of late AMD according to age and gender in Caucasians from industrialized countries (adapted from the meta-analysis by Friedman et al.(3))

 

3. Prevalence of late AMD in other ethnic groups

As shown in Table 2, the prevalence of late AMD is much lower in African-Americans than in Caucasians.

Author	Study Years study conducted	Country Number of subjects, age	Prevalence of late AMD(%)
African-Americans
Friedman, Baltimore Eye Study(28)	1985-1988, USA	N=1843, age >= 40 years	0.22
Bressler, Salisbury Eye Evaluation Project(30)	1993-1995, USA	N=666, age >= 65 years	Exudative AMD : 1.1 Geographic atrophy : 0.3
Klein, Atherosclerosis Risk in Communities	1993-1995, USA	N= 2548 , age 48-72 years (one eye only)	0.04
Study(31)
Klein, Cardiovascular Health Study(32)	1997-1998, USA	N=363, age 69-97 years (one eye only)	0.3
Hispanics
San Luis Vally(17)	1983, USA	N=3995, age 43-74 years	0.09
Proyecto VER(15)	1997-1999, USA	N=2776, age >= 50 years	0.5
Los Angeles Latino Eye Study(16)	2000-2003, USA	N=5875, age >= 40 years	0.43
Klein,Multi-Ethnic Study of Atherosclerosis(10)	2000-2002, USA	N=1280, age 45-84 years	0.2
Chinese
Klein, Multi-Ethnic Study of Atherosclerosis(10)	Klein, Multi-Ethnic Study of Atherosclerosis(10)	N=699, age 45-84 years	1.0

Table 2. Prevalence of late AMD in other ethnic groups from the United States

Most of the cited studies were multi-ethnic and performed direct comparisons between African-Americans and Caucasians. Statistical power was generally low for these comparisons, but these studies globally suggest that the prevalence of late AMD is at least 2-fold lower in African-Americans. In the Barbados Eye Study, bearing on subjects of African origin from the Barbados Island, the prevalence rate of late AMD (0.57%) was similar to those observed in African-Americans (Table 3)⁽⁷⁾.

Author	Study Years study conducted	Country Number of subjects, age	Prevalence of late AMD(%)
Barbados
Schachat, Barbados Eye Study(7)	11987-1992, Barbados	N=3444, age 40-84 years	0.57
Japan
Oshima, Hisayama Study(8)	1998, Japan	N=889, age >= 50 years	0.89
Kawasaki, Funagata Study(9)	200-2002, Japan	N=1625, age >= 35 years	0.5
		N=1037, age >= 55 years	0.8
India
Krishnaiah, Andrah Pradesh Study(38)	1996-2000, India	N=3723, age >= 40 years	1.9
Nirmalan, Aravind Eye Study(18)	1995-1997, India	N=4197, age >= 40 years	0.6
Krishnan, INDEYE(19)	2005-2007, India	N=4266, age >= 60 years	1.2
China
Li, Beijing Eye Study(11)	2001, China	N=4376, age >= 40 years	0.2
Taïwan
Chen, Shihpay Study(12)	1999-2000, Taïwan	N=1105, age >= 65 years	1.9
Singapore
Kawasaki, Singapore Malay Eye Study(13)	2004-2006, Singapore	N=3265, age 40-80 years	0.7
Greenland
Andersen, Greenland Inuit Eye Study(14)	2002-2003, Greenland	N=642, age >= 60 years	9.1

Table 3. Prevalence of late AMD in other countries

 

Therefore, people of African origin appear to be at much lower risk of late AMD than Caucasians, although no data are available from African countries. American Hispanics also appear to be at lower risk for late AMD than Caucasian, with prevalence rates ranging from 0.09% to 0.5% among studies (Table 2). The reasons for the lower prevalence rates in African-American and Hispanics are unclear. No data are available on minorities of other industrialized countries, in particular in Europe. This constitutes a limitation to the estimation of the prevalence of AMD in European countries, which have different types of ethnic groups than in the United States (populations originating from North Africa, Middle-east countries, India…). Given the major differences observed in ethnic groups in the United States, such epidemiological data in European minorities are warranted.

Other epidemiological studies have mainly been conducted in Asia (Japan, India and China). In contrast to what was originally thought, late AMD is not rare in these Asian populations, as shown in Table 3. In two Japanese studies^(8,9), the prevalence rates in Japanese men were similar to those observed in Caucasian men from industrialized countries, while late AMD appeared rare in Japanese women. This gender-effect, which is not observed in industrialized countries, may be due to gender-related smoking habits in Japanese. Indeed, smoking is a major risk factor for AMD, in all studied populations. In these two studies, smoking was very frequent in Japanese men, while rare in Japanese women.

In three epidemiological studies performed in India, the prevalence rates of late AMD ranged from 0.6% to 1.9% (Table 3). Below the age of 80 years, the prevalence rates were rather similar to those observed in industrialized countries. Above 80 years, the figures are mostly unreliable because of the very high prevalence of unoperated cataracts in this population, leading to a very high percentage of ungradable retinal photographs. The global prevalence rates were therefore probably underestimated.

Three studies have been performed in subjects of Chinese origin: one in the United States (Table 2)⁽¹⁰⁾, one in mainland Beijing⁽¹¹⁾and one in Taïwan⁽¹²⁾. While prevalence rates in Chinese from the United States and Taïwan seemed similar to those in Caucasians, the prevalence rates in Chinese from Mainland China were much lower.
The reasons for these differences are unclear, and may be related to gene-environment interactions, since Chinese from the United States and Taïwan tend to live a more industrialized lifestyle. More epidemiological data are needed to confirm and understand the reasons for these differences.

Finally, in a study from Singapore, the prevalence of late AMD was similar to that in Caucasians⁽¹³⁾. Globally, the prevalence of late AMD appears high in Asians, with the only exception of Chinese from Mainland China, which remains to be confirmed.

A study in Inuits from Greenland shows strikingly high rates of late AMD, with particularly high rates of neovascular AMD⁽¹⁴⁾. This is in contrast from the observations of European Nordic countries, which found higher rates of geographic atrophy^(5,6) and also in contrast with original observations in this population, which also found higher rates of geographic atrophy in Inuits. The reasons for these differences are unclear.

 

4. Prevalence of early AMD

Late AMD is preceded by early, usually asymptomatic, retinal abnormalities. The long-term cohort studies have helped to characterize the lesions that are the most predictive of incident late AMD. While large soft drusen (>125 microns) and pigmentary abnormalities clearly are the hallmarks of early AMD, there is no consensus on the precise definition of early AMD, and several classification systems coexist. This makes it difficult to assess and compare the prevalence of early AMD among geographical areas and ethnic groups.

Despite these difficulties, it appears that early AMD is frequent in the elderly. For instance, in the meta-analysis by Friedman et al, the prevalence of large drusen increased from about 1.5% in Caucasians aged 40-49 years, to more than 25% in those aged 80 years or more⁽³⁾. Similar rates were observed in the EUREYE Study, performed in Europe⁽⁴⁾.

Consistently with what was observed for late AMD, early AMD appears less frequent in African-Americans than in Caucasian^(3,10). By contrast, while late AMD is also less frequent in Hispanic-Americans, the prevalence of early abnormalities appears similar, or even higher than in Caucasians^(10,15-17). The reasons for these differences are unclear, and suggest that different factors may be implicated in the etiology of early and late AMD. In most studies performed in Asians, the prevalence of early AMD appears similar to that observed in Caucasians, consistently to what was found for late AMD^{(8-10,12,13,18,19).}

 

5. Prevalence of late AMD in Portugal

To our knowledge, there are no published data on the prevalence of AMD in Portugal. What can therefore be extrapolated from available data in other countries? As explained above, the prevalence of late AMD appears to be similar in Caucasian populations from industrialized countries. Its best estimate therefore derives from the meta-analysis by Friedman et al⁽³⁾. We, thus, applied the age and gender-specific prevalence rates of the meta-analysis to the demographic data from Portugal, leading to an estimation of about 84 000 cases of late AMD in Portugal (Table 4).

Population of Portugal		AMD prevalence (%) *	Number with AMD
Men
50-54	309 484	0,34	1 052
55-59	268 899	0,41	1 102
60-64	256 179	0,63	1 614
65-69	244 230	1,08	2 638
70-74	196 615	1,98	3 893
75-79	143 439	3,97	5 695
80+	123 934	11,90	14 748
Total men	1 542 780	1,99	30 742
Women
50-54	333 032	0,20	666
55-59	302 553	0,22	666
60-64	294 737	0,35	1 032
65-69	293 935	0,70	2 058
70-74	257 347	1,52	3 912
75-79	204 627	3,44	7 039
80+	229 366	16,39	37 593
Total women	1 915 597	2,76	52 965
Total	3 458 377	2,42	83 707

Table 4. Prevalence of late AMD in Portugal according to age and gender

 

A major limitation to this approach is that it does not take into account the prevalence rates of AMD in non Caucasians living in Portugal. As shown above, the prevalence of late AMD appear lower in subjects of African origin. However, no data are available on the prevalence of AMD in subjects of African, or other ethnic origins (for instance North Africa and Asia) in European populations. It appears difficult to assess the prevalence of AMD while taking into account the multi-ethnic structure of the Portuguese population. However, these ethnic subgroups represent small minorities in Portugal, so that this may only marginally affect the global prevalence rates.

According to these estimations, about two thirds of cases are women, and two thirds of cases are aged 80 years or more. Since it represents about 55% of late cases, neovascular AMD probably affects about 46 000 subjects.

 

>> References

1. Introduction

The epidemiological studies conducted in the past 25 years have helped identifying major modifiable risk factors for AMD. In particular, smoking and nutrition appear ever more important in determining the occurrence of AMD, and may, in the future, lead to prevention strategies.

 

2. Smoking

Smoking is the best characterized risk factor for AMD⁽¹⁾. The initial observations performed in Caucasian populations from Western countries⁽²⁾, are now being confirmed in other ethnic groups, such as African-Americans⁽³⁾, Latino-American⁽⁴⁾, or Asian populations^(5-7). In most studies, the risk for late AMD was multiplied by 2.5 to 4.5 in current smokers. In addition, the dose-response relationship was explored in some studies^(6,8-12). Most of these studies found that the risk for AMD increased with increasing number of cigarettes smoked per day, and, even more, with number of pack-years smoked, which is an indicator of cumulative smoking over the lifetime (mean number of packs smoked/day x duration of smoking (years). Moreover, the risk for AMD appeared to decrease with time from cessation of smoking.  Former smokers generally demonstrated a lower risk for AMD than current smokers. Several studies have shown that the risk for AMD in subjects having ceased smoking for more than 20 years was similar to the risk in never smokers^(8-10,13). One study suggested that passive smoking is also associated with an increased risk for AMD⁽⁹⁾, while this association did not reach statistical significance in another study⁽¹⁴⁾. Finally, smoking appeared to be related to similar risks for both types of late AMD (geographic atrophy and neovascular AMD)^(9,13,15-16). By contrast, associations with early AMD were weaker in the vast majority of published studies, and often not statistically significant^{(8-9,13,15,17-18)}. 

Overall, the strength of the association (about 3-fold increased risk in current smokers), its consistency across different populations, the observation of a clear dose-response relationship in most studies, and the decrease of the risk with stopping smoking are all strong arguments in favour of a causal role of tobacco smoking in the aetiology of late AMD. 

The exact mechanisms by which smoking increases the risk for AMD are unclear, and probably multiple, including oxidative stress, inflammation and decreased macular pigment. Finally, recent studies gave important insights on the joint effects of smoking and genetic polymorphisms, showing that the risk for AMD is particularly high in smokers bearing at-risk polymorphisms in the CFH or LOC387715 genes^(19-21).

Other vascular risk factors, such as systemic hypertension, obesity, diabetes, plasma lipids or alcohol drinking may be associated with an increased risk of AMD, but epidemiological studies have been inconsistent in this field⁽²²⁾. At the time being, they remain putative, but not clearly identified risk factors for AMD. 

 

Nutritional factors

More recently, epidemiological studies have focused on the potential association of AMD with nutritional factors. Mainly three types of nutritional factors have been investigated for their potential protection against eye ageing: antioxidants (mainly vitamins C and E, zinc), the carotenoids lutein and zeaxanthin and omega 3 polyunsaturated fatty acids (PUFA). 

The retina is particularly susceptible to oxidative stress because of the high level of in-site reactive oxygen species production, due in particular to light exposure and high metabolic activity⁽²³⁾.

Epidemiological studies are mostly in favour of a protective role of antioxidants for AMD⁽²⁴⁾.

Moreover, the Age-Related Eye Diseases Study (AREDS), a randomized clinical trial performed in the United States and including on almost 5000 subjects supplemented for five years, showed a significant 25% reduction of the incidence of late AMD with supplementation in antioxidants and zinc, by comparison with placebo⁽²⁵⁾. In this field, data from the United States should be extrapolated to European populations with caution. Indeed, vitamin supplements are widely used in the American population, while this is rarely the case in Europe. For instance, two thirds of the AREDS participants used vitamin supplements, in addition to the supplementation tested in the study⁽²⁵⁾. Plasma vitamin C concentration at baseline in the AREDS (before the initiation of the study supplementation) was 62 micromol/l⁽²⁵⁾, whereas it was 31.6 micromol/l in men and 40.5 micromol/l in women of the Pathologies Oculaires Liées à l’Age (POLA) Study, performed in the South of France⁽²⁶⁾. Similarly, in the EUREYE Study, plasma vitamin C concentrations ranged from 35.5 micromol/l to 48.4 micromol/l in seven European countries⁽²⁷⁾. Therefore, antioxidant intake is much lower in European populations than in the United States, with part of European populations being at risk of clinical deficiency in these vitamins. Two European studies suggested that the benefit to be expected from increased antioxidant intake may be more important in our populations with low antioxidant intake. Indeed, in the French POLA Study, we observed an 80% decreased risk for late AMD in the subjects with higher plasma vitamin E, by comparison to those with lower concentrations^(28), a much stronger effect than the 25% reduction in risk observed in the AREDS Study. Moreover, we observed a 25% reduction in risk for early ARM, whereas the AREDS Study showed no benefit of antioxidant supplements for early ARM. Similarly, results from the Rotterdam Study showed a decreased risk for early ARM in subjects with high dietary intake of vitamin E or zinc, by comparison with those with low intake⁽²⁹⁾. A European supplementation study would be needed to better assess the benefit of antioxidant supplementation in European populations.

A more recent research domain evaluated the role of two carotenoids, lutein and zeaxanthin, in the protection of the retina and the lens. These carotenoids accumulate in the macula, where they are known as the macular pigment⁽³⁰⁾. Besides their antioxidant properties, they probably act as a filter against the phototoxic effects of blue light⁽³⁰⁾. To date, five epidemiological studies have assessed the associations of the risk of AMD with plasma concentration of lutein and zeaxanthin^(31-35). As shown in Fig. 1, all five studies showed a decreased risk for AMD in subjects with high plasma concentrations of lutein and zeaxanthin, although the association was statistically significant only in 2 studies^(32-35). With regard to dietary intake, four prospective population-based studies were published^(29,36-38).

Figure 1. Association of the risk of AMD with plasma levels of lutein and zeaxanthin in cross-sectional and case-control studies (odds-ratios with 95% confidence interval)
OR below 1 suggest a protective role and OR greater than 1 suggest a deleterious role. References of the cited studies: Beaver Dam⁽³¹⁾; EDCC⁽³²⁾; NHANES III⁽³³⁾; Gale et al⁽³⁴⁾; POLA⁽³⁵⁾

 

These studies assessed the risk for developing AMD (in subjects initially free of AMD), according to their dietary intake of lutein and zeaxanthin. As shown in Fig. 2, the results for these dietary studies are less clear than for those on plasma measurements.

Figure 2. Associations of the risk for AMD with dietary lutein and zeaxanthin, in published epidemiological prospective studies. References of the cited studies: Beaver Dam⁽³⁶⁾; Health Professionals⁽³⁷⁾; Rotterdam⁽²⁹⁾; Blue Mountains⁽³⁸⁾.

 

Only one study found a significantly reduced risk for AMD in subjects with high dietary intake of lutein and zeaxanthin(38). However, dietary assessment methods rely on the subjects’ memory and perceptions and face the difficulties of the extreme day-to-day variability of human diet, the bias in reporting due to social standards and nutritional recommendations and the estimations of nutritional contents of food items. Biomarkers have the advantages of being objective, and of taking into account individual variations in bioavailability and metabolism. For instance, smoking and obesity are known to decrease the bioavailability of carotenoids^(39-40). Despite normal dietary intake in lutein and zeaxanthin, subjects may be at higher risk for AMD because of decreased bioavailability, associated with lower plasma concentrations of these components. However, currently available studies including plasma measurements are cross-sectional or case-control studies, where the plasma measurements were performed in subjects already affected by the disease. The stronger associations found in these studies may therefore be explained by reverse causality (i.e. plasma carotenoids were lower because of change of dietary habits in subjects with AMD, for instance).

Globally, the few available epidemiological studies suggest a protective role of lutein and zeaxanthin in AMD, but other studies are needed in this field, in particular larger, prospective studies including dietary and plasma measurements.

Finally, omega 3 PUFA include a precursor (alpha-linolenic acid (ALA)), and three long-chain derivatives (eicosapentaenoic acid (EPA), docosapentaenoic acid (DPA) and docosahexaenoic acid (DHA)). ALA is an essential nutrient, since humans cannot synthesize it de novo and therefore rely on diet as its sole source. Synthesis of the long-chain derivatives is very limited in humans⁽⁴¹⁾, who must therefore rely on their dietary supply, mainly by fish and seafood. Long-chain omega-3 PUFA have important structural and protective functions in the retina. DHA is a major component of the photoreceptors⁽⁴²⁾. They also have protective functions, including the systemic anti-inflammatory, anti-angiogenic and anti-apoptotic functions and specific actions in the retina such as increase in lysosomial acid lipase, leading to increased lipid degradation in the retinal pigment epithelium⁽⁴²⁾. As shown in Fig. 3 and Fig. 4, seven case-control^(43-44) or cross-sectional studies^(45-49) and five prospective studies^(50-54) assessed the associations of dietary intake of long-chain omega 3 PUFA or fish with AMD. In spite of differences in populations, methods and types of studies, all studies showed a reduced risk for AMD in subjects with high dietary intake of long chain omega 3 PUFA or fish, although some of these associations did not reach statistical significance.

Figure 3. Association of the risk for AMD with dietary long chain omega 3 or fish (cross-sectional and case-control studies). References of the cited studies: EDCC⁽⁴³⁾; Twin Study⁽⁴⁵⁾; AREDS⁽⁴⁴⁾; Beaver Dam⁽⁴⁶⁾; NHANES⁽⁴⁷⁾; POLA(⁴⁸⁾; EUREYE⁽⁴⁹⁾

 

Figure 4. Association of the risk for AMD with dietary long chain omega 3 or fish (prospective studies). References of the cited studies: Blue Mountains⁽⁵¹⁾; Nurses and Physicians’Health Study⁽⁵⁰⁾; Reykjavik Study⁽⁵²⁾; Seddon et al.⁽⁵³⁾; AREDS⁽⁵⁴⁾

 

Most of these studies were grouped in a meta-analysis performed in 2008, which concluded at a 38% reduction in risk for AMD in subjects with high dietary long-chain omega 3 PUFA⁽⁵⁵⁾. By contrast, two recent studies found an increased risk for ARM in subjects with high omega 3 PUFA(56-57). However, in these studies, only total omega 3 PUFA intake was studied, including ALA and long-chain omega 3 PUFA, while most studies found a reduced risk for ARM only with high long chain omega 3 PUFA, in accordance with the scientific rationale. As stated in one of these studies⁽⁵⁷⁾, the main source for ALA is vegetable oil, which is also the main source of omega 6 PUFA, and was found to increase the risk for ARM in some studies(^43,53). Future studies need to separate the precursor from the long chain derivatives.
Globally, available epidemiological studies strongly suggest a reduced risk for AMD in subjects with high consumption of long chain omega 3 fatty acids and fish. 
In addition, some studies have suggested that the risk for AMD may be decreased in subjects with high intake of vitamins B^(56-63). Consistently, supplementation with vitamins B reduced the incidence of AMD in a randomized interventional study⁽⁶⁴⁾. A role for vitamin D in the aetiology of AMD has also been suggested⁽⁶⁵⁾. 
 

4. Light exposure and cataract extraction

Sunlight exposure has been investigated as a potential risk factor for AMD. Light exposure may have deleterious effect on the eye, in particular through the production of reactive oxygen species⁽⁶⁶⁾. Only blue light reaches the retina, since ultraviolet radiations are absorbed by the cornea and the lens. Intense blue light exposure has been shown to induce retinal damage^(66), and the macular pigment, which absorbs blue light, is thought to protect the macula against photo-toxic damage⁽³⁰⁾. However, results have been inconsistent in the few studies that have investigated the associations of the light exposure with AMD in humans^(27,67-74). Globally, the available studies suggest that the effect of sunlight exposure in the aetiology of AMD is at most modest. Interestingly, a recent study evidenced an association of the risk for AMD with blue light exposure, only in those subjects with low plasma antioxidants and zeaxanthin⁽²⁷⁾. This suggests that light exposure may increase the risk for AMD only when defences against the produced reactive oxygen species are not appropriate. Sunlight exposure therefore does not appear to be a major determinant of AMD, but may be a risk factor in susceptible individuals. Appropriate nutritional intake in antioxidants and macular pigment may be particularly important in subjects highly exposed to light. These data will need to be confirmed in future studies.
Besides, cataract surgery was associated with a major increase in AMD incidence in a few studies^(7,75-81), although not all^(82-83). For instance, in a pooled analysis of two major population-based studies (Beaver Dam and Blue Mountains), eyes which had undergone lens extraction had a 5.7-fold increased risk of developing late AMD^(79). The reasons for increased risk of AMD in aphakic and pseudophakic eyes are unknown, but may include increased light exposure. Indeed, the lens naturally absorbs ultraviolet light, and, with the lens yellowing observed with ageing and cataract, also part of blue light. In this context, use of blue light filters in the implanted artificial lenses have been proposed⁽⁸⁴⁾, and are currently widely used, although their potential effect on the reduction of incidence of AMD has not been evaluated. 
Because cataract surgery is the most frequent surgical procedure in most industrialized countries, the potential increased risk of AMD in operated eye needs further study, in order to better characterize it, to determine its causes and to identify strategies to limit this potentially deleterious effect.

In conclusion, AMD is emerging as a disease resulting from major genetic susceptibility, the effect of which is modulated by lifestyle. Among lifestyle factors, smoking is the best characterized risk factor, now considered as causal, while the role of nutrition is increasingly identified. The respective roles of antioxidants, macular pigment and omega 3 fatty acids, together with other potential nutritional factors such as vitamins B and D, will be better understood in the future. In this field, several new epidemiological studies are being conducted, among which, in France, the Alienor Study⁽⁸⁵⁾. This population-based cohort study aims at assessing the association of AMD with nutritional, vascular and genetic risk factors. It bears on almost 1000 subjects, recruited from an existing cohort study on brain ageing (the 3C Study). The main nutritional factors studied are lutein and zeaxanthin, antioxidants (vitamins C, E, zinc) and omega 3 fatty acids, and are measured both in the diet, in plasma and, for lutein and zeaxanthin, on the retina. This study will add to the existing literature in this field, which is still relatively scarce and partially inconsistent. Moreover, several large controlled interventional trials, including the AREDS2 Study, will help demonstrating their causal role in the aetiology of AMD. Finally, the role of light exposure (in particular blue light) does not seem to be major determinant in this disease, but may be important in subgroups of the population (subjects with low antioxidant and macular pigment intake, subjects undergoing cataract surgery).

 

>> References

The outer retina

Age-related changes that predispose to age-related macular degeneration (AMD) occur in the outer retina, more specifically the region that includes the photoreceptors, the retinal pigment epithelium (RPE), Bruch

Retinal anatomy is highly organized and vascular and avascular compartments are strictly segregated in the retina⁽¹⁾. The blood-retinal barriers, inner and outer, are fundamental for the integrity of structure and optimization of function in neuro-sensorial retina⁽²⁾.

The outer blood-retinal barrier is formed, among its various components, by the RPE tight junctions. The intercellular cohesiveness of the RPE is not easily disrupted. Tight junctions appear as a necklace of strands that encircle each cell, binding each cell to its neighbors in a monolayer that separate the outer layer of the neural retina from the choriocapillaris⁽³⁾. Choriocapillaris is a great vascular network of fenestrated capillaries with high blood flow, fundamental for the metabolism of outer retina. This outer blood-retinal barrier retards transepithelial diffusion through the paracellular spaces⁽³⁾.

The RPE is a polarized epithelium that consist of a continuous pavement-like monolayer of cuboidal shaped cells that in macular area are tall, narrow and highly uniform in size and shape⁽⁴⁾. Interdigitation of the apical processes of the RPE with the cone and rod outer segments provides only a tenuous adhesion of the RPE to the sensory retina⁽⁵⁾.

RPE cells have at least ten known functions, but regeneration of visual pigments, transport of fluids and ions between photoreceptors and choriocapillaris, formation and maintenance of the interphotoreceptor matrix and Bruch’s membrane, phagocytosis of outer segments of photoreceptors, and supplying trophic factors such as VEGF-A should be emphasized^(6,7).

Bruch’s membrane is a thin, acellular and well-delineated membrane with five layers. From internal to external these layers are: the basement membrane of the RPE, the inner collagenous zone, the elastic tissue layer, the outer collagenous zone and the basement membrane of the choriocapillaris. It is composed of elements from both, the retina and choroid, but is an integral part of choroid⁽⁸⁾. Its inner surface is smooth, whereas its outer surface is composed of a series of collagenous protrusions that extend externally to form the pillars separating and supporting the choriocapillaris⁽⁵⁾. Due to its specific location and properties, this tissue is thought to be a vital limiting layer for metabolic transport between the RPE cells and the choriocapillaris⁽⁹⁾.

The choriocapillaris consists of a continuous layer of fenestrated endothelial cells surrounded by a basement membrane. In the macula the choriocapillaris is arranged in a lobular pattern of highly concentrated interconnecting capillaries supplied by a central arteriole and drained by circumferential venules^(5,10). The fenestrations, 60-80 nm in diameter, are abundant and seem to play an important role in permitting the passage of glucose and vitamin A to the RPE and retina. The choriocapillaris supplies oxygen and nutrients to Bruch´s membrane and the outer third of the retina, except in the macula, where it supplies the entire retina⁽⁸⁾. The peculiar structure of the choroidal vascular tree in the macula provides this area with the highest rate of blood flow of any tissue in the body⁽⁵⁾. 

 

Aging changes in outer retina and early AMD

With aging, the neurosensory retina was shown to develop thickening of the internal limiting membrane, diminution of neural elements with age-related loss of rods before cones, gliosis in the peripheral retina, and diminution of capillaries around the fovea, while the lumina of the choriocapillaris and the choroidal thickness become reduced by half^(11-13).

With the advancement of age, both the thickness and complexity of Bruch’s membrane increase primarily due to extracellular matrix remodeling and accumulation of inclusions in this region⁽¹¹⁾. Bruch’s membrane calcifies and doubles in thickness between the ages of 10 and 90 years⁽¹¹⁾. There is a linear thickening due to deposits of collagen, lipids and debris. After the 30’s its lipid concentration increases during life and consequently the fluid permeability and nutrient transport across the membrane decreases⁽¹⁴⁾. In normal conditions Bruch’s membrane acts as an intercellular matrix regulating survival of adjacent RPE and choriocapillaris cells. Its diminished function results in apoptosis of these cells from incorrect cell adhesion⁽¹⁴⁾. On the other hand extracellular deposits around Bruch’s membrane instigate chronic inflammation, invasion by dendritic cells and release of inflammatory cytokines, angiogenic factors and immune complexes^(15,16).

The RPE is a monolayer of regularly arranged hexagonal cells that spans the retina from the margins of the optic disc anteriorly to the ora serrata. The number of RPE cells diminishes with age. Macular RPE cells become wither, flatter and increase in height with advancing age^(4,17). In each RPE cell there is a progressive accumulation of lipofuscin during life and in people over 80 years of age, the debris can occupy more than one fifth of the total volume of an RPE cell^(18,19).  RPE cells have a brown color in young eyes but with age, they become increasingly more golden colored, owing to the accumulation of lipofuscin pigment granules⁽²⁰⁾. Lipofuscin in the RPE is the source of fundus autofluorescence. The major component of lipofuscin is N-retinylidene-N-retinylethanol-amine (A2E), a retinoid product of the visual cycle⁽²¹⁾. The A2E produced interferes with the function of RPE cells, leading to its apoptosis and subsequent geographic atrophy⁽²²⁾. Age-related changes also include a decrease in the number of melanin granules, loss of basal digitations and irregularity in shape. The RPE cells become separated from their basal membrane by membranous debris and abnormal secretory products and subsequently occurs deposition of collagen and fibronectin and latter formation of basal laminar deposits⁽²³⁾.

Basal laminar deposits are composed of basement membrane protein and long-spacing collagen located between the RPE plasma and basement membranes⁽²⁴⁾. Basal laminar deposits are considered the precursors of AMD and can appear around the age of 40 years⁽²⁵⁾.

Basal linear deposits consist of granular, vesicular or membranous lipid-rich material located external to the basement membrane of the RPE, in the inner collagenous layer of Bruch’s membrane, and represent a specific marker of AMD⁽²⁶⁾.

These two types of deposits can only be shown on pathological specimens and not by clinical evaluation⁽²²⁾.

The combination of the deposits with secondary changes in the RPE results in the formation of drusen. Drusen are localized deposits of extracellular material lying between the basement membrane of the RPE and the inner collagen layer of Bruch’s membrane^(23,27). Drusen often have a core of glycoproteins but they also contain fragments of RPE cells, crystallins, apolipoproteins B and E, and proteins related to inflammation such as amyloid P and β, C5 and C5b-9 complement complex^(28-31). 

Drusen change in size, shape, color, distribution and consistency with the passing years⁽³²⁾.

Small drusen are defined as being less than 63 μm in diameter⁽³³⁾. The presence of small, hard drusen alone is not sufficient to diagnose early AMD. These deposits are ubiquitous and the new development of small drusen in an adult eye without prior evidence of hard drusen is not age dependent⁽³⁴⁾.

Hard drusen are discrete nodules or deposits composed of hyaline-like material. During fluorescein angiography hard drusen behave as pin-point window defects⁽³⁵⁾.

Soft drusen are larger and associated with pigment epithelium detachment and diffuse abnormal Bruch’s membrane alterations^(36,37). Soft drusen have a tendency to cluster and merge with one another demonstrating confluence⁽³⁵⁾. During fluorescein angiography soft drusen hyperfluoresce early and either fade or stain in the late phase⁽⁵⁾.

Drusen can be visible in ophthalmoscopy when their diameter exceeds 25 μm as dots ranging in color from white to yellow⁽⁶⁾. However soft drusen are clinically identified whenever there is sufficient RPE hypopigmentation or atrophy overlying diffuse Bruch’s membrane thickening, or, when there are focal detachments within this material. These findings suggest that the clinical identification of soft drusen identifies an eye with diffuse changes at the RPE-Bruch’s membrane complex⁽³⁵⁾. When they become larger (>125 μm), and greater the area that they cover, the risk of late AMD becomes higher⁽³⁸⁾.

The RPE degeneration and nongeographic atrophy of the RPE are characterized by pigment mottling and stippled hypopigmentation with thinning of the neurosensory retina⁽³⁹⁾. Histopathology shows mottled areas of RPE hypopigmentation or atrophy overlying diffuse basal linear and basal laminar deposits⁽³⁶⁾. Incidence and prevalence rates of RPE depigmentation are age dependent⁽³⁴⁾.

Focal hyperpigmentation of the RPE, clinically evident as pigment clumping at the level of the outer retina or sub-retinal space, increases the risk of progression to the late phases of the disease^(34,36,40).

Aging is known to be associated with increased oxidative damage and the retina is a fertile environment for reactive oxygen species. Apart from the presence of two retinal blood supplies that generate an highly oxygenated environment, the exposure to high levels of cumulative irradiation, high levels of photosensitizers, large amounts of polyunsaturated fatty acids, readily oxidizable lipid, protein and carbohydrate substrates, and the huge proteolytic burden in the RPE contribute to this particular predisposition to oxidative stress^(41,42). This state of accumulation of toxic elements is said to ultimately tip the balance of the ocular immune privilege towards immune activation and inflammation. The end result of an innappropriate activation of diverse immune pathways, including classical and alternative complement pathways, is an immune-mediated retinal damage and/or an impaired immune-mediated retinal maintenance with RPE damage⁽¹³⁾. 

With all of the above in mind, Zarbin⁽⁴³⁾ has summarized his review on AMD pathogenesis in five sequential steps as follows: (1) AMD involves aging changes plus additional pathological changes (ie, AMD is not just an aging change); (2) in aging and AMD, oxidative stress causes RPE and, possibly, choriocapillaris injury; (3) in AMD (and perhaps in aging), RPE and, possibly, choriocapillaris injury results in a chronic inflammatory response within the Bruch membrane and the choroid; (4) in AMD, RPE and, possibly, choriocapillaris injury and inflammation lead to formation of an abnormal extracellular matrix, which causes altered diffusion of nutrients to the retina and RPE, possibly precipitating further RPE and retinal damage; and (5) the abnormal extracellular matrix results in altered RPE-choriocapillaris behavior leading ultimately to atrophy of the retina, RPE, and choriocapillaris and/or choroidal new vessel growth. In this sequence of events, patient's susceptibility to AMD would be determined by both the environment and his genetic profile⁽⁴³⁾.

 

Late AMD

The primary clinical characteristic of late dry AMD is the appearance of geographic atrophy of RPE. On microscopy, geographic atrophy is seen as abnormal RPE cells with hypotrophy, atrophy, hypertrophy, hypopigmentation, hyperpigmentation, migration, loss of photoreceptors, attenuation of Bruch’s membrane and choriocapillaris degeneration^(44,45). Geographic atrophy is clinically characterized by roughly oval areas of hypopigmentation that allows for increased visualization of the underlying choroidal vessels and is the consequence of RPE cell loss. Loss of RPE cells leads to gradual degeneration of photoreceptors and thinning of the retina that may extend to the outer plexiform and inner nuclear layers^{(6, 29)}. Compensatory RPE cell proliferation leads to hyperpigmentary changes frequently observed at the periphery of the hypopigmented areas⁽⁶⁾. The atrophy of RPE is usually more severe than the loss of choriocapillaris but the choriocapillaris seem to be highly constricted in areas of complete RPE cell loss⁽⁴⁴⁾.

In neovascular AMD early choroidal neovascularization occurs under the RPE⁽⁴⁶⁾and eventually breaks through⁽⁴⁷⁾, leading to accumulation of lipid-rich fluid under the RPE or neuroretina. In haemorrhagic forms blood breaks through the RPE into the subretinal space and sometimes through the retina and into the vitreous⁽³⁵⁾.

The pattern of growth of CNV often simulates that of a sea fan with radial arterioles and venules supplying and draining a circumferential dilated capillary sinus⁽⁵⁾.

As neovascularization of the sub-RPE space occurs, initially the blood flow through the neovascular network is sluggish and there is little or no exudation. This is a period of occult neovascularization and the overlying RPE and neuroretina may be minimally affected⁽⁵⁾. With an increase of blood flow through the network the endothelium decompensates and exudation extends into the subpigment epithelial space creating in some cases RPE detachments. The exudation may also extend through the RPE and detach the overlying retina.

In type II CNV the new vessels extend from the choroid through defects in Bruch´s membrane enters the space between the photoreceptors and RPE cells and growth laterally in the subretinal space⁽⁵⁾. This is usually accompanied by varying amounts of subretinal exudates and/or blood.

Macrophages have been documented both morphologically and functionally in neovascular AMD^(48,49). Activated macrophages and microglia may secret cytokines and chemokines that promote cellular damage and angiogenesis⁽⁵⁰⁾.

Involution of CNV eventually occurs and is associated with varying degrees of subretinal scar tissue, reactive hyperplasia of the RPE and/or atrophy, and can partially or totally replace the neuroretina⁽²¹⁾. The outer nuclear layer can be severely attenuated with a reduction of photoreceptor length of almost 70%⁽⁵¹⁾. Often anastomosis between the retinal circulation and the underlying choroidal circulation develops within these old disciforme scars^(32,52).

Other factors like complement factor H, that downregulates the alternative complement pathway⁽⁵³⁾, HtrA1 – a secretory protein and an inhibitor of transforming  growth factor β (TGF-β)⁽⁵⁴⁾- and ARMS2⁽⁵⁵⁾ play a role in development of AMD. However its specific role and relevance in development and progression to neovascular and atrophic forms of AMD are discussed in others chapters of this book.

 

Key points

• Age-related changes that predispose to AMD occur in the outer retina, more specifically the region that includes the photoreceptors, the RPE, Bruch’s membrane and the choriocapillaris.
• The deposition of insoluble material, the calcification and increase in thickness of Bruch’s membrane, and a less fenestrated and thinner choriocapillaris leads to photoreceptors/RPE hypoxia. The number of RPE cells reduces with age and in each cell there is a progressive accumulation of lipofuscin during life. Basal laminar and basal linear deposits, drusen, RPE degeneration and atrophy develop and finally, geographic atrophy of RPE and choroidal neovascularization occur.
• Choroidal neovascularization is a relative self-limited disease with development, growth and involutional stages leading to the formation of a scar and destruction of macular structure and function.

 

>> References

Introduction:

Age-related macular degeneration (AMD) is a complex disease with demographic and environmental risk factors (age, diet, and smoking) but also genetic risk factors.

In fact, instead of having a single contributory gene, there are multiple genes of variable effects that seem to be involved turning the issue of genetics of AMD a complex one: AMD involves environmental factors and varying susceptibilities to these external factors based upon diferent genetic backgrounds⁽¹⁾.

The genetic component of the disease has been suspected from family, twin and sibling studies. According to several family studies, patients with a family history of AMD have an increased risk for developing AMD^(2,3).

The concordance for the presence of the disease is greater among homozigous twins than among heterozygous twins.

In 2008, Luo et al.⁽⁴⁾ estimated the magnitude of familial risks in a population–based cross-sectional and case–control study.

Recurrence risk in relatives indicate increased relative risks in siblings (2.95), first cousins (1.29), second cousins (1.13) and parents (5.66) of affected elements.

Many linkage and association studies have showed that chromosomes 1q (1q25-31) and 10q (10q26) had genes involved in this pathology^(5-7).

It was the completion of the Human Genoma Project, 5 years ago, resulting in the knowledge of the sequencing of the human genome that allowed improved DNA sequencing and mapping technologies and consequently, identification of Single Nucleotide Polymorphism (SNPs).

There are several types of genetic sequences variations (polymorphisms) in the human genome: repeated polymorphisms, insertions and deletions. However the majority of the DNA sequence variations in the human genome is in the form of SNPs which are persistent single changes, substitutions or variants of a single base in at least a population and with a frequency of more than 1% referred as alleles and representing altered forms of a gene: different alleles may produce variations in inherited characteristics^(8,9).

These variants are also important because they serve as genetic markers and in this way they can help in determining those which confer increased or decreased risk of several diseases including AMD.

Dissection of the genetic background of AMD has undergone tremendous progress in the last 2 years. We know, now, some polymorphisms which modulate AMD risk.

 

Complement system factors and AMD

Complement factor H (CFH)

 

CFH is a negative regulator of alternative pathway of the complement system which means that in normal conditions, it inhibits the alternative pathway complement system.

It is encoded by a gene localized in 1q23-32 and its dysfunction may lead to excessive inflammation and tissue damage⁽¹⁰⁾.

Complement activity is very important for the imune responses against pathogens and dying cells but, over-activation can result in complement-mediated damage to nearby healthy tissue cells.

It is now accepted that CFH gene is an important susceptibility gene, harbouring variants and haplotypes (short DNA sequences containing alleles) associated with increased and reduced risk of AMD.

Six CFH gene variants have been reported in AMD association studies as major genetic factors for developing AMD in Caucasians^(11-15): rs1061170, (CFH Y402H); rs3753394; rs800292; rs1061147; rs380390; rs1329428.

However in the Chinese and Japanese populations only three of these CFH SNPs (rs1329498, rs800292 and rs3753394) were associated with risk of AMD^(16,17).

So it is possible that CFH could play a central role in AMD pathogenesis and that multiple SNPs that alter CFH function might contribute to the development of AMD.

Their importance varies among the race of the population.

In the variant (polymorphism) CFH Y402H of the CFH gene, there is a substitution on the nucleotide in exon 9 (1277) where thymine (T) is changed for cytosine (C) (rs1061170) which is the allele risk.

This change leads to the substitution of the aminoacid in the position 402 in the protein, from tyrosine (Y) to histidine (H).

Homozygote CC or heterozygote TC can account for 50% of AMD cases.

The risk attributable for a disease is the rate of disease among individuals with a given characteristic minus the rate of the disease among indivuals without that characteristic.

The population attributable risk (PAR) in individuals with this polymorphism for developing AMD is 43% to 50%^(11,12,18).

When compared with those with no risk allele TT, one copy of the Tyr402His polimorphysm (heterozygous for the risk allele TC), increases the risk of AMD by a factor of 2.2 to 4.6 (these individuals are at least twice and half more likely to develop AMD) and two copies of the risk allele (homozygous for the risk allele CC) increases the risk by a factor of 3.3 to 7.4 in Whites⁽¹⁹⁾.

In addition to the common risk haplotype carrying the C allele of CFH Y402H, haplotype analysis of CFH has revealed two common protective haplotypes: homozygous deletions CFHR1 or CFHR3.

The gene cluster of CFH includes other “CFH-related genes“: CFHR1, CFHR2, CFHR3, CFHR4 and CFHR5. This means that the CFH gene resides within the region of complement activation (RCA), which includes also five “CFH–related” genes.

While the function of the CFH related genes is largely unknown, the high degree of sequence similarity between these genes and the suggestion that they arose out of duplication events with CFH, suggest an overlapping function of the CFH-related genes in immune system function and /or regulation.

There is a common and widespread (commonly found in many different populations in the world) deletion within the RCA locus that encompasses the CFHR1 and CFHR3 genes.

However the frequency of homozygous CFHR1 or CFHR3 deletion shows considerable variation between ethnic groups and occurs in 17.3% of African populations, 15.9% of African-American, 6.8% of Hispanic, 4.7% of Caucasian and 2.2% of Chinese cohorts⁽²⁰⁾.

This is in agreement with AMD less frequency among African-Americans compared with Caucasians and Chinese populations.

CFH1 and CFH3 protein may compete with CFH for C3 binding and therefore interfere with normal regulation of the complement system.

Those individuals, who are homozygous for the CFHR1/CFHR3 deletions and, therefore do not express the respective proteins, are highly protected from developing AMD⁽²⁰⁾.

 

Complement factor B (BF), complement component 2 (C2)

Complement factor B (BF) is involved in the activation of the complement alternative pathway and complement component 2 (C2) is involved in the activation of the classical pathway of the complement and both have adjacent genes located 500 base pair apart on chromosome 6p21.3 within the major histocompatibility complex class III region⁽²¹⁾.

Haplotypes in BF and C2 have been linked to AMD. In particular, the L9H in BF and the E318D in C2 and also the R32Q in BF and a variant in intron 10 of C2 have been showed to be protective for AMD by Gold et al.⁽²¹⁾.

They hypothesized that the significance of the haplotypes is due largely to the BF variants, which are in strong linkage disequilibrium with C2.

BF is a complement activating factor and studies have demonstrated that at least one of the two variants associated with AMD (R32Q BF) leads to an impairment in the complement activation function of BF.

This means that the absence of these variants C2/BF can predispose patients to AMD⁽²¹⁾.

Thus, much like impaired CFH-mediated complement inhibition confers AMD risk, decreased complement activation by BF might serve to protect against AMD risk.

 

Complement component 3 (C3)

C3 is the central element of the complement cascade and a candidate gene to be involved in AMD, since its cleavage product, C3a, not only was found in drusen but also was proved to induce vascular endothelial growth factor expression and promote choroidal neovascularization in both in vitro and in vivo^(22-24).

The variants R102G and P314L of the C3 gene significantly increase the risk of early and all subtypes of AMD and this risk seems to be independent of CFH Y402H, LOC387715 A69S and smoking⁽²⁵⁾.

 

LOC387715/ARMS2 and HTRA1

In 2003 Majewski et al. suggested that chromosome 10q26 might contain an AMD gene⁽⁷⁾.

Later this finding has been replicated by other genome-wide linkage studies(26) and supported by a genome–scan meta-analysis⁽⁵⁾.

This locus contains three tightly linked genes: PLEKHA1, LOC387715/ARMS2 (age related maculopathy suceptibility gene 2) and HTRA1, a secreted heat shock serine protease.

In 2005 Jakobsdottir et al.⁽²⁷⁾ found that the strongest association was over LOC387715/ARMS2 and HTRA1, which share an extensive linkage disequilibrium (LD) block harbouring the high risk haplotype.

There has been more dispute than agreement between the studies in what concerns this locus.

All initial genetic studies, about ten years ago, lacked statistical power (because small samples were used), used cumbersome genotyping technologies and poorly defined cohort.

In recent years, there are some publications of preliminary and unconfirmed genetic associations of the genes in this locus to AMD⁽²⁸⁾.

 

ARMS2 (LOC387715) SNP and AMD

The association of ARMS2 gene and AMD has been now replicated in various independent studies especially the advanced form of the disease, that means to say, the “wet” or with choroidal neovascularization and the “dry” or geographic atrophy form of the disease^(29-32).

The risk confering polymorphism consists in a change in the 69 position aminoacid alanine (A) to serine (S).

According to Ross⁽³¹⁾, heterozygosis at the ARMS2/LOC387715 ( A69A/A69S) is associated with odds ratio (OR) of 1.69-3.0 for advanced AMD while homozygozity for the risk conferring allele (A69S/A69S) results in a OR of 2.20-12.1.

The frequency of the risk allele is higher in patients with advanced AMD than in those with early or intermediate AMD^(27,33).

Later two studies, based on semiquantitavive expression data of allele associated differences in HTRA1 mRNA or protein levels, suggested a diferent variant (rs11200638) in the same LD block, in the promoter of HTRA1 gene as the functional variant^(34,35).

 

HTRA1 (high temperature required factor A-1) SNP and AMD

HTRA1 gene is located on chromosome 10q26.3, extremely close to the locus of the ARMS2 gene (10q26.13) and because of its role in extracellular matrix homeostasis (its extracellular protease activity may favour neovascularization) and in cellular growth or survival (it is an inhibitor of TGF-ß family member⁽³⁶⁾ and it could play a critical role in controlling TGF-ß dependent neuronal survival⁽³⁷⁾ it seems a possible functional candidate gene.

Four significant SNPs have been reported in the promoter and the first exon of HTRA1: G625A (rs11200638); T487C (rs2672598); C102T, A34A (rs1049331); G108T, G36G (rs2293870). However the most well documented, statistically significant AMD associated SNP is rs11200638 (G625A) in the promotor region.

Caucasians, Chinese and Japanese heterozygous for the risk allele (G/A) have a high OR of 1.60-2.61 and Caucasians, Chinese and Japanese homozygous for the risk allele (A/A), 6.56-10.0^{(34,35,38-40)}.

According to Tam et al.⁽⁴⁰⁾, there is an increase in population attributable risk (about 5.5 fold increase) by the joint effect of smoking and HTRA1 allele.

This means that smokers homozygous for the risk allele had a substantially higher risk of developing wet AMD than non smokers with the risk allele.

However Deangelis et al⁽⁴¹⁾, in 2008 reported no interaction between this SNP and smoking.

In what concerns the studies which relate HRTA1 promoter polymorphisms to risk factors for developing AMD, three problems arise according to Allikmets and Dean⁽²⁸⁾.

The variant encoding the A69S (rs10490924) in ARMS2 and the rs11200638 variant in HTRA1 are almost in complete LD, so it is impossible to assign causality on the basis of allele frequency alone.

10q26 locus doesn’t harbour a wet AMD gene as the authors claimed but a late AMD gene as showned by Weber and colleagues in 2005⁽³⁰⁾.

All subsequent studies have failed to replicate the functional data^(32,42).

This basically means that, as there is strong linkage disequilibrium (LD) across ARMS2-HTRA1 region, genetic association studies alone are insufficient to distinguish between the two candidates.

It is also necessary not only the characterization of the extent of the variants associated to the disease but also the analysis of their possible functional relevance in the disease process⁽⁴²⁾. Doing this, Fritsche et al.⁽⁴²⁾ claimed that the functional variant in this locus is the deletion-insertion polymorphism variant 372-815delins54 in the ARMS2 gene.

The deletion removes the polyadenilation signal sequence at position 395-400 exclusively used for the addition of a poly A tract 19 bp downstream.

The insertion introduces a 64 bp AU-rich element, known for its properties to control mRNA decay in many transcripts that encode a wide variety of proteins^(43,45).

They demonstrated that it is a major risk factor for AMD: individuals carrying a single copy of the risk allele deletion-insertion in ARMS2 gene have a 2.8–fold increased risk compared with an 8.1–fold increased risk in homozigous individuals.

Their work, also revealed that in homozygous for the deletion-insertion variant, expression of ARMS2 is absent.

They localized the ARMS2 protein within the photoreceptor layer namely, to the mitochondria–enriched ellipsoid region of the inner segments and in accordance; they proposed a functional role of ARMS2 in mitochondrial homeostasis.

According to Fritsche et al., this suggests, that this polimorphysm is the sought-after functional variant with relevance in AMD etiology in 10q26 locus.

However, as Fritsche et al. recognize, it is ultimately required formal exclusion of functional consequences for the remaining polymorphisms on the risk haplotype namely, A69S in ARMS2 gene and HTRA1 promoter variant.

The A69S and the InDel are in 100% LD and on the same haplotype and so the effects are not independent to each other^(46).

The work of Fritsche and colleagues does not eliminate all other possibilities⁽²⁸⁾, nobody disputes the role of complement genes in AMD in spite of the functional consequences of the disease associated variants being not known for CFH, CFB/C2 and/or C3.

 

Apolipoprotein E gene (ApoE)

The ApoE gene, located on chromosome 19q13.2, is polymorphic and has three isoforms wich are common, E2, E3 and E4, coded by different alleles: the ancestral E3 and the SNPs, E2 and E4⁽⁴⁷⁾.

Most studies favour a protective role for the ApoE4 SNP and a slight risk-conferring role for ApoE2^(47-51).

However other studies do not^(52-55).

 

SNP genotype and therapeutic responses

Genotype and response to antioxidative/zinc therapeutics

One of the first works in this area was that of Michael Klein et al.⁽⁵⁶⁾.

These authors correlated the CFH and LOC387715 A69S genotypes with the therapeutic responses to supplementation with antioxidants and zinc.

They concluded that, in homozygous individuals for the non-risk phenotype (Y402Y/Y402Y), 34% of those treated with placebo progressed to advanced AMD, compared to 11% of those treated with antioxidants and zinc: a reduction of approximately 68%.

In homozygous individuals for the risk allele (Y402H/Y402H), 44% of those treated with placebo progressed to advanced AMD, compared with 39% of those treated with antioxidants plus zinc: a reduction of only 11%.

A similar interaction was observed in the groups taking zinc versus those not taking zinc: intake had a more protective effect in patients with non-risk alleles compared to patients with risk alleles.

These results suggest that the zinc plus antioxidative treatment seems to have less impact on those with the high-risk CFH variant.

These authors found no association between LOC387715/ARMS2 A69S and the response to AREDS treatment.

 

Genotype and response to intravitreous bevacizumab

Brantley et al⁽⁵⁷⁾ investigated 86 wet AMD patients for the association between CFH and LOC387715/ARMS2 genotypes and the response to treatment with bevacizumab.

For the CFH genotypes results show that only 10.5% of patients homozigous for the risk - conferring allele Y402H/Y402H genotype demonstrated improved vision with treatment compared with 53.7% of patients homozigous for the non-risk allele Y402Y/Y402Y and the heterozigous for the non-risk allele Y402Y/Y402H variants.

They found that the CFH variants are associated to the responses to bevacizumab but that LOC387715/ARMS2 are not.

 

Genotype and response to photodynamic therapy

Goverdhan et al⁽⁵⁸⁾ and Brantley et al⁽⁵⁹⁾ studied the association of the CFH and LOC387715 genotypes with the response to PDT.

They found no statistical association with the LOC387715 genotype but a statistical association with CFH genotype: risk allele genotypes have better results than non-risk allele genotypes.

However more studies are warranted before any definitive conclusions.

 

Conclusions

1. Genetics variants at two chromosomal loci, 1q31 and 10q26, confer major disease risks, together accounting for more than 50% of AMD pathology^{(11-14, 27,30)}.

At present SNPs are the best available markers of AMD risk: SNPs in complementt factor H and ARMS2/HTRA1 capture a substancial fraction of AMD risk and permit the identification of individuals at high risk of developing AMD.

Genetic markers can successefully identify individuals whose lifetime risk of age-related macular degeneration ranges from 1% to greater than 50%.

2. Understanding the genetic basis of AMD has important implications for the ophthalmologists as it allows the identification of the biochemical pathway for a large proportion of AMD patients, raises the possibility to perform pre-symptomatic diagnostic testing of risk genotypes and to stratify the response to therapy based on genetic risks and supports the development of new therapies being the inhibition of complement a potential one among others that are already being tested.

 

>> References

Definition

Blood vessels develop and grow by three different basic mechanisms: vasculogenesis in which vessels form by concatenation of vascular precursor cells into solid cords that then lumenize; angiogenesis that is the growth of new blood vessels from pre-existing ones; and intussusception in which new blood vessels form by the proliferation of endothelial cells that form a pre-existing vessel into the vessel lumen, originating two blood vessels that split into two opposite sides⁽¹⁾.

Angiogenesis, the growth of new vessels from pre-existing ones by sprouting of endothelial cells into a previously avascular tissue, is an essential process both in embryonic development and in adulthood^(1,2). It is a complex multistep process involving extracellular matrix degradation and proliferation, survival, migration and anastomosis of endothelial cells⁽²⁾.

The release of extracellular matrix proteases leads to the degradation of the basal membrane of blood vessels. Endothelial cells change shape, proliferate, invade stroma and form tubular structures that coalesce. This requires the coordinated action of a variety of anti and pro-angiogenic factors and cell-adhesion molecules in endothelial cells. Angiogenesis is of paramount importance as it promotes tissue repair; however, in certain conditions it may cause tissue damage. If not tightly regulated, the angiogenic process is frequently imbalanced, and associated with several pathological situations^(1,3).

 

Angiogenic mediators and modulation of their expression

The angiogenic process requires the activation of a series of receptors by numerous ligands including Placental Growth Factor (PIGF), Fibroblast Growth Factors (FGFs), Angiopoietin-1 and -2 (Ang-1 and -2), Platelet-derived Growth Factor (PDGF), Hepatocyte Growth Factor (HGF), Connective Tissue Growth Factor (CTGF) and Transforming Growth Factors (TGF-α e TGF-β), among many others^{(1, 3-9)}.

However, there is a consensus that Vascular Endothelial Growth Factor (VEGF) is the most important angiogenic factor and represents the crucial rate-limiting step during angiogenesis^(3,10,11).

VEGF-A is the prototype member of a gene family that also includes placental growth factor (PlGF), VEGF-B, VEGF-C, VEGF-D, and the orf-virus-encoded VEGF-E⁽¹¹⁾. Alternative exon splicing results in the generation of four main VEGF isoforms, which have respectively 121, 165, 189, and 206 amino acids (VEGF121, VEGF165, VEGF189, VEGF206). Less frequent splice variants have also been reported, including VEGF145, VEGF183, VEGF162, and VEGF165b^(8,11).

VEGF mediates its biological functions at the endothelial level by binding two highly related tyrosine kinase receptors (RTKs), VEGFR-1 and VEGFR-2. It is generally agreed that VEGFR-2 is the major mediator of the mitogenic, angiogenic and permeability-enhancing effects of VEGF-A⁽³⁾. VEGFR-1 binds to both VEGF-A and Placenta Growth Factor and fails to mediate a strong mitogenic signal in endothelial cells. It is now generally agreed that VEGFR-1 plays a role in the modulation of VEGF activity⁽¹⁰⁾. VEGFR-3 is mainly a VEGF-C receptor but it plays important roles in lymphangiogenesis and angiogenesis⁽¹²⁾.

The mediators of the angiogenic process can be modulated by some molecules and microenvironmental conditions. VEGF is upregulated by cyclo.oxygenase (COX-2)⁽¹³⁾. Inflammatory cells within a hypoxic environment release huge amounts of factors that exert effects on endothelial cells and degrade the extracellular matrix⁽²⁾.

Angiogenesis can also be suppressed by inhibitory molecules, such as interferon-α, thrombospondin-1, angiostatin, endostatin or pigment epithelial-derived factor (PEDF)^(14-18).

It is the balance of stimulators and inhibitors that tightly controls the normally quiescent capillary vasculature. When this balance is upset angiogenesis develops⁽¹⁹⁾.

Cell-cell and cell-matrix interactions may also play an important role in angiogenesis. Special focus has been recently given the Rac1 GTPase⁽²⁰⁾. Recently, microRNA regulation of gene expression has been implicated in the control of pathologic ocular angiogenesis⁽²¹⁾.

The angiopoietin (ANG)–TIE signalling pathway has been identified as the second vascular tissue-specific receptor Tyr kinase system. The ANG–TIE pathway is required for lymphatic and blood vessel development and is important for the development of mature blood vessels that originate from the VEGF induced endothelial sprouting. This pathway controls vascular permeability, inflammation and pathological angiogenic responses in adult tissues⁽²²⁾. Maturing of blood vessels include pericyte and coating of endothelial cell walls.

PDGF is mainly believed to be an important mitogen for connective tissue. PDGF promotes migration and proliferation of endothelial cells as well as an increased recruitment of pericytes. These findings suggest that PDGF is not only important in formation of new blood vessels but it is also very important for their maturation and stabilization⁽²³⁾.

Evidence suggests that processes of inflammation and angiogenesis are connected. Newly formed blood vessels enable the recruitment of inflammatory cells, which release a variety of proangiogenic cytokines and growth factors that will perpetuate angiogenesis⁽²⁴⁾.

 

Angiogenesis during development of retinal vasculature

During embryogenesis retinal vascularization begins in the most superficial (or inner) retinal layers at the optic nerve head, and radiates outwards from this central point. It reaches the retinal periphery just before birth⁽²⁵⁾.

The migration of large numbers of vascular precursor cells (VPCs) from the optic disc is the first event in human retinal vascularization, and it is apparent before 12 gestational weeks⁽²⁶⁾. They proliferate and differentiate to form a primordial vascular bed centered on the optic disc. Thus, vasculogenesis is responsible for the formation of the primordial vessels of the inner (superficial) plexus in the central human retina⁽²⁷⁾. Formation of retinal vessels via vasculogenesis seems to be independent of metabolic demand and hypoxia-induced VEGF expression⁽²⁸⁾.

Angiogenesis is responsible for the formation of the remaining retinal vessels, including increasing vascular density in the central retina, vessel formation in the inner plexus of the peripheral retina, and formation of the outer plexus and the radial peri-papillary capillaries⁽²⁸⁾. Formation of the outer plexus begins around the incipient fovea between 25 and 26 weeks of gestation, coincident with signals that indicate a functional visual pathway and photoreceptor activity⁽²⁷⁾. The timing and topography of angiogenesis in the human retina supports the “physiological hypoxia” model of retinal vascular formation, in which angiogenesis is induced by a transient but physiological level of hypoxia as a result of the increased metabolic activity of retinal neurons as they differentiate and become functional⁽²⁹⁾.

 

Angiogenesis in retina and choroidal pathologies

Retinal anatomy is highly organized and vascular and avascular compartments are strictly segregated in the retina⁽¹⁾. The blood-retinal barriers, inner and outer, are fundamental for the integrity of structure and optimization of function of the neuro-sensorial retina⁽³⁰⁾.

Pathological retinal and choroidal angiogenesis generates chaotically orientated and physiologically deficient vessels that do not conform to neuronal histology, which can lead to vision-threatening oedema, exudation and haemorrhage⁽¹⁾.

Angiogenesis is a key aspect in many ocular pathologies that are leading causes of blindness in the world, such as neovascular age-related macular degeneration (AMD), diabetic retinopathy, retinopathy of prematurity, central retinal vein occlusion and other diseases associated with ischemia and neovascularization⁽³¹⁾.

Although angiogenesis is a highly complex and coordinated process requiring multiple receptors and ligands in endothelial cells, VEGF is a hypoxia-inducible cytokine that appears to be a pivotal element required for the process in a variety of normal and pathological circumstances^{(3, 10)}. VEGF is a surrogate angiogenic marker, since it acts not only as a mitogen, but also as a survival factor for endothelial cells⁽²⁾. Furthermore, it is also involved in the stimulation of the invasive and migration capacity of endothelial cells and in the enhancement of vascular permeability⁽¹⁰⁾. Bone-marrow derived cells have also been described in choroidal neovascular lesion. Their importance in the pathologic process is still under debate⁽³²⁾.

 

Angiogenesis and Age-related Macular Degeneration

The diagnosis of AMD is based on fundoscopic signs observed on the macula, irrespective of visual acuity⁽³³⁾. The stages of AMD are categorized as early, in which visual symptoms are inconspicuous, (moderate) and late, usually associated with severe loss of vision⁽³⁴⁾. Early AMD is characterized by the presence of drusen and/or hyperpigmentations or small hypopigmentations⁽³³⁾. Late AMD has “dry” and “wet” forms. However, in the same patient we can find the dry form in one eye and the wet in the other eye, or both forms in the same eye. Moreover with time we can see the conversion of wet in dry or dry becoming wet⁽³⁵⁾.

Age-related changes that predispose to AMD occur in the outer retina, more specifically the region that includes the photoreceptors, the retinal pigment epithelium (RPE), Bruch’s membrane and the choriocapillaris.

The aging-dependent alterations in the outer retina have been already discussed in another chapter. AMD-related visual loss is a complex process starting by the deposition of debris in the outer retina⁽³⁶⁾. The deposition of insoluble material, the calcification and increase in thickness of Bruch’s membrane, and a less fenestrated and thinner choriocapillaris leads to photoreceptors/RPE hypoxia resulting in a stimulus for VEGF release^{(35, 37-39)}.

All the aging changes in outer retina compromise the nutrition of photoreceptors and RPE and create a favourable environment for the development of choroidal neovascularization (CNV). However other factors – genetic and environmental^{(40, 41)} – are also important, but its role in the development of CNV is discussed in other chapters of this book.

In general terms there are two basic CNV growth patterns, based on the anatomical position of the abnormal vessels with respect to the RPE monolayer, which are related to Gass’s classification of CNV^{(42, 43)}. In type 1 CNV, the neovascular complex is located in the plane between the RPE and Bruch’s membrane and in type 2 neovascularization the vessels have penetrated the RPE layer to proliferate in the subneurosensory space⁽⁴²⁾.

In the type 1 growth pattern, after breaking through Bruch’s membrane, the CNV extend laterally under the RPE in a horizontal fashion that is facilitated by the natural cleavage plane between basal laminar deposits and a lipid rich Bruch’s membrane. This growth pattern recapitulates the choriocapillaris and can provide some nutrients and oxygen to an ischemic RPE/outer retina^{(43, 44)}. 

The type 2 growth pattern occurs usually with one or few ingrowth sites with vascular leakage under the RPE/outer retina that usually lead to acute visual simptoms⁽⁴³⁾.

Yannuzzi proposed a type 3 neovascularization, for retinal angiomatous proliferation (RAP), indicating proliferating vessels within or below the retina itself⁽⁴⁵⁾. This mixed neovascularization, with a presumed dual origin, may have intraretinal neovascularization driven by angiogenic cytokines from Müller cells, endothelial cells, pericytes, and retinal glial cells, and CNV driven by cytokines from the RPE⁽⁴⁵⁾.  There is hypothetically neovascularization extending anteriorly from the choroid in conjunction with retinal neovascularization progressing posteriorly, with both circulations eventually anastomosing. 

The reason that growth patterns vary according to disease and individuals may be related to genetic predispositions, environmental mechanisms, variations in composition and anatomy of Bruch’s membrane, cytokine distribution, or other causes⁽⁴³⁾.

During the dynamic process of development of CNV there is a balance of angiogenesis promoters and inhibitors. In the initiation stage the RPE and photoreceptors produce VEGF⁽⁴⁶⁾. There is also production by RPE of Interleukin-8 (IL-8) and Monocyte Chemoattractant Protein-1 (MCP), witch attract monocytes from the choriocapillaris along the outer surface of Bruch’s membrane⁽⁴⁷⁾. The macrophages tend to concentrate around sites of vascular ingrowth through the Bruch’s membrane and express Tumor Necrosis Factor-α (TNF-α) and Interleukin-1 (IL-1), which up-regulate complement factor-B, activates the complement alternative pathway in the subretinal space, and stimulates RPE cells to produce more VEGF^(47,48).

After initiation, CNV grows to a certain size and progresses through the tissue planes by the action of Matrix Metalloproteinases (MMP) produced by endothelial cells and macrophages⁽⁴⁹⁾. During this stage of active growth, Angiopoietins (Ang-1 and 2) are expressed, FGFs are produced by RPE and endothelial cells, and TGF-β is produced by the RPE^(50-52). 

CNV stabilizes during the active stage due to a steady state established between MMP and tissue inhibitors of metalloproteinases, Ang-1 and 2, PEDF and VEGF, PDGF and VEGF, plasminogen and fibrin, and others^{(43, 53)}.

At some point the balance shifts toward antiangiogenic, antiproteolytic and antimigratory activity resulting in the involutional stage of CNV. When this occurs the angiogenic/proteolitic/migratory cytokine production decreases with a shift toward TGF-β and tissue inhibitors of metalloproteinases production by the RPE⁽⁴³⁾. In this involutional stage the CNV may become collagenized and form a disciform scar. Subretinal fibrosis is the hallmark of this stage of the disease and is the result of the activity of inflammatory and endothelial cells and fibroblasts⁽⁵⁴⁾.

 

Key points

• Angiogenesis, the growth of new vessels from pre-existing ones, is a complex multistep process involving the activation of series of receptors by numerous ligands.
• Vascular Endothelial Growth Factor (VEGF) is the most important angiogenic factor and represents the crucial rate-limiting step during angiogenesis.
• There are three basic CNV growth patterns, based on the anatomical position of the abnormal vessels with respect to the RPE monolayer and to the proliferation beginning within or below the retina itself: occult (type 1), classic (type 2) and RAP (type 2) lesions.
• Interleukin-8 (IL-8), Monocyte Chemoattractant Protein-1 (MCP), Tumor Necrosis Factor-α (TNF-α), Interleukin-1 (IL-1), Matrix Metalloproteinases (MMP), Angiopoietins (Ang-1 and 2), Fibroblast Growth Factors (FGFs) and Transforming Growth Factor β (TGF-β), among others, play an active role during the stage of active new vessel growth.

 

>> References

Introduction

Age-related macular degeneration (AMD) is the leading cause of blindness in older population in industrialized nations^(1,2). It is characterized by ageing changes at the photoreceptors, retinal pigment epithelium (RPE), Bruch’s membrane and choroid^(3,4).

The clinical hallmark of the early stage of AMD is drusen, which are yellowish deposits at the level of the RPE. Focal RPE hyperpigmentation and atrophy can also be seen. The advanced stage of non-neovascular AMD is Geographic Atrophy (GA) where areas of atrophy become confluent and cause visual loss. Neovascular AMD (NV-AMD) is also an advanced manifestation of AMD and is characterized by choroidal neovascularization (CNV).

This review summarizes the various biomarkers of AMD and analyses whether or not they may, one day, be exploited to determine risks of disease onset, measure progression of disease or even assess the effects of treatment of AMD.

 

Environmental Risk Factors

See chapter “Modifiable Risk Factors for AMD”.

 

Genetic biomarkers

Genetic predisposition for AMD has been demonstrated by familial aggregation studies and twin studies^(5,6). Using genome linkage scan and association studies, multiple potentially causative genes and/or single nucleotide polymorphisms (SNPs) have been identified^(7-9).

The most commonly implicated genetic risk loci are located in chromosome 1q25-31, which are related to several complement system regulators, such as complement factor H (CFH) and CFH-related proteins (CFHR1-5)⁽¹⁰⁾. CFH, in particular, is a key regulator of the complement system of innate immunity: it maintains the homeostasis of the complement system and protects bystander host cells and tissues from damage by complement activation⁽¹¹⁾. At least 6 genetic variants related to AMD development in Caucasians have been identified^(7,12-14), while only three of these have been described in Chinese and Japanese populations^(15,16).

The most documented risk-conferring single-nucleotide polymorphism results is a tyrosine-to-histidine substitution in 1q26 at position 402 of the CFH protein. This CFHY402H polymorphism increases AMD risk 2 to 4-fold in heterozygotes and 3 to 7-fold in homozygotes^(7,17). It causes a significant loss of binding of factor H to heparin, C-reactive protein and to peroxidised lipids on dead cells combined with reduced anti-inflammatory functions by factor H, leading to increased local complement activation⁽¹⁸⁾.

The CFHR genes comprise five plasma proteins (CFHR1-5) that bind to the central complement component C3b. These are located at 1q32 within the RCA (Regulation of Complement Activation) gene cluster⁽¹⁰⁾. A homozygous deletion of CFHR1 and CFHR3 is a relatively common finding in several populations and appears to confer a lower risk of AMD⁽¹⁹⁾. A possible explanation for the protective effect of these deletions might be that CFHR1 and CFHR3 compete with CFH for binding to C3b, thus a deficiency of these proteins increases the efficacy of factor H and its regulatory effect on complement activation⁽¹⁰⁾. The prevalence of this deletion differs between ethnic groups, occurring in 17.3% of Africans, 15.9% of Afro-Americans, 6.8% of Hispanics, 4.7% of Caucasians and 2.2% of Chinese. This finding is consistent with the AMD prevalence rates, which are lower in African-Americans than in Caucasians and Chinese⁽¹²⁾.

For an in-depth review of genetic biomarkers of AMD, please refer to the chapter “Genetics of AMD”.

 

Inflammatory biomarkers

Besides components of the complement system⁽²⁰⁾, several immunological molecules and inflammatory mediators have been identified at the site of AMD lesions⁽²¹⁾. In fact, chronic inflammation appears to be a causative factor for the development of AMD by causing endothelial dysfunction and facilitating interactions between modified lipoproteins, monocyte-derived macrophages, T-cells and normal cellular elements of the retinal vasculature⁽²²⁾. Activated macrophages and microglia may cause cellular damage, Bruch’s membrane degradation and angiogenesis by secreting chemokines and cytokines⁽²²⁾. In fact, using electron microscopy or immunohistochemistry methods, macrophages can be found in the area of geographic atrophy phagocytising pigment debris⁽²³⁾ and around CNV in wet AMD⁽²⁴⁾.

In order to grasp the role of inflammation on the development and progression of AMD, several markers of systemic inflammation have been extensively studied. C-reactive protein (CRP), an acute phase serum protein, is a surrogate for interleukin-6 (IL-6)⁽²⁵⁾. Classically seen as an inflammatory marker, CRP is now regarded as an independent risk for both cardiovascular and peripheral arterial disease⁽²⁶⁾. It directly upregulates endothelial cell adhesion molecules and promotes the release of chemoatractant chemokines, which have a negative effect on the retinal microvasculature^(22,27).

A meta-analysis of data from more than 41.000 patients found a two-fold increase of late AMD risk in patients with CRP higher than 3 mg/L when compared to those with serum levels <1mg/L⁽²⁸⁾. The effect of CRP levels on AMD progression might be due to chronic inflammation leading to oxidative damage, endothelial dysfunction, drusen development and the degeneration of Bruch’s membrane⁽²⁹⁾. Additionally, CRP may also have a direct role in AMD development through its ability to induce complement activation⁽³⁰⁾.

IL-6 is a marker for systemic inflammation, such as acute pancreatitis and chronic arthritis. It has been implicated in angiogenesis, along with VEGF, in several in vitro and cancer studies⁽³¹⁾. Seddon and colleagues found a correlation between the level of IL-6 and chances of AMD progression⁽³²⁾. Animal studies have shown that CNV induction by laser stimulated IL-6 expression in the RPE–choroid complex, and that blockade of IL-6 led to a significant suppression of CNV⁽³³⁾. Interestingly, high aqueous IL-6 levels might to predict resistance to intravitreal bevacizumab in patients with wet AMD⁽³⁴⁾.

Fibrinogen is also an established biomarker of acute and chronic inflammation⁽³⁵⁾. Lip and colleagues found elevated levels of plasma fibrinogen in AMD cases compared with controls⁽³⁶⁾, and a case–control analysis from the large Blue Mountains Eye Study in Australia detected significantly elevated plasma fibrinogen levels in AMD patients compared with controls (p < 0.05)⁽³⁷⁾. In another study using patients recruited from the Muenster Aging and Retina Study population in Münster (Germany), plasma fibrinogen levels were found to be elevated as the degree of AMD severity increased⁽³⁸⁾. In fact, fibrinogen plasma levels appear to be particularly increased in patients with exudative AMD⁽³⁹⁾.

 

Multimodal imaging evaluation of AMD progression

1) Fundoscopy and Colour Fundus Photographs

Despite the increasingly sophisticated imaging techniques available, fundoscopy and colour fundus photographs (CFP) remain invaluable tools for the assessment of the severity and risk of progression of patients with AMD.

Drusen size, area and location are important predictors of AMD progression^(40,41): large soft drusens with ill-defined borders and extensive drusen area are the most important risk factors for the progression to advanced AMD identifiable on CFP^(42,43). Yet, earlier macular changes can provide valuable information about the risk of AMD progression. This has been extensively demonstrated in population-based observational studies and provides the basis for several AMD severity classifications.

The Age-Related Eye Disease Study (AREDS) collected stereoscopic CFP of more than 3000 AREDS over 5 years of follow-up and identified large drusen and macular pigmentary abnormalities as features that, cumulatively, increase the risk of progression to advanced AMD⁽⁴³⁾. A simplified risk scale was developed based on these findings (Table 1). It considers the presence of at least 1 large drusen (diameter greater than or equal to that of a large vein at the disc margin) and the presence of any pigment abnormality as 1 point each, and sums their presence across both eyes when both are free of advanced AMD. One point is assigned to patients who have no large drusen in either eye but intermediate-sized drusen (diameter ≥ one half that of a large drusen) in both eyes. The 5-year risk of advanced AMD using this scale increases as more risk factors are present: 0 factors, 0.5%; 1 factor, 3%; 2 factors, 12%; 3 factors, 25%; and 4 factors, 50%. This simplified scale is particularly useful in clinical practice because it allows fast risk stratification and is useful when discussing with patients their risk of progression to vision-threatening AMD.

 

In recent years, several studies have studied the significance of reticular pseudodrusen (RPD). They were first described in 1990 as a yellowish interlacing network most commonly located on the superotemporal macula⁽⁴⁴⁾. RPD are difficult to identify in clinical fundoscopic examination and are poorly visualized with regular CFP, so they are best seen with multimodal imaging⁽⁴⁵⁾. In two population-based studies, the overall prevalence of RPD was 0.7%-1.95% for the general population^(45,46). In AMD patients, the prevalence of RPD is significantly higher: they are present in up to 58% of patients^(47,48). This finding is particularly relevant due to the fact that RPD are associated with greater risk of progression to advanced AMD: the presence of RPD at baseline in fellow eyes of patients with unilateral neovascularization is a significant predictor of progression to advanced AMD (OR =2.5) over a 5-year period, particularly the wet form⁽⁴⁸⁾. Similarly, depending on the follow-up, up to 14% of patients with AMD ultimately develop GA^(47-49). Unfortunately, RPD presence remains an underreported and underresearched retinal phenotype, but the advent of new imaging modalities will certainly improve our knowledge about this phenotype⁽⁴⁸⁾.

 

2) Fundus Autofluorescence

Fundus autofluorescence (FAF) is a non-invasive method that supplies additional information to that obtained using CFP and fluorescence angiography⁽⁵⁰⁾. It provides an indirect evaluation of RPE function, an important component of the physiopathology of AMD⁽⁵¹⁾. In GA, the atrophic areas appear as hypoautofluorescent patches surrounded by non-atrophic retina of variable autofluorescence⁽⁵²⁾. The FAF features of these borders of GA lesions appear to be particularly important to predict the progression of pre-existing GA, as variation in GA growth rates are dependent on the specific phenotype of FAF at baseline^(53-55).

Additionally, FAF has a great sensitivity for the detection of RPD: areas covered with RPD usually show a reticular FAF pattern with small areas of decreased FAF surrounded by normal FAF⁽⁵⁵⁾.

For further data on FAF, please refer to the chapter “Fundus autofluorescence in age-related macular degeneration”.

 

3) Optical Coherence Tomography

Due to the quantitative and highly reproducible data that it provides, optical coherence tomography (OCT) has quickly become an invaluable tool to evaluate AMD progression and treatment response.

Spectral domain OCT (SD-OCT), in particular, enables automated measurement of drusen area and volume⁽⁴¹⁾, important predictors of AMD progression⁽⁵⁶⁾, and has been validated as a method of following GA development and progression⁽⁵⁷⁾. In addition, RPD can be easily seen on SD-OCT B-scans as triangular shaped structures located above the RPE line⁽⁵⁸⁾.

Detailed information about the role of OCT in the management of AMD can be found on the chapter “Optical Coherence Tomography in Age-related Macular Degeneration”.

 

Conclusion

AMD is a complex disease caused by a combination of genetic predisposition and environmental factors and the interplay between these remains a mystery. A better understanding of the involved pathophysiologic process or the identification of biomarkers for the conversion would enhance our ability to diagnose and treat the wet AMD, to develop better therapies and, eventually, to prevent vision loss associated with the disease.

 

>> References

Introduction

Fluorescein angiography (FA) was introduced in ophthalmology by Novotny and Alvis in the sixties of the last century.

They took serial fundus photographs after intravenous injection of sodium fluorescein to study the retinal and choroidal circulation⁽¹⁾.

Initially, they used this technique in diabetic and hypertensive patients and after, the technique was used in age-related macular degeneration (AMD).

Although the clinical diagnosis of AMD can be established based on patient’s history and fundus examination, FA is the most important ancillary test for classifying the disease in its different subtypes, especially in its wet form.

Nowadays, optical coherence tomography (OCT) is being more used than FA for monitoring the response to treatment, although FA is still very useful in some cases. Sodium fluorescein is a small molecule, with a molecular weight of 376.27 daltons, and it is highly soluble in water.

It is stimulated by light in the range of 465-490 nm and then it enters into a higher energy state.

The molecule emits longer wavelength fluorescence, between 520 and 530 nm, as it decays to a lower energy state.

In clinical application, two filters are used.

An excitation filter allows the passage of blue light, which stimulates the fluorescein in the eye, which emits yellow-green light.

In addition, a barrier filter is used to block some reflected blue light, allowing only the yellow-green light to pass through.

This resultant fluorescence is recorded by a camera as an image⁽²⁾.

Sodium fluorescein diffuses through the fenestrated vessels of the choriocapillaris, but does not cross the internal and the external blood-retinal barriers.

Thus, any condition that compromises these barriers, obstructs blood flow, or changes the normal pigmentation of the retinal pigment epithelium (RPE) can cause abnormalities on FA.

The dye is removed from the vascular compartment by the kidney.

It is relatively inert, making intravenous injection safe and severe adverse reactions rare. Nevertheless, the patient should be properly informed of the potential risks of FA injection⁽³⁾.

 

Interpretation

Several FA patterns can be observed in AMD patients.

They can be classified at those leading to decreased fluorescence (hypofluorescence) or increased fluorescence (hyperfluorescence).

Hypofluorescence either represents blocked fluorescence or a vascular filling defect (Table 1).

Hypofluorescence

Blocked

Intraretinal or subretinal hemorrhage/exudate Sub-RPE hemorrhage Pigment proliferation Pigment epithelial clumping (RPE rip)

Vascular filling defect

Choroidal vascular atrophy Retinal capillary occlusion

Table 1 - Causes of hypofluorescence.

 

Hyperfluorescence can be the result of loss of the normal barrier to background choroidal fluorescence known as transmitted fluorescence.

A second reason for hyperfluorescence can arise from extravascular accumulation of the dye or from leakage from abnormal vessels (Table 2)⁽⁴⁾.

Hyperfluorescence

Transmitted fluorescence

RPE atrophy RPE rip Hard, basal laminar drusen

Extravascular fluorescence

Serous pigment epithelial detachment Soft drusen Disciform scar Loculated fluid Cystoid macular edema

Abnormal vessels

Choroidal neovascularization Retinal angiomatous proliferation

Table 2 - Causes of hypofluorescence.

 

By convention, leakage of fluorescein into a space is referred to as pooling, while leakage into a tissue is called staining⁽⁵⁾.

 

Angiographic patterns in AMD

Drusen and RPE abnormalities

The majority of patients with AMD have drusen and RPE abnormalities with no significant visual loss.

FA is not usually indicated in these cases unless we suspect the presence of choroidal neovascularization (CNV).

Several types of drusen can be identified. Hard drusen are small (<63 µm), round, well-defined, yellowish deposits that correspond to accumulation of hyaline material in the inner and outer collagenous zones of Bruch’s membrane.

On FA, they appear hyperfluorescent as transmission defects due to overlying RPE thinning⁽⁶⁾.

On occasion there may be a myriad of small drusen, termed cuticular or basal laminar drusen, which appear as a “starry sky” on FA (Figure 1).

Figure 1 - Cuticular drusen with the typical pattern of “starry sky”

 

Soft drusen are larger (>63 µm) with poorly defined borders and they tend to coalesce and become confluent.

Their angiographic appearance depends on the thinning of the overlying RPE, the histochemical composition and the age of the patient.

They are hyperfluorescent with phospholipid accumulation and in younger patients⁽⁷⁾.

Soft drusen represent localized detachments of the RPE.

It is very usual to find both hard and soft drusen in the same eye of a patient (Figure 2).

Figure 2 - Coexistence of hard and soft drusen in the same eye of a patient with AMD.

 

The confluence of soft drusen can produce a drusenoid pigment epithelial detachment (PED), which shows hyperfluorescence and dye pooling without leakage beyond its margin with typical areas of focal hyperpigmentation (Figure 3).

Figure 3 - Drusenoid pigment epithelial detachment.

 

In addition to drusen we can find RPE abnormalities, namely hyperpigmentation.

Focal hyperpigmentation is a risk factor for the development of choroidal neovascularization (CNV) and angiographically appears as a blocked fluorescence (Figure 4).

Figure 4 - RPE abnormalities with focal hyperpigmentation.

 

Histopathologically it is characterized by focal RPE hypertrophy and pigment migration into the subretinal space.

It also displays focal hyperautofluorescence suggesting that these cells contain lipofuscin⁽⁸⁾.

 

Atrophic AMD

Age-related macular degeneration (AMD), is a leading cause of severe central vision loss and legal blindness among people over 55 years of age in the industrialized countries ^(1-4), and thus represents a major worldwide sociosanitary problem.

Exudative (neovascular) AMD is a less common form of the disease, but is responsible for approximately 80-90% of all cases of severe vision loss ^(5-7). Choroidal neovascularization (CNV) in its different forms of presentation is the main cause of complications and vision loss in these patients.

The classification of CNV based on its angiographic characteristics was regarded as very important in the past, when treatment fundamentally consisted on thermal laser ablation or photodynamic therapy (PDT). However, at present, the development of spectral domain optical coherence tomography (SD-OCT) and the use of intravitreal injections of antiangiogenic drugs have led to a genuine revolution in the diagnosis, follow-up and prognosis of these patients.

Anti-vascular endothelial growth factor (VEGF) therapy is the treatment of choice in patients with exudative AMD, though there are cases where other treatments may be used alone or in combination in order to increase efficiency.

OCT has displaced other diagnostic techniques used for patient control in the frequent follow-up visits of the disease. Today, SD-OCT and noninvasive fundus autofluorescence imaging can provide additional anatomical information beyond the fluorescein angiography (FA), including visualization of the layers of the retina and choroid involved with the neovascular membrane. Moreover, swept source OCT (SS-OCT) is a new OCT technology, which have enabled a even faster acquisition and a reduction in the signal-to-noise ratio. The use of OCT has increased dramatically during the past decade, whereas the use of FA had declined considerably. 

In addition, the more recent introduction of OCT angiography (OCTA) and the increasing experience in interpreting OCT images has reduced even further the need for FA and its associated side effects.

FA remains however, in our opinion, a necessary technique at the time of diagnosis, since it affords prognostic information, and is useful for re-evaluating patients with a poor response to treatment. In fact, the American Academy of Ophthalmology AMD Preferred Practice Pattern advocates the use of intravenous fundus FA when a patient complains of a new unexplained blurred vision and it is also indicated in the following situations: to detect the presence and determine the extent, type, size, and location of CNV; and, to detect persistent or recurrent CNV following treatment⁽⁹⁾. 

Nowadays, the retina and imaging community is excited to see if improvements in OCT technology and the introduction of the OCTA could eliminate the FA⁽⁹⁾.

Indocyanine green angiography (ICGA) is useful for the study of occult CNV (OCNV), and particularly in identifying the characteristic patterns of idiopathic polypoidal choroidal vasculopathy (IPCV) and retinal angiomatous proliferation (RAP). It is important to correctly diagnose these disorders, since they respond differently to treatment. The main advantage of ICGA is that it allows specific analysis of choroidal circulation and its interactions with the retina. However, interpretation of the ICGA images is complex, and in some cases requires contrasting with the data obtained by means of other techniques such as FA or OCT, and with correct biomicroscopic assessment of the macula.

The present chapter analyses the information that we can obtain from ICGA in application to the diagnosis and follow-up of AMD, and defines the current most precise indications of the technique, and its usefulness in daily clinical practice.

Thus, an analysis is made of the characteristic ICGA images found in patients with AMD, the precise indications of the technique, and the situations in which ICGA proves essential for establishing a correct diagnosis to ensure adequate treatment of this complex disease.

 

INDOCYANINE GREEN ANGIOGRAPHY (ICGA)

Indocyanine green (ICG) is a tricarbocyanine dye with both hydrophilic and lipophilic properties⁽¹⁰⁾. The empirical formula of ICG is C₄₃H₄₇N₂NaO₆S₂. The molecular weight is 775 Daltons. When dissolved in saline solution, ICG tends to form polymers if the solution concentration is high and monomers if the concentration is low. For this reason the dye is dissolved in water for intravenous injections.

Peak absorption of the monomers and polymers is observed at 785 and 690 nm, respectively. Binding to proteins in turn elevates the absorption spectrum to 810 nm, since polymer formation decreases. The emission spectrum in aqueous solution is 810-820 nm, though in blood it shifts 10 nm towards longer wavelengths. The ICG molecule can be considered active at a wavelength of between 790 and 830 nm.

Following intravenous administration, ICG binds to blood lipoproteins (fundamentally phospholipids) in a proportion that can reach 98%^(11-13). The emitted fluorescence intensity increases as a result of such binding. Elimination takes place in the liver, with secretion in bile.

The adverse reactions are similar to those observed with fluorescein, but are less frequent⁽¹⁰⁾. In 1994, explorations with ICG documented a 0.15%, 0.2% and 0.05% incidence of mild, moderate and severe reactions, respectively⁽¹⁰⁾. The dye should not be used in patients with shellfish or iodine allergies, liver disease, and end-stage renal disease. No studies in animal models have been made regarding use of the dye in pregnancy. However, current practice patterns regarding the use of ICGA in pregnant woman may be unnecessarily restrictive⁽¹⁴⁾.

Exploration is carried out in the same way as in FA, though with the corresponding excitatory light and barrier filter. The contrast is injected and images are obtained documenting the fluorescence passing through the ocular fundus during the filling phases, and posteriorly after 1, 3, 5, 10, 20 and 30 minutes.

The ICG molecule has biophysical properties that make it useful for visualizing the choroidal circulation. The dye is excited by light close to infrared, increasing its penetration through the retinal pigment epithelium (RPE), melanin, xanthophyll pigment and thin blood layers. ICG circulates almost entirely bound to plasma proteins, and gradually diffuses across the choriocapillaris – thus allowing better visualization of its circulation within the choroidal vessels. These properties make it possible to visualize CNV beneath blood layers, exudates or RPE detachments (PEDs)^(15-16) by differentiating (in the latter case) between the hypofluorescent serous component and the hyperfluorescent vascular component (Figure 1).

Figure 1. (A) The ICG molecule has biophysical properties that make it useful for visualizing the choroidal circulation. (B, C and D) These properties make it possible to visualize choroidal neovessels beneath blood layers, exudates or RPE detachments (PEDs) by differentiating (in the latter case) between the hypofluorescent serous component and the hyperfluorescent vascular component. With FA, both the serous and the vascular component appear hyperfluorescent, and the limits of CNV may be masked as a result. A and B. Visualization of the choroidal circulation under thin blood layers. (C) Hot spots visible in an area of exudation. (D) PED with small CNV at the margin of the lesion. With FA, both the serous and the vascular component appear hyperfluorescent, while with ICGA only the vascular component of PED appears hyperfluorescent.

 

The most important contribution of ICGA to the study of CNV in AMD is its capacity to detect choroidal new vessels which in some cases remain occult when using FA^(15,16). Following the development of ICGA in the early 1970s^(17,18), two decades went by before scanning laser ophthalmoscopy (SLO) and digital angiography made it possible to improve ICGA visualization^(19-21)and registry, and routine application of the technique in clinical practice.

In the 1990s, different studies⁽²²⁾ showed that ICGA allows the visualization of OCNV, the most frequent type of CNV in the context of neovascular AMD. The capacity of ICGA to locate new vessels particularly in those cases with associated PEDs made it possible to increase the number of cases of CNV amenable to laser photocoagulation therapy – the latter being the only treatment option available at that time for such patients.

Until the end 1990s, the visualization and delimitation of neovascular membranes was regarded as very important, due to the inclusion of extra- and juxta-foveal locations as an indication for laser photocoagulation⁽²²⁾. In this context, the use of ICGA was of great help due to its capacity to delimit OCNV. In 1992, Yannuzzi et al.⁽²³⁾ showed that late frames of ICGA allowed visualization of the entire choroidal neovascular membrane^(22). Before this time, ICGA had already been found to be useful in detecting CNV in cases of block caused by media opacity^(24,25).

ICGA served to guide photocoagulation of OCNV without classic CNV, particularly when this OCNV was located inside or at the margin of PED^(26,27). The images obtained with this technique in such cases of vascularized PEDs allowed the individualization of a new form of CNV known as chorioretinal anastomosis, later termed retinal angiomatous proliferation (RAP) or type 3 neovascularization^(28-30). Scanning laser ophthalmoscopy (SLO), more than vid

In his description, Yannuzzi (58) defended this retinal onset of neovascularization, while Gass (38) suggested that the process starts in the choroid as occult CNV.

When FA is used, RAP manifests as an occult form of CNV with PED and perilesional microhemorrhage and frequent intraretinal edema, associated to neurosensory retinal detachment (NSRD) and RCA in advanced stages.

eoangiography, truly improved the quality of the early frames of ICGA, making it possible to record a dynamic sequence of choroidal filling and improving visualization of this particular neovascular network corresponding to chorioretinal anastomosis⁽²²⁾.

Using SLO, other authors also described the use of high-speed ICGA for visualizing the feeding vessels in cases of OCNV, allowing their selective photocoagulation without the need to cover the entire membrane^(31-33). However, the usefulness of ICGA in these cases is limited, since in many cases there is not only one but several feeding vessels; there may be failures in identification; and laser photocoagulation is often not enough to ensure definitive sealing of the afferent vessel⁽³⁴⁾.

The introduction of verteporfin photodynamic therapy (V-PDT) in the late 1990s allowed the treatment of subfoveal CNV. Different randomized clinical trials were performed, such as the TAP (Treatment of Age-Related Macular Degeneration with Photodynamic Therapy) and VIP (Verteporfin in Photodynamic Therapy) trials, which respectively, included patients with classic and occult CNV^(35,36). New vessels were classified according to the FA findings, while ICGA was not performed because PDT became standard care treatment for classic subfoveal CNV.

Anti-VEGF therapy inaugurated a new era in the management of exudative AMD, because it was the first treatment to improve the mean visual acuity of eyes treated with intravitreous injections of anti-VEGF^(37,38), which proved effective in both classic and occult subsets of CNV. In randomized clinical trials, CNV was classified according to the FA findings, and ICGA did not form part of the protocol. However, eyes with PEDs were not included in the initial randomized clinical trials, and it is known that such conditions show a poorer treatment response and predispose to RPE rupture⁽³⁹⁾.

New developments of OCT include better visualization of the choroid^(40-43), and this noninvasive technology reduced the use of angiographic testing in routine clinical practice in patients with exudative AMD in which fixed-regimen monthly intravitreous reinjection are used. OCT brings another classification of the neovascular type secondary to AMD: type I is a lesion which grows under the RPE, type II is subretinal and type III corresponds to RAP.

ICGA remains necessary when the diagnosis proves uncertain despite FA and OCT; in certain cases of OCNV with PEDs (particularly in cases of suspected IPCV and RAP) or conditions such as central serous chorioretinopathy (CSC), which may require a different therapeutic approach from that used in neovascular AMD; and in the re-evaluation of non-responding patients.

Due to the economical and medical care burden implied by this type of treatment, an early and correct diagnosis is very important in order to prescribe adequate patient management.

 

Current Importance of ICGA in AMD

Nonexudative (Dry) AMD

In nonexudative AMD, ICGA has been used to improve drusen classification^(44,45); to facilitate diagnosis of basal laminar drusen and the associated subretinal deposits⁽⁴⁶⁾; and to improve understanding of reticular pseudodrusen^(47,48), due to the affinity of some lesions for this dye.
However, ICGA is not routinely used to study dry AMD⁽²²⁾, except in cases of drusenoid PED to confirm the absence of CNV (Figure 2).

Drusenoid PED could be defined as the progressive increase and confluence of small soft drusen. These appear as yellow-gray spots, sometimes with a characteristic hyperpigmented ring, due to their slow and progressive growth. OCT in application to these drusenoid forms allows us to identify elevations of the pigment epithelium, sometimes variable in number and sometimes as a single and larger PED, filled with hyper-reflecting material.

Drusenoid PEDs require no treatment, since they do not imply the presence of neovascular activity, and ICGA is important for establishing this differentiation (Figure 2). 

Figure 2. Drusenoid PEDs require no treatment, since they do not imply the presence of neovascular activity, and ICGA is important for establishing this differentiation. Drusenoid PEDs appear hyperfluorescent in the fluorescein study, while ICGA allows the exclusion of CNV if there is no hyperfluorescent component associated to the hypo- or isofluorescence of the PED.

 

Recently, ICGA has been reported to be useful for evaluating the atrophic areas present in Stargardt disease, in comparison with the atrophy of dry AMD⁽⁴⁹⁾. The hypocyanescence evidenced by ICGA in the atrophic areas (“dark atrophy”) has been shown to be more frequent in Stargardt disease than in atrophic AMD – suggesting a possible selective damage of the choriocapillaris in Stargardt disease.

 

Exudative (Wet) AMD

Using ICGA, CNV has been reclassified into three different morphologic types: focal spot or “hot spot”, plaques (well or poorly defined), and mixed (i.e., a combination of the previous two)^(50,51). These alterations may be located at the margin of the lesion (marginal spot), above the lesion (overlying spot), or at a distance from the lesion (remote spot) (Figures 3-4).

Figure 3. Using ICGA, CNV has been reclassified as focal spots (or hot spots), plaques or mixed forms (a combination of the first two forms). Figure A shows a plaque with poorly defined margins, in the late phases of ICGA. Figure B shows two hot spots (the lower spot being more active).

 

 

Yannuzzi(23,60) reported that hypofluorescence is maintained throughout the duration of the test in non-vascularized PEDs. On analysing these changes in serous PED with CNV, this author established two well defined groups in 96% of the patients: (a) cases where ICGA shows a solitary and well defined hyperfluorescence spot which he refers to as focal CNV or a “hot spot”; and (b) a “plaque CNV”, which is a larger and poorly defined lesion with less intense fluorescence. In the remaining 4% of the cases the author observed no CNV with ICGA.

Figure 4. CNV may be located: (A) At the margin of the lesion (marginal spot); (B) Above the lesion (overlying spot); or (C) At a distance from the lesion (remote spot).

 

Plaques were found to be the most common type (61% of the cases) and had a poor visual prognosis, whereas the focal spots or “hot spots” (29%) had a better prognosis, and they were considered to be potentially treatable by ICG-guided laser photocoagulation^(50,52).
Moreover, ICGA is useful to identify underlying causes of spontaneous submacular hemorrage, allowing for different approaches that may improve outcomes and safety⁽⁵³⁾.

 

Classic CNV

Classic CNV is better delimited by FA than by ICGA^(54,55). ICGA does not offer relevant supplementary information unless associated OCNV is present. As a result, it is likewise not routinely used to study classic CNV.

 

Occult CNV (OCNV)

The term “occult” refers to a type of CNV that is difficult to visualize, analyse and localize on FA. This difficulty is due to their polymorphic features, but especially to the fact that they are ill-defined and ill-delimited⁽⁵⁶⁾. The fundamental advantage of ICGA is that it allows early detection and localization of OCNV and analysis of the filling and staining pattern of the CNV network of the lesions.

OCNV is very common, and most cases of CNV in AMD (60-85%) correspond to this type^(57,58).

These cases may present a variable course:

• A slow and almost asymptomatic course for many months.
• A course characterized by exacerbation episodes.
• A slowly progressing course with impairment of visual acuity, metamorphopsia and gradual growth of the lesion.

The natural outcome of the disease is generally poor, with severe central vision loss due to hemorrhages, sometimes associated with serous PEDs and/or tearing of the pigment epithelium (RPE Tear), and eventually, fibrous proliferation with formation of a fibrous scar.

OCNV can be grouped into two main types⁽⁵⁶⁾:

1. OCNV without serous PED;
2. OCNV associated with serous PED.

The classification proposed by the MPS (Macular Photocoagulation Study) investigators is similar and based on FA:

• Type I: Fibro-vascular PED (Fibro-vascular pigment epithelium detachment with irregular and poorly demarcated hyperfluorescence);
• Type II: Late leakage of an undetermined source (without PED and without classic CNV).

RAP or type 3 neovascularization in turn is added to the above^(29,30).

The importance of ICGA in the study of OCNV is that it can afford an early diagnosis of the disease in cases that prove doubtful with FA and OCT particularly cases involving OCT based differential diagnosis with neurosensory detachment, which can also occur in CSC.
 

1. OCNV without serous PED

FA shows irregular and poorly delimited hyperfluorescence with late phase diffusion suggestive of OCNV. OCT in these cases is essential for evaluating the degree of activity due to the presence of neurosensory detachment and thickening of the pigment epithelium membrane – choriocapillary band, which is more suggestive of OCNV than of CSC in the absence of leakage points in FA. OCT is also fundamental for the follow-up of these patients.

2. OCNV with serous PED

Serous or serohematic PED appears in the ocular fundus as a yellow-orange elevation with margins that are very well defined by the solid adherence of the pigment epithelium to Bruch’s membrane at this level.
Serous PED may be avascular or vascular. For all eyes with PEDs, the purpose of the initial evaluation is to identify or exclude the presence of CNV, as this feature will change the treatment in the anti-VEGF era. ICGA is very useful in these cases for establishing the presence or absence of associated neovascularization, since such CNV may be masked by the hyperfluorescence of PED when FA is used (Figure 5).

Figure 5. ICGA is very useful in these cases for establishing the presence or absence of associated neovascularization, since CNV may be masked by the hyperfluorescence of PED when FA is used. When FA proves inconclusive, ICGA may confirm the absence of neovascularization.

 

When ICGA is used, PED appears as a hypofluorescent lesion in all phases of the angiogram. In some cases, as a result of the accumulation of lipids on Bruch’s membrane, the lesion may have an isofluorescent or even somewhat hyperfluorescent appearance, while in contrast CNV always appear hyperfluorescent. Gass⁽⁵⁹⁾ described a blood or pigment meniscus in a zone of the PED, forming a notch that affords a kidney-like shape and which represents the zone of firm adherence between the CNV and the pigment epithelium. This notch which gives the PED a kidney shape allows us to suspect where the CNV is located.

Yannuzzi^(23,60) reported that hypofluorescence is maintained throughout the duration of the test in non-vascularized PEDs. On analysing these changes in serous PED with CNV, this author established two well defined groups in 96% of the patients: (a) cases where ICGA shows a solitary and well defined hyperfluorescence spot which he refers to as focal CNV or a “hot spot”; and (b) a “plaque CNV”, which is a larger and poorly defined lesion with less intense fluorescence. In the remaining 4% of the cases the author observed no CNV with ICGA.

ICGA is a very useful diagnostic technique and superior to FA in diagnosing and classifying CNV associated to serous PEDs. As can be seen in the figures, all types of neovascularization can be found associated to serous PED: classic, occult, RAP, and also quite often associated to IPCV (as shown in the Figure 6).

Figure 6. ICGA is very useful for diagnosing and classifying CNV associated to serous PED, due to its capacity to distinguish between the serous and the vascular component. All types of associated CNV may be found. Figure A shows classic CNV which is better visualized with FA in the early phases, since it is masked in the late phases by the hyperfluorescence of the entire lesion. The rest of the cases involve OCNV, where ICGA can show hot spots on the black image of the PED associated to RAP (yellow arrows) or polyps (red arrows).

 

OCTA used to enhance the detection of neovascularization could be possible in some cases where neovascularization is not clearly detected with OCT or dye-based angiographies. Nowadays, it could be interesting to use a multimodal approach in these cases⁽⁶¹⁾.

 

Retinal Angiomatous Proliferation (RAP)

RAP has been described as a variant of exudative AMD characterized by the presumed retinal origin of CNV⁽⁶²⁾ (Figure 7). 

Figure 7. RAP has been described as a variant of exudative AMD characterized by the presumed retinal origin of CNV. (A) Image of retinal communication under Zeiss FF 450 plus IR (Carl Zeiss Meditec, Jena, Germany). (B) HRA (Heidelberg Retinal Angiography 2) image (HRA2, Heidelberg Engineering, Germany).

 

The pathogenesis and classification of RAP remain subject to debate⁽⁶³⁾.

Yannuzzi ⁽⁶²⁾ was the first to describe this condition, establishing three different stages on the basis of the clinical and angiographic findings:

• STAGE I. Proliferation of intraretinal capillaries; intraretinal neovascularization;
• STAGE II. Subretinal new vessels without PED (IIA) or with PED (IIB);
• STAGE III. Choroidal neovascularization (CNV);
• Such CNV often consolidates as retinochoroidal anastomosis (RCA), which posteriorly has been classified as:
• STAGE IV. Retinochoroidal anastomosis (RCA) ^(64,65).

RAP represents approximately 12-20% of all cases of exudative AMD ^(30,66,67), and currently the term type III neovascularization has been proposed for this condition, precisely in order to distinguish it from other types of CNV in exudative AMD ^(28,65,68).

This disorder is known to show a different natural course and response to treatment compared with other forms of AMD^(69,70,71). The existence of retino-retinal anastomosis (RRA) and RCA(62,64) is characteristic of RAP type lesions and contributes to the presence of high-flow vascular alterations with double retinal and choroidal circulation. This and the presence of associated PED and cystoid macular oedema in the more advanced stages complicates the treatment of this disease⁽⁷²⁾ (Figure 8).

Figure 8. The existence of RRA is characteristic of RAP type lesions and contributes to the presence of high-flow vascular alterations with double retinal and choroidal circulation. This, and the presence of associated PED and cystoid macular edema, complicates the treatment of this disease.

 

In some cases, ICGA in the early phases allows us to identify RRA with a nutrient vessel and a draining vessel⁽⁷³⁾ – this representing a retinal neoangiogenic process that subsequently spreads to the subretinal space (Figure 8).

In some cases, ICGA in the early phases allows us to identify retino-retinal anastomosis with a nutrient vessel and a draining vessel ⁽⁶⁹⁾ – this representing a retinal neoangiogenic process that subsequently spreads to the subretinal space (Fig.8).

In his description, Yannuzzi⁽⁶²⁾ defended this retinal onset of neovascularization, while Gass⁽⁴⁰⁾ suggested that the process starts in the choroid as OCNV.
When FA is used, RAP manifests as an occult form of CNV with PED and perilesional microhemorrhage and frequent intraretinal edema, associated to neurosensory detachment and RCA in advanced stages.

ICGA allows us to study the existence of single or multiple hot spots at neovascularization level. This is very useful in the earliest stages, especially to the effects of treatment, if V-PDT is to be associated. ICGA also allows us to observe spot persistence or disappearance after treatment, as well as the reappearance of spots in the case of patients who are re-evaluated due to a lack of response or resistance to therapy. In some early stage cases, ICGA can reveal the characteristic retino-retinal vascular communication in the early phases of the angiogram (Figure 9).

Figure 9. ICGA allows us to study the existence of single (A and C) or multiple (B) hot spots at neovascularization level (white arrows). In some early stage cases, ICGA can reveal the characteristic retino-retinal vascular communication in the early phases of the angiogram (red arrows).

 

It is important to distinguish RAP from other forms of AMD, not only in view of the poor vision prognosis if exhaustive patient follow-up and correct treatment are not carried out, but also due to the high risk of neovascularization in the other eye^(70,74,75). In this context, ICGA helps us to establish a correct diagnosis (Figure 10).

Figure 10. It is important to distinguish RAP from other forms of AMD due to the high risk of neovascularization in the other eye when the disease is present in one eye. (A) RAP first manifesting in the left eye with spontaneous rupture of the RPE; we must keep a very close follow-up because of a possible involvement of the fellow eye. (B) After 6 months the fellow eye is affected, and ICGA helps us to establish a correct early diagnosis. We can see the communication in early phases and the small hot spot in late phases of ICGA.

 

The precise mechanism giving rise to this peculiar form of vascularization remains unclear, though some authors recently have suggested that there is ischemia at external retinal level, induced by diminished choroidal perfusion^(76-78).

RAP requires more intensive management, often involving combined treatments, in order to control the activity of the disease^{(70, 79-82)}. Recently, favourable results have been published with the use of anti-VEGF drugs, though with a high recurrence rate after suspending the treatment^(63,83-97).

V-PDT has been associated to anti-VEGF drugs and to intravitreous triamcinolone (IVT), with promising results^(98-114). Combined treatment proves effective in maintaining or improving patient vision and in reducing exudation, without long-term adverse effects⁽¹¹⁰⁾. Rouvas et al.⁽¹¹²⁾ have reported stabilization in patients administered with ranibizumab alone and in patients administered with ranibizumab associated to V-PDT, with functional and anatomical improvements in the patients subjected to V-PDT with IVT.

Regarding the combined treatments, it must be taken into account that there is an increased risk of developing atrophy in patients with RAP subjected to V-PDT(112). Frequent geographic atrophy has been described in these patients after the treatment⁽⁷⁷⁾, and recently there have also been reports⁽⁷⁸⁾ of specific changes in the choroid and in drusen distribution, which may have long term diagnostic and therapeutic implications (Figure 11).

Figure 11. Regarding the combined treatments, it must be taken into account that there is an increased risk of developing atrophy in patients with RAP subjected to V-PDT. Frequent geographic atrophy has been described in these patients after the treatment (white arrow).

 

In coincidence with other authors(111), we perform ICGA-guided PDT in some patients in stages I and II, with the smallest possible spot covering the neovascularization, associated to anti-VEGF drugs. This technique would use smaller spots, thereby contributing to lessen the risk of long-term atrophy in these patients (Figure 12). 

Figure 12. We perform ICGA-guided PDT in some patients in stages I and II, with the smallest possible spot covering the neovascularization, associated to anti-VEGF drugs. This technique would use smaller spots, thereby contributing to lessen the risk of long-term atrophy in these patients. The figure shows the outcome, three years after the treatment.

 

It is also known that the prognosis is poorer in cases of large PEDs, where pigment epithelial rupture may be more frequent after treatment^(115-117) (Figure 13), and that better functional results are obtained when these patients are diagnosed and treated in early stages of the process⁽¹¹⁸⁾. The diagnosis should be, therefore, established as soon as possible, and ICGA is very important to this effect (Figure 14).

Figure 13. It is also known that the prognosis is poorer in cases of RAP (A, red arrow) with large PEDs, where pigment epithelial rupture may be more frequent after treatment. (A) PED is observed/marked with a thin wall (A, white arrow) where pigment epithelial rupture occurs after intravitreous injection of ranibizumab (B, white arrow).

 

Figure 14. The diagnosis therefore should be established as soon as possible, and ICGA is very important to this effect. (A) Stage I RAP is observed, easily identifiable with ICGA. (B) A patient with bilateral drusen where ICGA identifies an incipient communication (B1, white arrow) that produces alterations with intraretinal edema as evidenced by OCT (B1); ICGA proves negative in the other eye, however (B2).

 

OCT may be used to diagnose type 3 neovascularization in patients with neovascular AMD with relatively high concordance compared with ICGA-based diagnosis⁽¹¹⁹⁾.

 

Idiopathic Polypoidal Choroidal Vasculopathy (IPCV) 

IPCV is characterized by the presence of a network of abnormal choroidal vessels with aneurysmal dilatations in the form of polyps visualized as red-orange nodules in the ocular fundus and located mainly at peripapillary level and in the macular zone^(120-123). These choroidal vascular lesions can produce serous exudation and hemorrhage that have been associated to recurrent serohematic PEDs and neurosensory detachment^{(122,124-128)} (Figure 15). 

Figure 15. (A and B) Peripapillary polyps with a large exudative component at superior temporal arcade level in the first patient. (C) Serohemorrhagic PED at macular level, with a small polyp, evidenced by ICGA (C, white arrow). (D) Circinate macular lesion centered by a network of polyps (D, white arrows).

 

The vascular lesions of IPCV can simulate an occult or minimally classic membrane with FA, while ICGA clearly reveals the polypoidal lesions and anomalous vascular network within the choroid. In concordance with this, a literature review published by Stanga et al. concluded that ICGA is highly recommended for the identification of choroidal lesions in patients with IPCV⁽⁵¹) (Figure 16).

Figure 16. Patients with OCNV in FA, where the polyps are seen by ICGA. The three patients responded poorly to anti-VEGF therapy and where therefore re-evaluated due to suspected IPCV. In case C, FA reveals a hyperfluorescent vascular PED (yellow arrow), while only the polyps at the margin of the PED appear hyperfluorescent with ICGA (white arrow).

 

ICGA is able to identify two elements in these anomalies: polypoidal structures projecting from the internal choroid towards the retina, and a branched choroidal vascular network (BVN) manifesting with early vascular hyperfluorescence^(129-131). The underlying pathogenesis remains subject to debate (Figure 17). In this sense, some authors suggest that the polyps are a form of CNV^(132-135), while others regard them as an alteration of the internal choroidal vessels^(136-138).

Figure 17. ICGA is able to identify two elements in these anomalies: polypoidal structures projecting from the internal choroid towards the retina, and a BVN manifesting with early vascular hyperfluorescence. The figure shows this choroidal vascular network (short white arrows), with the formation of a polyp at marginal level (large white arrow).

 

The polyps are difficult to identify on FA, as they are located beneath the RPE. On ICGA, however, the polyps are clearly visible as single or multiple vascular aneurysms with a diameter of 100-500 μm which fill after a short delay and remain hyperfluorescent until the late phases. These lesions are often located in the juxtapapillary zone, but can also be found at macular level, in the vascular arches, and even at peripheral level. The choroidal vessels in the area of the polyps may appear irregular and dilated, forming a fine vascular network. The angiographic sequence would be as follows: filling of the fine choroidal vessels and polyps with early choroidal vascular hyperfluorescence when the dye flows through the network of anomalous vessels. This fluorescence is maintained over time and disappears as the dye washes out in the late phases, in the presence of polypoidal lesions without active leakage (Figure 18).

Figure 18. Patient diagnosed with IPCV subjected to laser treatment in the macular area. Scantly active polyps persist at inferior temporal arcade level, evidencing dye washout in the last phases of the angiogram. The angiographic sequence would be as follows: filling of the fine choroidal vessels and polyps with early choroidal vascular hyperfluorescence when the dye flows through the network of anomalous vessels. This fluorescence is maintained over time and disappears as the dye washes out in the late phases, in the presence of polypoidal lesions without active leakage.

 

Guidelines have recently been published for the clinical diagnosis and treatment of IPCV⁽¹³⁹⁾, in which ICGA is regarded as the gold standard for the diagnosis of this disorder. This panel of experts considers that such angiography should be performed when routine ophthalmoscopic examination indicates a serosanguineous maculopathy with one of  the following features:

• Red-orange subretinal nodules visible with the ophthalmoscope.
• Spontaneous massive subretinal hemorrhage.
• Notched or hemorrhagic PED.
• Occlusion of polyps with lowresponse rate to anti-VEGF therapy.

In conclusion, ICGA is the gold standard for diagnosis of IPCV, which is defined as the presence of single or multiple focal nodular areas of hyperfluorescence arising from the choroidal circulation within the first 6 minutes after injection of ICG, with or without an associated BVN⁽¹³⁹⁾.

These authors propose algorithms for both the diagnosis and the treatment of the disease. IPCV may be classified clinically as: quiescent (polyps in the absence of subretinal or intraretinal fluid or haemorrhage), exudative (exudation without haemorrhage, which includes variously sensory retinal thickening, neurosensory detachment, PED, and subretinal lipid exudation) or hemorrhagic (any subretinal or sub-RPR haemorrhage with or without other exudative characteristics). Treatment should be initiated for active and symptomatic IPCV, and can be considered for active asymptomatic presentations (Figure 19).

Figure 19. Treatment should be initiated for active and symptomatic IPCV, and can be considered for active asymptomatic presentations. The figure shows both eyes of the same patient. (A) The patient is symptomatic, with a visual acuity of 0.6 in the right eye. The polypoidal lesions can be observed with ICGA, and retinal thickening and intraretinal cysts can be seen by OCT. (B) The patient maintains a visual acuity of 1.0 in the left eye. ICGA shows a peripapillary polyp, likewise visible in the OCT acquisition, but there is no retinal edema, and the patient therefore remains asymptomatic.

 

ICGA is necessary not only for the diagnosis but also for continuous monitoring of the disease, which may present polyp relapse or the appearance of new polyps from the BVN, with phases of quiescence, exudation or hemorrhage (Figure 20). 

Figure 20. ICGA is necessary not only for the diagnosis but also for continuous monitoring of the disease, which may present polyp relapse or the appearance of new polyps from the BVN, with phases of quiescence, exudation or hemorrhage. (A) Late phases of FA showing the hyperfluorescent PEDs. (B) ICGA image. (B1 and B2) Only the PEDs are seen, and the BVN alteration is noted in early phases with ICGA. (B3 and B4) Formation of polyps from the BVN, visible with ICGA 6 months later.

 

Focal photocoagulation with ICGA-guided thermal laser still plays a role in the treatment of extrafoveal polyps, but it is not recommended in application to juxta- or subfoveal lesions(139,140). In these cases, PDT has been the treatment of choice, with maintenance or improvement of patient visual acuity in 80-95% of the cases(141-146) (Figure 21).

Figure 21. Focal photocoagulation with ICGA-guided thermal laser still plays a role in the treatment of extrafoveal polyps, but it is not recommended in application to juxta- or subfoveal lesions. (B) Resolution of the polyps with atrophy scarring secondary to thermal laser in extrafoveal polyps at inferior temporal arcade level and after PDT at macular level. OCT shows resolution of the retinal thickening, with disappearance of the hard exudate.

 

However, long-term follow-up studies have described frequent recurrences after PDT^{(142, 147)}. In some cases PDT is unable to fully occlude the polyps, though in other cases persistence of the BVN can give rise to new active polypoidal lesions. As a result, periodic controls with ICGA are needed during follow-up to determine whether there are persistent or recurrent lesions with polyps or BVN.

Although recurrence and retreatments do not seem to decrease best corrected visual acuity, long-term studies are still needed in order to evaluate the efficacy and safety of PDT treatment of IPVC⁽¹⁴²⁾. PDT can modified the natural progression of the polypoidal CNV as a result of ischemia and increased VEGF expression⁽¹⁴⁸⁾.

After the introduction of anti-VEGF therapy, different studies reported the efficacy of combined treatment with PDT and anti-VEGF agents for IPCV^(149,150), as well as the way to minimize the complications associated with the use of PDT, and the need to use selective(151) or with reduced fluence PDT⁽¹⁵²⁾.

The Everest study⁽¹⁵³⁾ concluded that V-PDT combined with ranibizumab  0.5 mg or alone was superior to ranibizumab monotherapy. In our own experience with long-term patient follow-up⁽¹⁵⁴⁾, the association of ICGA-guided PDT for the occlusive therapy of polyps and intravitreous ranibizumab is effective in prolonging the recurrence-free interval, as well as in maintaining good visual acuity (Figure 22).

Figure 22. The association of ICGA-guided PDT for the occlusive therapy of polyps and intravitreous ranibizumab is effective in prolonging the recurrence-free interval, as well as in maintaining good visual acuity. Patient with a single juxtapapillary polyp associated to a large PED at macular level. Three PDT sessions are performed (A, B and C), without resolution of the condition. Indeed, worsening is noted, with the appearance of neurosensory detachment of the fovea (C). At the time where the polyp is not visualized by ICGA, and since the PED persists, three intravitreous injections of ranibizumab are administered, followed by resolution of the condition (E). The patient remained stable, with a visual acuity of 20/20 until month 36 of follow up.

 

Although the efficacy and safety of PDT and ranibizumab versus PDT and bevacizumab remain to be established, the experts have concluded⁽²³⁾ that PDT with standard fluence guided by ICGA with or without the combination of three intravitreous injections of ranibizumab 0.5 mg on a monthly basis is the most widely recommended treatment for this disease. ICGA-guided thermal laser photocoagulation can be considered for extrafoveal polyps.

The use of ICGA is very important for guiding laser treatment and PDT, both as regards the indication of such treatments and to minimize the spot size. This in turn contributes to limit as far as possible the risk of atrophy and other side effects derived from the multiple treatments usually required in the long-term follow-up of these patients.
A study with ICGA may in some cases help to unmask patients purportedly refractory to anti-VEGF therapy⁽¹⁵⁵⁾, because it allows a precise diagnosis and adequate treatment indication.

ICGA will retain a key role in the management of PCV because of its ability to better identify polyps although SS-OCTA is effective at detecting vascular network⁽¹⁵⁶⁾. However, SS-OCTA may be more sensitive than SD-OCT in detecting early recurrence.

On the other hand, the combination of En Face and cross-sectional OCTA images provides anatomical information about polypoidal structures that is comparable to ICGA. OCTA is safe and can provide supplementary information but due to its limitations it does not replace ICGA in a clinical setting nowadays^(157-159).

 

>> References

Definitions

Fluorescence is the capability of absorbing light at a specific wavelength and releasing it at a longer, less energetic wavelength. This phenomenon raises an especial interest when the released radiation is found within the spectrum of visible light permitting its visualization, recording and measurement.

Autofluorescence is the spontaneous fluorescence that some substances present naturally.

Fluorophore is the part of a molecule that makes it fluoresce.

The human eye contains autofluorescent substances in the retina, especially within the retinal pigment epithelium (RPE). The main autofluorescent component of the RPE is lipofuscin (LF), containing at least ten different fluorophores presenting discrete emission spectra within the green, golden-yellow, yellow-green, and orange-red emitting range⁽¹⁾.

When LF granules are stimulated with light within the blue range, a characteristic yellow fluorescence is emitted⁽²⁾.

 

Basic considerations on fundus autofluorescence

The RPE plays an important role in the physiopathology of age-related macular degeneration (AMD)^{(3, 4)}. The study of RPE can help to achieve a better understanding of AMD and to find new ways of early diagnosis as well as new prognostic and progression markers for this condition. LF originates from the constant phagocytosis of the shed outer segment disks of the photoreceptors and is accumulated in the cytoplasm of RPE cells^(5,6). This accumulation is considered to be a hallmark of RPE aging ⁽⁷⁾. Some experimental studies have addressed the molecular mechanisms of the interaction of excessive LF with the normal cellular functions of RPE^(8,9). According to these studies, A2-E (N-retinylidene-N-retinylethanol-amine) has been identified as the main fluorophore of LF. A2-E may play a toxic effect including phototoxic and detergent actions, as well as an inhibitory effect on lysosomal function^(10,11). It has been suggested that the photo-oxidation by-products related to LF may trigger the complement cascade, thus contributing to the pathological chronic inflammation of the macular area⁽¹²⁾. New studies have been designed to improve our understanding of potential underlying molecular mechanisms. The autofluorescence of LF, its distribution in post-mitotic human RPE cells and its accumulation with age have been extensively studied in post-mortem eyes with fluorescence microscopy^(14,15). In the past few years new research on ocular autofluorescence in vivo has also been performed.

Ultraviolet light is frequently used to visualize LF by fluorescence microscopy ex vivo, since light absorption limits the transmission of ultraviolet light within the retina in human living eyes. However, due to the wide range of excitation of LF (from 300 nm to 600 nm), visible light can be used to visualize its fluorescence in vivo. The emitted spectrum ranges from 480 to 800 nm and peaks within the range of 600 to 640 nm⁽¹⁵⁾.

Fundus autofluorescence (FAF) imaging of the human living eye is a relatively new imaging method that provides a topographic map of the distribution of LF in the RPE. The detection of FAF is limited by its low intensity (approximately two orders of magnitude lower than the peak background fluorescence of an ordinary fluorescein angiography), and by the FAF characteristics of the anatomic structures of the eye, including those of the optical media, especially of the lens⁽¹⁶⁾.

 

Imaging methods

1. Fundus spectrophotometry

Fundus spectrophotometry was developed by Delori et al. and was designed to determine the spectrum of excitation and fluorescence emission from small areas of the retina (2º diameter)^(17,18). The authors were able to determine the amount of FAF and compare it with in vitro fluorescence microscopy. They found out that the spectrum of in vivo excitation was slightly broader, and peaked at a longer wavelength than those of A2E and native LF, concluding that considering the spatial distribution, spectral characteristics and age relationship, LF is the main source of fluorescence in the FAF in vivo⁽¹⁸⁾.

Currently there are two available systems to examine the FAF of the human eye in vivo in the clinical practice: confocal scanning laser ophthalmoscope and fundus camera.

 

Scanning laser ophthalmoscope

Confocal scanning laser ophthalmoscope (cSLO) was originally developed by Webb et al. using a low-energy laser source to scan the retina in two directions termed as x and y⁽¹⁹⁾. The confocal nature of the optics ensures that the reflectance and fluorescence correspond to the same focal plane. cSLO overcomes the limitations of the low-intensity signal of FAF and the lens interferences. The defocused light is almost completely suppressed, thus reducing the FAF from the optical media anterior to the retina, such as the lens or the cornea. In order to reduce the background noise and to increase the contrast of the image, a series of FAF images are usually recorded^(20,21). Following the aligning of the images in order to correct the movement of the eye during the acquisition, the final image is calculated (usually from 4-32 frames) and the values of the pixels are normalised. 

The FAF image can be obtained with low excitation energies within the limits of maximum retinal irradiance established by the American National Standard Institute and other international standards⁽²²⁾. cSLO enables the acquisition of FAF images from wide areas of the retina (55º with one frame and even larger areas using the composite mode)^(20,22). Although limited by the optical properties of the human eye, SLO succeeds in imaging the posterior pole with a high contrast⁽²³⁾.

Currently, there are three different cSLO systems for FAF imaging: the Heidelberg Retinal Angiograph (HRA) (based on the HRA classic, HRA 2 and the Spectralis HRA) (Heidelberg Engineering, Dossenheim, Germany); the Rodenstock cSLO (RcSLO; Rodenstock, Weco, Düsseldorf, Germany); and the Zeiss prototype SM 30 4024 (ZcSLO; Zeiss, Oberkochen, Germany). HRA is the only currently commercially available system with the cSLO system to capture FAF images. HRA uses an excitation wavelength of 488 nm from an Argon laser or a solid-state laser. A barrier filter with a short-wavelenght cut-off at 500 nm is inserted just opposite the detector, blocking the laser light and letting the autofluorescent light through.

There are currently 3 kinds of cSLO-based FAF techniques:

1. Short wavelength FAF (486 nm blue laser): The most extended cSLO since 2009, which basically excites the RPE LF;
    
2. Long wavelength FAF (787 nm infrared laser): Less used, it excites the melanin of the RPE mainly;
    
3. 518 nm green laser FAF has also been employed lately. This wavelength is absorbed by the macular pigment to a lesser extent, showing a clearer image of the foveal RPE. 

The device uses both lasers and it is able to simultaneously capture both images and use them to measure the distribution of the macular pigment. Blue FAF images acquired with a cSLO operating at an excitation wavelength of 486 nm present a dark area in the fovea caused by absorption of blue light by the macular pigment. However, green FAF images acquired at a wavelength of 518 nm show much less of a shadowing effect as green light is less absorbed by macula pigment (Figure 1). Both blue and green FAF images can be used to quantify the macular pigment distribution⁽²⁴⁾.

In patients with geographic atrophy (GA) that present with foveal sparing, the border of the atrophic area is easily recognizable on green FAF images. The quantification of areas of GA becomes more reliable when using green FAF images. The patterns of hyperfluorescence seem similar in blue and green FAF images. Meanwhile, it has been made possible to acquire real time images, a technique known as real-time averaging.

Figure 1. Differences between blue and green FAF.

 

The Rodenstock cSLO and the Zeiss prototype SM 30 4024 have also been used to acquire clinical FAF images. Both systems use an excitation wavelength of 488 nm (the same as the HRA), and barrier filters at 515 nm and 521 nm, respectively^(20,21,25). Bellmann et al. have noticed marked contrast and brightness differences as well as in the grey range (an important marker of the image quality between the different systems of cSLO). These limitations must be taken into consideration when comparing images from different cSLO systems⁽²⁵⁾.

The default software of the HRA system normalizes the pixel distribution of the final image in order to improve the distribution of the FAF intensity. Even though this final step facilitates the evaluation of the localized topographic differences, it allows a relative estimation of the intensities of the FAF. Thus, it should not be used for quantitative calculation and absolute comparison between different FAF images. The normalization of the average images can be easily turned off, and brightness and contrast can be manually adjusted to permit an adequate visualization of the distribution of FAF in areas with a very high or very low signal in order to improve the visualization of small details.

 

Fundus camera

Fundus cameras are widely used in clinical routine for imaging the retina as fundus photographs, reflectance photographs and fluorescein angiography. Fundus cameras use a single flash to capture images from large retinal areas. When confocal optics are not available, the FAF signal from all the ocular structures with fluorescent properties reaches the camera, and scattered light, anterior and posterior to the plane of interest can influence the detected signal^(26-28). The lens contributes significantly to the FAF signal when similar wavelengths are used in the blue-range, as for the cSLO (wavelength = 488 nm), particularly in older patients with lens yellowing and nuclear opacities. Flashlight intensities and detector gain have to be set at relatively high levels in order to obtain reasonable FAF images. However, the signal-to-noise ratio decreases simultaneously which may result in reduced image quality.  To reduce interference from lens fluorophores that mainly emit in the range between 510 to 670 nm, Spaide modified the excitation filter (peak 580 nm, bandwidth 500-610 nm) and the barrier filter (peak 695 nm, bandwidth 675-715 nm). 

A further modification was introduced in 2007 using a slightly different filter set (excitation bandwidth 535-580 nm, emission bandwidth 615-715 nm)⁽²⁹⁾, thus improving signal-to-noise ratio and image quality. Furthermore, as this setup of the fundus camera uses different excitation and emission filters compared with the cSLO, it may even visualize other retinal fluorophores. However, a systematic comparison of different pathologies with clinico-pathological correlations between cSLO and fundus camera, particularly in patients with AMD, has not yet been performed.

In this chapter we will present a series of images of patients suffering from different forms of AMD using both fundus camera and Spectralis cSLO. 

Originally, a fundus camera that enabled imaging with a field of 13º was used. Afterwards, Spaide has obtained images of the spatial distribution of FAF intensities over larger retinal areas up to 50º with his new modified fundus camera^(26-29). In the near future it will be possible to improve FAF imaging with the aid of new filters and some other innovations, and increased experience. Furthermore, it is already possible to visualise different fluorophores from the retina with the configuration of the fundus camera using excitation and emission filters for the cSLO.

 

Autofluorescence imaging in the human eye in vivo

FAF images show the spatial distribution of the intensity of FAF of each pixel in grey values (arbitrary values from 0 to 225); low intensities are commonly known as low pixel values (dark) and high intensities as high pixel values (light).

 

Normal fundus

FAF imaging shows a consistent pattern of FAF distribution in normal eyes⁽²¹⁾. Such common findings have been reported in children as young as four years old⁽³⁰⁾. The macular FAF signal is reduced at the fovea because it is limited by the presence of lutein and zeaxanthin in the neurosensory retina. The signal is higher in the parafoveal area and tends to increase as we move away from it, peaking at the most peripheral retinal areas. It has been suggested that this FAF pattern is caused by the melanin deposition and density of LF granules at the different areas of the retina^(18,31). The optic nerve head typically appears dark mainly due to the absence of RPE. The retinal vessels are associated with a markedly reduced FAF signal because of the blocked fluorescence (Figure 2).

Figure 2. Colour fundus and FAF from a normal subject. (A) Right eye colour fundus and (B) FAF photographs. (C) Left eye colour fundus and (D) FAF photographs.

 

The common ratios of gGeographic atrophy (GA)rey intensity between the fovea and the perifoveal area have been established^(32,33). Considering these findings, any deviation from the normal pattern in a specific location can be easily identified; hence the qualitative description of the local changes in the FAF is widely used. The changes in signal intensity are qualitatively described as decreased, normal, or increased as compared to the background signal of the same eye.

 

FAF imaging in AMD

When examined with FAF, the fundus of patients with AMD may show a range of signal changes^(20,34-38). Assuming that RPE has an important role in the pathophysiology of AMD and that the major fluorophores in the retina are located within RPE cells, FAF imaging can show changes in the concentration and distribution of RPE LF and hence establish the condition of RPE in patients with AMD. Therefore, atrophic RPE typically appears as dark patches in FAF and can be clearly delineated, even better than in normal fundus photograph (Figure 3)^(21,39).

Figure 3. Atrophic area of RPE. These areas typically appear as dark patches in FAF images and can be clearly delineated. (A) Colour fundus and (B) FAF photographs.

 

All this information can be obtained from a quick and minimally invasive exploration with FAF. The decreased FAF intensity may also be associated with hyperpigmented areas due to the melanin absorption of light^(36,40). However, it should be considered that other fluorophores than LF can be found in RPE and become more prominent in AMD patients, and hyperpigmented areas may also cause an increase in the signal, which is supposed to result from the accumulation of melanolipofuscin.Other changes in FAF which are not related to RPE defects may appear in AMD. Fresh haemorrhages typically appear dark due to blocked fluorescence (Figure 4). However, these haemorrhagic areas eventually synthesize substances and fluorophores, which are observed in the fundus as yellowish areas and in FAF images as increased signals (Figure 5)⁽⁴¹⁾.

Figure 4. Fresh haemorrhage in the left eye from a patient with choroidal neovascularization (CNV) secondary to AMD. Fresh haemorrhages typically appear dark due to blocked fluorescence. (A) Colour fundus and (B) FAF photographs.

 

Figure 5. Haemorrhage in the left eye of a patient with CNV secondary to AMD. Fluorophores eventually appear in haemorrhagic areas which are observed as yellowish areas in the fundus and as an increased FAF signal. (A) Colour fundus photograph of the fresh haemorrhage. (B) Colour fundus photograph from the same eye one month later. (C) FAF of the haemorrhage.
 

Pigment epithelial and neurosensory detachment and areas with extracellular fluid accumulation associated with exudative lesions can be observed in FAF as increased or decreased signal intensity. Fluid accumulation under pigment epithelium detachment (PED), extracellular deposition of material under the RPE (drusen), and fluid originated from CNV can occur with increased, normal or decreased FAF intensity. This phenomenon is a consequence of the presence of unknown autofluorescent molecules other than LF, in the same spectral range than LF. FAF imaging alone may not distinguish between melanolipofuscin from RPE cells migrated into the neurosensory retina and LF within the normal RPE layer. It is always necessary to compare the FAF findings with those from other techniques such as fundus photograph, reflectance image, fluorescein angiography or optical coherence tomography (OCT)^{(36, 40)}.

 

FAF in early AMD

Early AMD is characterised by the appearance of localized RPE hypo or hyperpigmentation and drusen. Drusen are formed by the accumulation of extracellular deposits in the inner aspects of Bruch’s membrane⁽³⁾. Depending on their size and morphology, they can be classified as hard or soft drusen. The molecular composition of drusen is quite complex and has not been completely elucidated (Figure 6).

Figure 6. Small drusen from a patient with early AMD in the left eye. (A) Colour fundus and (B) FAF photographs.

 

FAF changes in early AMD have already been reported by several authors^{(9,21,26,28,34-37,42);} all of them concluding that the changes in ophthalmoscopy and fluorescein angiography are not necessarily related to FAF, suggesting that FAF may provide new information regarding the stages and activity of the disease.

Differentiation between RPE LF and sub-RPE deposits with FAF images in vivo can be a hard work. An analysis of the variability of FAF in patients with early AMD was reported in an international workshop on FAF phenotype in early AMD. Among their conclusions, a new classification system with eight different FAF patterns was given⁽⁴⁰⁾.

Normal pattern characterized by a homogeneous background autofluorescence with a gradual fluorescence decrease in the inner macula towards the foveola (blocked fluorescence caused by yellow macular pigments). AF may be normal even in the presence of soft or hard drusen. 

Minimal change pattern characterized by a limited and irregular decrease or increase of background FAF, not associated to any obvious or important topographic pattern.

Focal increased pattern is defined by the presence of at least one well defined spot (<200 micron diameter) of markedly increased FAF much brighter than the surrounding background fluorescence. These areas may or may not correspond to large, soft drusen and to areas of hyperpigmentation.

Patchy pattern, characterized by the presence of at least one large area with well-defined borders (>200 micron diameter) of markedly increased FAF. Again, these areas may or may not correspond to large, soft drusen and areas of hyperpigmentation.

Linear pattern defined by the presence of at least one linear area of markedly increased FAF with well-defined borders. Linear structures of increased FAF usually correspond to hyperpigmented lines. This pattern shows multiple branching lines of increased FAF forming a lacelike pattern. The borders may be hard to define, and FAF may gradually decrease from the centre of the linear area towards the surrounding background. This pattern may correspond to hyperpigmentation or to non-visible abnormalities.

Reticular pattern is defined by the presence of multiple small, well-defined areas (<200 micron diameter) of decreased FAF. This pattern has been found to occur not only in the macular area, and may be associated with multiple small soft drusen, hard drusen or areas with pigment changes or non-visible abnormalities.

Speckled pattern is characterized by the simultaneous presence of different types of abnormalities in a large area. The changes reach beyond the macular area and may cover the entire posterior fundus. These abnormalities include multiple small areas of irregular increased or decreased FAF corresponding to hyper and hypopigmented areas and multiple subconfluent and confluent drusen.

The speckled pattern has been reported to be the most frequent (26%) followed by the patchy pattern (23%). The most infrequent patterns are the normal pattern (2%) and the lacelike pattern (2%). The study confirmed that visible drusen on fundus photography are not always correlated with noticeable FAF changes and that areas of increased FAF may or may not correspond to areas of hyperpigmentation or soft or hard drusen. Several authors have also mentioned the different FAF patterns in eyes with drusen.

Delori et al. described a pattern consisting of decreased FAF in the centre of the drusen surrounded in most of the cases by a ring of increased FAF⁽²⁷⁾. They also observed that the decreased drusen signal was not as intense as in the areas with RPE atrophy. The authors hypothesized that it might be caused by a displacement of the cytoplasm and LF granules in RPE cells instead of an actual RPE atrophy⁽³⁶⁾.

Von Rückmann et al. further reported that crystalline drusen are characterised by a decrease in FAF, signalling the onset of atrophy⁽²¹⁾. Lois et al. confirmed that areas of confluent drusen are usually associated with focal, mildly increased FAF and that only large subfoveal soft drusen (drusenoid RPE detachments) topographically correspond with focal changes of FAF (Figure 7)⁽³³⁾.

Figure 7. Drusen in a patient with early bilateral AMD. (A) Right eye colour fundus and (B) FAF photographs. (C) Left eye colour fundus and (D) FAF photographs.

 

Smith et al. reported their results after using image analysis software to study drusen and pigmented areas on fundus photographs from AMD patients⁽⁴²⁾. The authors initially used image analysis algorithms, including automated background levelling and thresholding. Areas of focally increased FAF intensities were compared to the normal background signal. By overlapping fundus photographs and FAF, the topographic correlation of drusen and pigmented areas with focally increased FAF signals was established.
Smith and co-workers reported that eyes with isolated drusen or pigment abnormalities were better correlated with FAF abnormalities than eyes with GA⁽⁴²⁾.

Regarding areas with changes in RPE, hypopigmented areas are usually associated with a corresponding decreased FAF signal, suggesting an absence or degeneration of RPE cells, with reduced content of LF granules (Figures 8 and 9). 

Figure 8. RPE hypopigmentation in the macular area secondary to AMD. Hypopigmented areas are usually associated with correspondingly decreased FAF signals, suggestive of RPE cells loss or degeneration with reduced content of LF granules. (A) Colour fundus and (B) FAF photographs.

 

Figure 9. RPE hypopigmentation in the macular area secondary to AMD. (A) Colour fundus and (B) FAF photographs.

 

However, hyperpigmented areas frequently show a higher FAF signal, which may be caused by a higher concentration of autofluorescent melanolipofuscin (Figure 10)⁽³⁶⁾.

Figure 10. RPE hyperpigmentation in the macular area secondary to AMD. Areas with hyperpigmentation frequently show a higher FAF signal which may be caused by a higher amount of autofluorescent melanolipofuscin. (A) Colour fundus and (B) FAF photographs.

 

Reticular pseudodrusen are seen as small triangular deposits above the RPE in OCT imaging. Infrared pictures show them as small white spots with a surrounding darker halo and FAF typically show an orange skin pattern (Figure 11). 

Figure 11. Reticular pseudodrusen. FAF and SD-OCT of a patient with bilateral reticular pseudodrusen.

 

FAF in advanced AMD

Advanced AMD is characterized by geographic atrophy (GA), choroidal neovascularization (CNV), pigment epithelial detachment (PED), RPE tears and disciform scars.

Geographic atrophy (GA)

GA is thought to be the natural end stage of the atrophic AMD process when CNV does not appear. GA occurs in areas where the RPE is dead and the outer neurosensory retina and choriocapillaris disappeared^(43,44). Due to the loss of RPE and LF, the atrophic area appears dark in FAF imaging⁽³⁶⁾. High contrast between the atrophic and the non-atrophic retina defines the area of GA more precisely than colour fundus photographs, permitting a clearer and more specific study of GA, as well as its natural development and evolution (Figures 12 and 13)^(39,45).

Figure 12. GA secondary to AMD. The loss of RPE and LF causes a characteristic absence of FAF signal (a dark area in FAF images). (A) Right eye colour fundus and (B) FAF photographs. (C) Left eye colour fundus and (D) FAF photographs.

 

Figure 13. GA secondary to AMD. The high contrast between atrophic and non-atrophic retina enables the delineation of the atrophic area more precisely than can be performed from conventional fundus photographs. (A) Colour fundus and (B) FAF photographs.

 

The GA patches usually become larger and coalesce as AMD progresses^(46,47). An excessive accumulation of LF, and therefore an increased FAF in the junction are highly suggestive of the appearance or progression of pre-existing GA (Figure 14).

Figure 14. GA secondary to AMD with increased autofluorescence in the junction. New areas of GA and the extension of pre-existing areas are characterised by an excessive accumulation of LF, and therefore an increased FAF signal. (A) Colour fundus and (B) FAF photographs.

 

Preliminary observations suggest that different phenotypes may appear associated with junction FAF changes⁽⁴⁸⁾. 

A new classification for junction FAF patterns has been proposed in GA patients(Figure 15)⁽⁴⁹⁾.

Figure 15. GA secondary to AMD with a banded pattern of increased autofluorescence in the junction. (A) Colour fundus and (B) FAF photographs.

 

Proposed classification for junction FAF patterns in GA⁽⁴⁹⁾:

• Focal increased FAF is defined by single or multiple spots of focal markedly increased FAF localized at the border of the atrophic patch;
• Band pattern of increased FAF is characterized by a continuous stippled band of increased FAF surrounding the entire atrophic area;
• Patchy increased FAF are large patches of increased FAF outside the GA area. FAF tends to be less intense than that in the focal pattern described above;
• Diffuse increased FAF is the most frequent pattern of increased FAF in eyes with GA.

 

FAF changes are not limited to the border of the atrophic area and may show inter individual differences that have been further classified into four subtypes:

• Reticular pattern, characterised by several lines of increased FAF usually following a radial pattern;
• Branching pattern shows a diffusely increased FAF with a fine branching pattern of increased FAF;
• Fine granular pattern is defined by a large area of increased FAF with a granular appearance surrounding the GA area and a clear border between the granular increased FAF and the surrounding normal background FAF;
• Fine granular with peripheral punctate spots pattern is characterised by diffuse FAF changes surrounding the atrophic area with elongated small lesions and increased FAF.
   

Refined phenotypes help to identify the prognosis and seem to be a prerequisite to determine specific genetic factors in a complex, multifactorial disease such as AMD. An analysis of the follow-up of junction FAF patterns in GA and the rate of progression of atrophic lesions revealed that variation in GA growth rates are dependent on the specific phenotype of FAF at baseline⁽⁵⁰⁾. Atrophy enlargement was slowest in eyes with normal FAF pattern (median, 0.38 mm²/year), followed by focal FAF pattern (median, 0.81 mm²/year), diffuse FAF pattern (median, 1.77 mm²/year) and banded FAF pattern (median, 1.81 mm²/year). The rate of progression of GA in eyes with patchy FAF pattern were not included in this analysis because of their low frequency, insufficient for statistical analysis. The rate of progression of band and diffuse FAF patterns were significantly higher compared to eyes without FAF abnormalities and focal FAF pattern.

Another interesting finding of this study was the identification of eyes with extremely rapid progression of the atrophy, showing distinct FAF features of atrophy that had not been previously reported.

The authors introduced the term diffuse trickling for a pattern associated with a significantly faster enlargement of atrophy. 

Areas with increased FAF and consequently higher concentrations of RPE LF precede the development of new areas of GA or the enlargement of the pre-existing atrophic areas⁽⁵⁰⁾.

The phenotypic features of FAF abnormalities may play a stronger influence on the progression of atrophy than any other previously reported risk factors such as smoking, arterial hypertension or diabetes. 

The different rates of enlargement of atrophy may be related to heterogeneity at a cellular and molecular level in the disease. The high degree of symmetry in GA suggests that genetic determinants may be involved, rather than nonspecific aging changes.

GA is a relatively slow progressing disease. Currently, FAF patterns are mostly accepted as prognostic predictors for GA progression. 

Risk factors for the progression of GA include the perimeter of damaged RPE surrounding the GA lesion, the number and size of foci and the FAF pattern. 

It has been shown that hyperautofluorescence of the junction zone surrounding regions of GA is related to a higher progression rate. Holz et al.⁽⁵⁰⁾ showed that progression rates in eyes with the banded and the diffuse FAF pattern (especially those with diffuse trickling pattern) were significantly higher compared to eyes with focal patterns or with no FAF abnormalities, and these changes were later confirmed by other authors⁽⁵¹⁾.

Other authors evaluated initial GA lesions and grouped them into three categories: predominantly hyperautofluorescent (hyperAF), both hyper- and hypoautofluorescent (mixed AF), or predominantly hypoautofluorescent (hypoAF). In their series, the FAF characteristics were significantly dependent on the type of atrophic area: initial GA lesions were most commonly mixed AF, while drusen-associated atrophy was most commonly hypoAF⁽⁵²⁾. Progression of GA areas seems to be symmetrical between both eyes of the same patient and in some series it has been observed that lesion growth is faster in larger lesions, suggesting that progression rate is related to basal area⁽⁵³⁾. However these findings are controversial. Some groups support a linear model of growth over a quadratic model⁽⁵⁴⁾. A square root transformation of baseline GA area appears to ameliorate the effect of baseline area on GA progression. A classification based on the area and the irregularity of the lesion perimeter shows that there is an association with growth rates by providing a measure of the relative border of RPE at risk, since the extent of the junction zone of damaged RPE increases with non-circularity for a GA area^(55,56). Measuring the lesion growth in mm² may not accurately reflect the relative dynamics of progression of the lesion, and a concept such as proportion of growth might be more intuitive. Growth of atrophy is not linear within time, no matter what measurement is used (mm², border, proportion or square root).

Caire et al. tried to identify the genetic risk factors that contribute to the presence and progression of established GA in a Spanish population. According to their results, genetic risk factors associated with the presence of GA (genetic polymorphisms within CFH, ARMS2 and FHR1-3) are not identical to those related to its progression (polymorphisms of CFB and CFH)⁽⁵⁷⁾. 

Some samples of patients with GA can be seen in Figures 16, 17 and 18 (FAF defect progression during one year of follow-up).

Figure 16. Patient with GA: FAF defect progression during one year of follow-up (example 1).

 

Figure 17. Patient with GA: FAF defect progression during one year of follow-up (example 2).

 

Figure 18. Patient with GA: FAF defect progression during one year of follow-up (example 3).

 

Figure 19. FAF patterns in GA

 

Figure 20. Atrophic AMD progression. FAF of a 78-year-old woman suffering from an atrophic form of AMD. Progression of the disease can be clearly appreciated over a 6 year follow-up (A to F).

 

Figure 21. GA case 1: 88-year old woman showing a bilateral atrophic AMD. Although the atrophy is extensive, the foveal centre is still preserved and her visual acuity remains at 0.4 and 0.7 respectively.

 

Figure 22. GA case 2: 71-year-old male suffering from dry AMD. FAF shows us the hyperfluorescent drusen and the limits of the GA which allows us to properly follow these patients.

 

Figure 23. Study of GA with FAF and SD-OCT. Study of dry AMD combining FAF and SD-OCT. The disappearance of layers of the external retina is clearly visible.

 

Choroidal neovascularization (CNV)

CNV is considered to cause almost 90% of the cases of severe visual loss related to AMD⁽⁵⁸⁾. CNV is usually studied by fluorescein angiography and OCT to assess the extent, location and nature of the lesion⁽⁵⁸⁾. Fluorescein angiography shows changes in retinal vascularization, but does not reveal how deeply RPE is affected.FAF imaging shows RPE damage, with the advantage that is a non-invasive test, less time consuming than angiography. Several studies have reported that CNV may show irregular FAF alternating areas of increased, normal and decreased fluorescence intensity (Figure 24)^(20,35-38). These studies have also reported that areas with previously high levels of FAF may show decreased FAF six months later⁽³⁷⁾. Data comparing FAF findings in occult and classic CNV are limited. These changes may be secondary to photoreceptor loss, RPE atrophy, replacement of normal phenotypes of RPE cells with scar and increased melanin deposition. These findings may have therapeutic implications and clarify long-term visual prognosis. For example, a person with an active CNV on fluorescein angiography and normal FAF may show a much better outcome than another with an abnormal basal FAF. Spital et al. reported that classic CNV usually shows more focal areas of decreased FAF than occult CNV⁽³⁵⁾. These findings have been confirmed by McBain et al.⁽⁶¹⁾ who guessed that low FAF at the site of the CNV are related to blocked fluorescence induced by the presence of CNV in the subretinal space, rather than to severe damage to the RPE. Another study did not find significant differences in FAF patterns in early classic and occult CNV secondary to AMD (Figures 25 and 26)⁽⁶⁰⁾. A continuous preserved autofluorescence pattern was observed in the central macula in most of the cases. These findings suggest that neovascular complexes, regardless if classic or occult, would be external to the RPE in most cases. Additional studies with a higher number of patients and longer monitoring are required to verify with these changes in patients with CNV (Figure 27).

Figure 24. Occult CNV. These lesions show irregular FAF intensities with alternating areas of increased, normal and decreased signal intensity. (A) Colour fundus and (B) FAF photographs. (C) and (D) Early frames fluorescein angiography. (E) Late frames fluorescein angiography.

 

Figure 25. Retinal angiomatous proliferation (RAP). (A) Colour fundus and (B) FAF photographs.

 

Figure 26. Classic CNV. (A) Colour fundus and (B) FAF photographs.

 

Figure 27. CNV with chorioretinal anastomoses. (A) Colour fundus and (B) FAF photographs. (C) Early frame fluorescein angiography. (D) Late frame fluorescein angiography.

 

Pigment epithelial detachment (PED)

PED can show different FAF patterns^{(11,16,32,62)}. FAF can provide complementary information to that of fundus colour photograph and fluorescein angiography. In most of the cases, a moderately and diffusely increased FAF can be found, surrounded by a clearly defined ring with decreased fluorescence^(35,36,62). Occasionally, intermediate or even decreased FAF can be found that may not correspond to the atrophic RPE or to fibrovascular scars. These changes in the FAF could correspond to different stages in the PED evolution⁽¹⁶⁾. These findings in FAF should be compared with those in fluorescein angiography. We should bear in mind that areas with increased FAF do not always correspond to increased or decreased LF. Besides, the presence of other fluorophores, extracellular fluid or degraded photoreceptor remnants should be considered (Figure 28)⁽¹⁶⁾.

Figure 28. Retinal PED. (A) Colour fundus and (B) FAF photographs.

 

RPE tears

FAF imaging is a useful tool to diagnose RPE tears^(35,63). RPE tears usually occur in association with PED in patients with neovascular AMD, either spontaneously or following therapy⁽⁵⁸⁾. FAF imaging reveals absence of autofluorescence in the area denuded from RPE. These areas are clearly identifiable by their very low signal, whereas a heterogeneous FAF signal is seen in the area where the RPE is rolled. Therefore, the exact location of the tear can be delineated in most cases.

 

Disciform scars

The appearance of disciform scars in FAF imaging depends on their duration and evolution^(35,37). Disciform scars may show different variations and alterations of FAF signal. A decreased signal is typically observed in scarred and fibrotic areas. It has been reported that approximately 50% of the disciform scars may be surrounded by a rim of increased FAF^(35,37). These areas of increased autofluorescence correspond to irregularly pigmented areas and may have been caused by a multi-layered RPE, a well-illustrated finding in histopathology (Figures 29 and 30)⁽³⁶⁾.

Figure 29. CNV with fibrosis. FAF outlines the marked atrophic lesions in the RPE surrounding the CNV/fibrosis. These changes are inconspicuous in colour photographs. (A) Colour fundus and (B) FAF photographs.

 

Figure 30. Fibrous scar secondary to CNV after treatment with anti-VEGF. The damaged RPE appears hyperpigmented in fundus photograph, whereas FAF imaging shows an increased signal. (A) Colour fundus and (B) FAF photographs.

 

Areas of GA and type II CNV may appear simultaneously (Figure 31). 

Figure 31. GA and type II CNV. FAF images of a 67-year-old woman who presents a bilateral type II AMD CNV, along with an evident RPE atrophy.

 

FAF study also reveals the RPE changes after anti-VEGF injections to treat a type I CNV (Figure 32).

Figure 32. CNV - RPE changes. FAF of a 82-year-old man after a series of 8 anti-VEGF injections to treat a type I AMD CNV of his left eye. The right eye shows mottled RPE alterations.

 

In cases of a type I CNV, the detached retina is seen as a hyperfluorescent area while blood spots produce the typical mask effect (Figure 33).

Figure 33. CNV - detached retina. FAF image of a 75-year-old woman at the moment of the diagnosis of a type I CNV. The detached retina is seen as a hyperfluorescent area while blood spots produce the typical mask effect.

 

>> References

Introduction

Geographic atrophy (GA) is considered the late stage of the dry form of age-related macular degeneration (AMD)⁽¹⁾.

GA is less common than neovascular AMD and it is responsible for 10-20% of cases of legal blindness in this condition(2,3,4), affecting more than 5 million people wordwilde⁽⁵⁾.

Currently there is no approved or effective treatment to prevent either the onset or progression of GA, however, in recent years, significant progress has been made in understanding the pathogenesis of GA, which has led to a number of new potential therapies currently undergoing clinical trial evaluation^(6,7,8).

 

Definition

Usually defined as any sharply delineated round or oval area of hypopigmentation, or apparent absence of the retinal pigment epithelium (RPE), in which choroidal vessels are more visible than in surrounding areas, that must be at least 175 μm in diameter (Figure 1)⁽¹⁾.

Figure 1. Fundus photography showing a sharply delineated area of hypopigmentation or apparent absence of RPE in both eyes of the same patient.

 

However larger dimensions have also been proposed for GA minimum limit. The Complications of Age-Related Macular Degeneration Prevention Trial (CAPT) specified a diameter ≥ 250 μm for definition of GA and in the AREDS2 study the minimum diameter of the area was increased to 433 μm. Currently, there is no international consensus on the minimum diameter for the diagnosis of GA⁽⁸⁾.

In addition, changes have been proposed in the grading of AMD severity level regarding the presence of GA. In the AREDS grading system, the presence of GA qualifies patients as having either intermediate or advanced AMD, depending on whether or not there is central involvement⁽⁹⁾.

In the classification system recently developed by the Beckman Initiative for Macular Research, the presence of any GA regardless of location qualifies as evidence of advanced AMD⁽¹⁰⁾.So far there is no approved or effective treatment to prevent the onset and progression of GA

One eye with GA and choroidal new vessels (CNV) is considered as having exudative form⁽¹¹⁾.

 

Epidemiology and risk factors

The global prevalence of GA is 0.66% in all ages but it occurs in 0.34% between 65-74 years old, 1.3% between 75-84 and 4.4% over 85 years old.

The prevalence increases to 22% after 90 years of age^(2).

It corresponds to half of exudative form prevalence^(4,12-14).

A recent meta-analysis based on data in the United States reported that the incidences of GA and neovascular AMD were 1.9 and 1.8 per 1,000 American whites aged ≥ 50 years. Prevalence rates for the two forms of advanced AMD were comparable across all age groups⁽¹⁵⁾.

For subjects with mild or intermediate AMD the 15-year cumulative incidence of neovascular AMD is approximately 2.0% and for progression to pure GA it is approximately 1.3%⁽¹⁶⁾.

The most consistent risk factors are age, familiar history and tobacco smoking, like in other forms of AMD(^(2,4,13).

In AREDS, GA was also associated with: high BMI (body mass index); use of calcium channel blockers and ß-blockers; not using anti-acids; not using hormone replacement (woman); light iris color and less education level^(14-17).

The prevalence of GA is lower in blacks than in whites (2.1% vs 4.8%) like in exudative forms^(18,19).

In Rotterdam and Beaver Dam Eye Study, serum HDL cholesterol was directly associated with GA, however this association was not found in Blue Mountains Eye Studyl⁽²⁰⁾. 

In this study, diabetes and the ratio total/HDL cholesterol were linked to increased risk of GA.

As complement system seems to play an essential role in this disease, the complement factor H (CFH) gene located at chromosome 1q32, and others as CFB, CFI, C2 and C3, have been implicated in the development of both forms of AMD^(21-23). 

Some studies have linked specifically 5p region and 4q 32 region with GAl^(24,25).

Only recently, more detailed genome-wide association studies have investigated whether the 2 subtypes of advanced AMD segregate separately in families and associate with different disease variantsl⁽²⁶⁾.

The variants in the 10q26 locus confer increased risk for both advanced AMD subtypes, but imparts greater risk for CNV than for GA.

Other loci were detected with suggestive associations that differ for advanced AMD subtypes and deserve follow-up in additional studiesl⁽⁶⁾.

 

Pathology

Accordingly to AREDS the most common sequence of events leading to GA is the progression of a large drusen to hyperpigmentation, followed by regression of the drusen, hypopigmentation and ultimately RPE cell death, with development of an atrophic area of retina and underlying choriocapillaris, sometimes preceded by the appearance of refractile deposits.

This evolution can be longer than 6 years^(27-29).

The average time from entering the AREDS Study to initial appearance of GA has been shown to be approximately 5 to 6 years in the presence of confluent or large (> 125 μm) drusen and hyperpigmentation, as compared with 2.5 years in the presence of hypopigmentation⁽⁶⁾.

The presence of large confluent drusen is a significant risk factor for developing GA^(30).

Drusen were found in 100% of patients at the site of later GA development; drusen > 125 μm, confluent drusen and hyperpigmentation were present in > 90% of eyes, whereas drusen > 250 μm and hypopigmentation were present in > 80% eyes⁽³¹⁾.

Less frequently, GA can follow a drusenoid RPE detachment, regression of a CNV membrane or a RPE rupture^(11,32,33).

Higher prevalence of reticular pseudodrusen (RPD) has been reported in patients with GA compared with early or intermediate AMD. Also, the presence of RPD appear to represent a meaningful risk factor associated with progression to GA^(34,36).

In some eyes, atrophy was related to a micro reticular pigment pattern distributed around the perimeter of the fovea⁽³³⁾.

Most histopathologic studies suggest that RPE cells are the primary target in GA and its death results in choriocapillaris atrophy^(37,38).

Autofluorescent pigments as lipofuscin that accumulate in RPE cells, contribute to a decline of the cell function and degeneration conducting to GA⁽³⁹⁾.

Although RPE cells are the most involved, the outer nuclear layer is also severely affected with dysfunction and death of photoreceptors. Probably rods are the first affected photoreceptors^(40,41).

The mechanisms of RPE death are best studied at junctional zone.

Here, lipofuscin may occupy 30% of RPE cell and may interfere with its metabolism, conducting to death.

These mechanisms include oxidative stress and inflammation^(42-45).

Macrophages are often seen in areas of GA, apparently phagocytosing pigment and debris resulting of normal cells deletion⁽⁴⁶⁾.

 

Diagnosis

GA can be distinguished from other forms of dry AMD based on stereo biomicroscopy and color fundus photography⁽⁶⁾.

The definition of GA is based on clinicohistopathologic investigations that have shown that clinically visible areas of atrophy are characterized by cell death in the RPE, outer neurosensory retina and choriocapillaris.

With modern in vivo imaging technology, these findings can be confirmed⁽⁶⁾.

Spectral-domain optical coherence tomography (SD-OCT) and fundus autofluorescence (FAF) allow for noninvasive and rapid quantitative morphological assessment of GA in the clinical setting⁽⁶⁾.

 

Fundus

Fundoscopy in GA typically shows a well-circumscribed oval or round area of pigment epithelium atrophy with sharply demarcated borders and increased visibility of choroidal vessels, usually sparing the fovea until late stages  (Figure 2).

Figure 2. Fundus photography showing a well-circumscribed round area of pigment epithelium atrophy, sparing the fovea in right and left eyes.

 

All precursor lesions of this final appearance can also be present: large drusen (>125 microns), focal pigmentation changes and refractile deposit^(27,28).

 

Angiography

On fluorescein angiography, GA appears as a sharply delineated window defect due to atrophy of overlying layers of RPE (Figure 3).

Figure 3. Fluorescein angiography showing a sharply delineated window defect.

 

A prolonged choroid filling phase has been described as a clinical marker for changes in Bruch’s membrane and as a risk factor for development of GA⁽²⁹⁾.

Despite these aspects in GA, fluorescein angiography may be indicated only in atypical cases, in order to allow the correct diagnosis⁽⁴⁷⁾.

 

Optical coherence tomography (OCT)

OCT scan shows thinning of hyperreflective external band, corresponding to attenuation of RPE/Bruch’s complex, and deeper hyperreflectivity because of loss of outer layers including photoreceptors (Figure 4)^(48,49).

Figure 4. Spectralis OCT: Thinning of hyperreflective external band (because of attenuation of RPE/Bruch’s complex) and deeper hyperreflectivity.

 

In high resolution OCT the atrophic area shows hyperreflective clumps at different levels, segmented plaques of the outer band and elevations with variable reflectivity⁽⁴¹⁾.. In the perilesional area there are elevations of the outer retinal layers, as well as thickening of outer hyperreflective band. At the junction area the outer band shows different degrees of loss⁽⁵⁰⁾.

Recently, changes in OCT that precede the development of GA have been identified. These changes include the presence of hyperreflective foci in the retina overlying drusen, a subsidence of the inner nuclear layer (INL) and outer plexiform layer (OPL) with the development of hyporeflective, wedge-shaped bands, and increased signal transmission below the level of the RPE.

These anatomic changes might be used to identify patients early in the course of GA development who might still be at a reversible stage and therefore amenable to intervention^(51-53).

OCT imaging may also contribute to a better understanding of the underlying pathologic mechanisms in AMD and GA, may suggest new biomarkers related to disease progression, and might potentially indicate new therapeutic targets in AMD⁽⁸⁾.

 

Fundus autofluorescence (FAF)

FAF is the current standard imaging technology for the morphological assessment of GA⁽⁸⁾.

Fundus spectrophotometric studies in vivo by Delori and co-workers have shown that FAF represents an accumulation of lipofuscin in the lysossomes of RPE cells, mainly derived from photoreceptors outer segments degradation.

The compound is found as micrometer-sized spherical particles and is characterized by yellow autofluorescence when exposed to blue light^(54-56).

It has been shown with confocal scanning laser ophthalmoscopy (cSLO) that FAF response is very low or extinguished in areas of atrophy.

The lack of RPE cells or its low number and therefore of lipofuscin, (the dominant fluorophore) explain this reduction⁽⁵⁷⁾.

Increased FAF precedes development of GA^(58,59).

FAF is increased in junctional zone around areas of atrophy, and intensity seems to correlate with extension of the atrophic area and also with reduction of retinal sensitivity detected by fundus perimetry (Figure 5)^(60,61).

Figure 5. FAF showing increased fluorescence in the junctional zone around areas of atrophy.

 

The FAM-Study Group developed the classification of abnormal FAF according to distinct patterns in the junctional zone of the GA area and identified the following five primary phenotypes based on the presence of increased hyper-autofluorescence: None, Focal, Banded, Patchy and Diffuse⁽⁶²⁾.

Holz et al. showed that the phenotypic features of FAF abnormalities had a much stronger impact on atrophy progression than any other risk factor that has been addressed in previous studies on progression of GA attributable to AMD, and introduced the ‘diffuse trickling' pattern that is associated with an extremely rapid progression of atrophy⁽⁶³⁾.

These hyperfluorescent areas in the junctional zone are believed to represent regions of cells that are stressed and more likely to become atrophic⁽⁸⁾.

 

Psychophysical Tests of Visual Function

GA develops gradually with the formation of scotomas that spare the fovea initially, expanding into the central visual field late in the course of disease⁽⁸⁾.

The most common assessment of visual function, best corrected visual acuity (BCVA), usually fails to reveal the functional deficits experienced by patients with foveal-sparing GA⁽⁸⁾.

Functional assessments beyond BCVA including multifocal electroretinography (mfERG), microperimetry, low-luminance visual acuity (LLVA), reading speed and contrast sensitivity, should be used to more fully assess visual impairment⁽⁸⁾.

 

Clinical evolution

Eyes with GA may also develop CNV. In one study, 7% of eyes with GA developed CNV in 2 years.

The strongest risk factor for developing CNV in one eye with GA is the presence of CNV in the fellow eye⁽⁶⁴⁾.

The 4-year rate of developing CNV is 11% if the other eye has pure GA but increases to 34% if there is CNV.

As other forms of late AMD, GA tends to be bilateral (over 50% of cases) and there is high symmetry between eyes for total atrophic area, presence of peripapillary atrophy and enlargement rate(^(64-66).

On the other hand there is a high interindividual variability⁽⁶⁷⁾.

Longitudinal studies have shown that progression rates vary widely among patients⁽¹⁰⁾.

The mean overall enlargement rate of atrophic area is 2.6 mm2/year.

Eyes with larger areas of atrophy at baseline tend to have larger enlargement rates⁽⁶⁵⁾.

The Beaver Dam Eye Study showed that eyes with multifocal disease had larger increase in area of GA and progressed to foveal involvement more frequently than eyes with single foci of disease over 5 years (1.2 vs 2.24 mm²)⁽⁶⁸⁾.

GA often first develops surrounding the fovea, sparing the central area. Because of that, the correlation between visual acuity and area of atrophy can be complex^(69-71).

The loss of three lines (ETDRS scale) was observed by Suness et al. in 31% of studied eyes by 2 years and 53% by 4 years. More than half of patients with GA encroaching upon the fovea will suffer this severity of vision loss in the same time frame when compared with extrafoveal GA⁽⁷¹⁾.

Even when GA is not center involving, patients can suffer significant deficits in visual function, such as compromised reading ability and impaired vision in dim lighting⁽⁷²⁾.

The occurrence of GA with concomitant CNV is associated with an even higher risk of severe vision than GA alone^(73-75).

Suness et al. showed that 86% of eyes with GA that developed CNV lost 3 or more lines of visual acuity over 2 years compared with only 27% in eyes with GA that did not develop CNV⁽⁶⁴⁾.

The risk of visual acuity loss was higher in eyes with better visual acuity at baseline and with lightest iris color⁽⁷¹⁾.

Usually vision loss is bilateral because of lesion symmetry, and evolution since first signs to legal blindness is quite variable^(65,66).

Because of rods paucity, reduction of foveal cone function and also because of switching from central to eccentric fixation, other visual functions as the contrast sensitivity, reading rate and dark adaptation are decreased in GA^(69,72,72).

The low luminance visual dysfunction and the reduction of the maximum reading rate seem to be significant risk factors for subsequent visual acuity loss⁽⁶⁹⁾.

 

Management

Prevention

In AREDS, dietary supplements as zinc, anti-oxidants, vitamins C and E and beta-carotene have been shown to reduce the risk of progression in participants in categories 3 and 4 to advanced AMD (25% in 5 years), however, in the group with GA away from the center (category 3), this reduction was not statistically significant.

Despite of that the AREDS Report nº 8 concluded that those with noncentral GA also should consider taking a supplement of antioxidants plus zinc^(9,17).

Macular xantophylls and polyunsaturated fatty acids seem to be associated with a lower risk of advanced AMD^(76,77).

Because of that antioxidant effect of macular pigments, lutein, zeaxanthin and omega-3 fatty acids have been tested in the AREDS 2 study^(76,78).

The researchers also tried substituting lutein and zeaxanthin for beta-carotene, which prior studies had associated with an increased risk of lung cancer in smokers.

The AREDS 2 study found that while omega-3 fatty acids had no effect on the formulation, lutein and zeaxanthin together appeared to be a safe and effective alternative to beta-carotene.

Subgroup analysis indicated that those with the lowest intake of lutein and zeaxanthin, supplemental lutein and zeaxanthin were protective⁽⁷⁹⁾.

Low dietary glycemic index also seems to reduce the risk of evolution to advanced AMD⁽⁷⁹⁾.

Other behavioral factors such as stop smoking, as it represents a key modifiable risk factor and control of BMI may play an important role on prevention^(80-82).

 

New treatments

So far there is no approved or effective treatment to prevent the onset and progression of GA^(6,30).

Although the main target for GA remains unknown, several trials are investigating strategies regarding these pathways and therapeutic targets:

1. Visual cycle toxic products (visual cycle modulators);
2. Inflammation, complement, and ECM (mTOR inhibitors, complement inhibitors, MMP inhibitors);
3. Lipoprotein accumulation (LDL lowering drugs);
4. Beta-amyloid accumulation (anti-amyloid beta);
5. Oxidative stress (antioxidants and neuroprotectant);
6. Choriocapillaris atrophy (choroidal perfusion enhancers);
7. RPE and photoreceptor loss (Stem cell therapy and neurotrophins);⁽⁶⁾

 

Visual cycle Inhibitors

Accumulation of lipofuscin and melanolipofuscin granules have been observed at the sites of RPE atrophy in GA and associated with GA pathogenesis⁽⁷⁾.

ALK-001 is a modified vitamin A molecule, which slows the formation of lipofuscin and RPE apoptosis.

This compound was initially designed to treat Stargardt’s disease. It may also benefit GA by reducing toxic A2E, all-trans retinal, and lipofuscin accumulation.

A phase 1 study was designed to assess the safety and pharmacokinetics of oral ALK-001 capsules in 40 healthy volunteers with no results posted yet^(6,7).

Fenretinide and Emixustat (ACU-4429) were also studied but there was no evidence of benefit^(6,7,83-85).

 

Anti-inflammatory Agents

Chronic inflammation is believed to play an important role in AMD pathogenesis.

Copaxone (or Glatiramer Acetate) functions to induce suppressor T cells and down-regulates inflammatory cytokines.

A phase I trial was designed to test its safety as well as its efficacy in the prevention of GA progression or conversion of dry AMD to neovascular AMD.

The participant recruitment was suspended with no results posted yet⁽⁷⁾.

Lampalizumab is an intravitreal (IVT) administered recombinant monoclonal antibody fragment directed against complement factor D in the alternative complement pathway.

In phase 2 trial (MAHALO study) monthly Lampalizumab showed 20.4% reduction rate in GA area at 18 months in patients with this form of dry AMD compared with monthly sham treatment⁽⁸²⁾.

Lampalizumab is currently being evaluated in a large, multicenter, phase 3 clinical trial for GA.

Primary outcomes will measure change in GA area after 48 weeks and BCVA up to 2 years after beginning the study^(7,30,83).

LFG 316 is an antibody against the C5 portion of the complement pathway administered IVT⁽⁶⁾.

The phase 2 trial designed to test the safety and efficacy of lower and higher doses of IVT LFG316 using 18 successive monthly injections in patients with GA was completed, but the results have not yet been published as reported on ClinicalTrials.gov.

APL2 is a derivative of compstatin, a small peptide inhibitor of complement factor C3, that is currently in phase 2 trial (FILLY) to assess the safety, tolerability and evidence of activity of multiple IVT injections of APL-2 in subjects with GA as reported on ClinicalTrials.gov.

Zimura (ARC-1905) is a chemically synthesized aptamer that inhibits complement factor C5.

Phase I trial for Other inflammation suppressor is fluocinolone acetonide, a glucocorticoid used as an intravitreal implant.

Although currently there is no drug proven effective, in the next decade some of the research lines will probably be able to find a more effective treatment for the atrophic form of AMD.

Dry AMD evaluated the safety and tolerability of IVT Zimura injection in patients with GA. The study was completed with no results posted, but Ophthotech plans to initiate Phase II/III clinical trial⁽⁷⁾.

Flucinolone acetonide, Eculizumab and Sirolimus (Rapamicina) were also studied, but phase 2 trials failed to demonstrate efficacy^(6,7,30).

 

Lipid Modulators

Lipids are also found in drusen and several studies demonstrate that lipids accumulate at the site of subsequent formation of AMD depositsl⁽⁸⁶⁾.

Barbosa et al. showed that statin use had a statistically significant effect in reducing AMD incidence in participants aged 68 years and olderl⁽⁸⁷⁾.

However the association of statin intake and AMD remains controversial as demonstrated by a 2012 Cochrane database review and to date no trials in GA with this target have been identifiedl⁽⁸⁸⁾.

 

Amyloid

The rationale for this new treatment strategy for AMD is related to the presence of amyloid β in drusen.

GSK933776 is a humanized monoclonal antibody intended to modulate amyloid-β levels in patients with GA secondary to AMD administered via intravenous infusion^(6,7).

A phase 2 study is completed, however the study results have not been posted (ClinicalTrials.gov).

RN6G is a humanized monoclonal antibody intended to prevent accumulation of amyloid β-40 and β-42 and preserve the photoreceptors and the RPE delivered by intravenous injections^(6,7).

A phase 2 study has terminated, but has not been successful, as reported on ClinicalTrials.gov.

MRZ-99030 is a dipeptide containing d-tryptophan and 2-amino-2-methylpropionic acid in clinical development for the topical treatment (eye drops) of glaucoma and AMD^(7,89).

 

Neuroprotection

Neuroprotectants have been shown to protect photoreceptors in in vitro models of retinal degeneration and animal models of glaucoma and retinitis pigmentosa.

The mechanism by which neurotrophic factors, such as ciliary neurotrophic factor (CNTF) and basic fibroblast growth factor, protect photoreceptors is not fully understood, and it should be noted that the relevance of these animal models in AMD is not clear.

The RPE and retina both produce fibroblast growth factor and the retina produces CNTF⁽⁶⁾.

Brimonidine stimulates the production of neurotrophic factors and can protect photoreceptor cells in animal models of retinal degeneration; it has been formulated as an intravitreal implant which delivers the drug to the retina over a 3-month period.

A Safety and Efficacy Study of Brimonidine Intravitreal Implant in Geographic Atrophy Secondary to Age-related Macular Degeneration (BEACON) (phase 2) is ongoing, with progression of the area of GA at month 12 being the primary endpoint (clinicalTrials.gov)⁽⁶⁾.

According to the preliminary results launched at the AAO (American Academy og Ophthalmology) 2016 annual meeting, Brimonidine Intravitreal Implant (22-gauge polymer similar to the one used in ozurdex®, one injection twice a year) caused a significant slowing of disease progression with no adverse effects.

Allergan has completed enrollment of a follow-up study aiming to develop a more potent delivery method(90).

UF-021 (isopropyl unoprostone) is a prostaglandin analog that was investigated in Japan in a phase 2 clinical study in patients with mid- to late-stage retinitis pigmentosa⁽⁶⁾.

A phase 2 trial for dry AMD in United States is completed but the study results have not been published yet (ClinicalTrials.gov).

A sustained-release platform with encapsulated human RPE cells engineered to release CNTF has been developed and was tested in a phase 2 study.

However, no benefit in the progression of lesion expansion was found^(6,91).

Tandospirone (AL-8309B) a selective serotonin 1A agonist that protects the retina from light damage was investigated, but the phase 3 trial (GATE study) was discontinued in 2012 owing to lack of efficacy⁽⁶⁾.

 

Oxidative stress protection
The AREDS trial has reported that daily high doses of antioxidants β-carotene, vitamins C and E and zinc reduced progression to advanced AMD by 25%.

However, the 25% benefit was reported only for the progression to neovascular AMD; there was no benefit in the progression to central GA. Furthermore, the AREDS antioxidant mineral formulation was reported as having no significant effect on the progression of GA in a more recent study⁽⁹²⁾.

The AREDS2 trial assessed the addition of lutein, zeaxanthin, and/or long-chain ω-3 fatty acids to the original AREDS formulation on progression to advanced AMD.

The study completed in early 2013 found that omega-3 fatty acids had no overall additional effect on the formulation, in a median follow-up period of 5 years^(7,93).

Analyses of the AREDS participants over a 12-year period found that participants with the highest omega-3 fatty acids intake were 30% less likely than their peers to develop central GA and neovascular AMD⁽⁹³⁾.

Several studies are ongoing investigating the effects of ω-3 fatty acids, the ratio of omega-3 to omega-6 fatty acids, and lutein and zeaxanthin (AREDS2 study) in patients with AMD^(6,94,95).

OT-551, a small molecule with antioxidant and antinflammatory properties was investigated, however phase 2 study results showed no significant effect on lesion enlargement, retinal sensitivity or total drusen area⁽⁹⁶⁾.

 

Choroidal Perfusion Enhancers

Choroidal circulation provides the nutrition and removes the waste from the retina/RPE.

As a consequence of a reduced choroidal blood flow, metabolic wastes are accumulated in photoreceptor cells, Bruch’s membrane and RPE cells.

Those events can lead to development of GA.

Therefore, improving choroidal blood flow could facilitate the removal of metabolic wastes from RPE, Bruch’ membrane and photoreceptor cells to halt AMD disease progression⁽⁷⁾..

MC-1101 is a topical agent with anti-inflammatory and antioxidative properties shown to increase choroidal blood flow; its intended action is to prevent rupture of Bruch’s membrane.

Phase Ib clinical trial showed that topical instillation of 1% MC-1101 produced no significant cardiovascular or ocular toxicity.

A phase II/III is ongoing, 60 patients will receive topical 1% ophthalmic solution, and be assessed for visual function over 24 months^(6,7).

Trimetazidine was also studied but failed to prevent progression of GA⁽⁶⁾.

 

Stem cells

Replenishing the lost or degenerating RPE cells in GA before the photoreceptors are irrevocably damaged with stem cell-derived RPE cells represents the forefront in the practice of regenerative medicine⁽⁷⁾.

RPE can be differentiated from human embryonic stem cells (hESCs) or human induced pluripotent stem cells (iPSCs).

Both display RPE-like morphology and express typical RPE markers and have the ability to phagocytose photoreceptor segments.

When they were transplanted subretinally into rat or mouse model of RPE insufficiency, the grafted cells were retained and retinal function improved⁽⁷⁾.

Two prospective phase 1/2 studies were done to assess safety and tolerability of subretinal transplantation of hESC-derived RPE in nine patients with Stargardt's macular dystrophy and nine with atrophic AMD.

Transplanted patients were followed up for a median of 22 months.

There was no evidence of adverse proliferation, rejection, or serious ocular or systemic safety issues related to the transplanted tissue.

BCVA, monitored as part of the safety protocol, improved in ten eyes, improved or remained the same in seven eyes, and decreased by more than ten letters in one eye, whereas the untreated fellow eyes did not show similar improvements in visual acuity.

Vision-related quality-of-life measures increased for general and peripheral vision, and near and distance activities in both diseases.

The results of this study provide the first evidence of the medium-term to long-term safety, graft survival, and possible biological activity of pluripotent stem cell progeny in individuals with any disease.

The results suggest that hESC-derived cells could provide a potentially safe new source of cells for the treatment of various unmet medical disorders requiring tissue repair or replacement⁽⁹⁷⁾.

There are currently 2 other clinical trials (phase 1/2) using hESC-RPE cells (MA09-hRPE).

Both are designed to evaluate safety and tolerability of subretinal injection or transplantation of MA09-RPE cells in patients with dry AMD.

Secondary outcomes will measure the mean change of BCVA, autofluorescence photography and reading speed^(6,7).

In another study, clonogenic human central nervous system stem cells (HuCNS-SC) will be evaluated for treatment of dry AMD.

Phase I/II study will investigate the safety and preliminary efficacy of unilateral subretinal transplantation of HuCNS-SC cells in subjects with GA secondary to AMD^(6,7).

 

Conclusion

GA is a devastating blinding disease without any approved or effective treatment currently available.

Numerous clinical trials are ongoing with the purpose of finding a viable solution to prevent or treat the disease.

The AREDS trials show that AREDS formulation reduces the risk of AMD progression by 25%, while transplanted hESC-derived RPE cells show long-term safety, graft survival and possible biological activity shown by improved visual acuity in GA patients^(6,7,30).

Future studies should focus on understanding the pathogenesis of the disease, which remains unclear.

Moreover, the development of advanced imaging system will provide state-of-art tools for analyzing GA pathophysiology and testing new therapeutics⁽⁷⁾.

 

>> References

Introduction

 

Geographic atrophy (GA) and choroidal neovascularization (CNV) represent the advanced forms of age-related macular degeneration (AMD).

GA is defined as a well circumscribed area of atrophy of the retinal pigment epithelium (RPE) where the large choroidal vessels can be seen by ophthalmoscopy and show thinning or absence of the RPE, closure of the choriocapillaris and degeneration of the overlying photoreceptors^(1,2).

Visual loss in GA is due to areas of atrophy of the RPE larger than 175 µm and subsequent loss of tissue in the outer retina (photoreceptors) and choriocapillaris; these areas tend to coalesce progressively and may not affect the fovea until late in the course of the disease (the so-called “foveal sparing”), when visual acuity (VA) finally ensues.

Due to the large paracentral areas of atrophy but preservation of the fovea visual function is often very poor in spite of an apparently good VA.

Patients with a VA of 20/20 may be functionally blind (Figure 1).

Figure 1 - Foveal sparing may allow a 20/20 visual acuity in spite of very severe visual function impairment

 

Relatives of the patients, and even ophthalmologist, have often confused VA for visual function, thus, frequently these patients have felt poorly understood⁽¹⁾.

GA is responsible for one-third of the cases of end stage disease⁽³⁾ and accounts for 20% of cases of severe visual loss due to the disorder⁽⁴⁾.

At the age of 85 years or older incidence of GA is four times the one of CNV. GA is not a benign disease; the atrophy of the external layers may progress at a speed of 1.5-2.6 mm² per year
(Figure 2)⁽⁵⁾.

Figure 2 - GA atrophy progression in one year of the right eye of the same patient (fundus autofluorescence images).

 

It remains a significant challenge because of its potential to cause blindness, its relentless progression and the lack of current effective treatment.

With the progressive increase of the longevity in developed countries GA secondary to AMD represents a true epidemic. Recent research has pointed towards lipofuscin, a fluorophore that accumulates in the RPE, as a triggering agent in the development of atrophy.

Lipofuscin derives mainly from phagocytosed photoreceptor outer segments and accumulates in RPE lysosomes, where one of its many compounds, A2E (N-retinylidene-N-retinylethanol-amine), increases the pH by inhibition of ATPase proton pump function, difficulting its phagocitation and inducing cellular apoptosis^(6,7).

Using a fundus spectrophotometer, Delori et al.^(8,9) were able to visualize lipofuscin due to its autofluorescent properties (when stimulated with blue light in the range of 488 nm, lipofuscin emits a yellow fluorescence).

Current developments allow the clinical in vivo visualization of the distribution of lipofuscin by means of confocal scanning laser ophthalmoscope (cSLO) or specific filters in the fundus cameras. by macular pigment, lutein and zeaxanthin.

A relatively high degree of inter-individual variability and technical difficulties limit the use of the absolute quantification of pixel gray values for longitudinal or transversal studies, and therefore the interpretation of the images is based on qualitative observations, ie decreased (dark), normal or increased (white) FAF, in a similar way to that of conventional fluorescein angiography.

 

Fundus autofluorescence

Fundus autofluorescence (FAF) is a novel, non-invasive method for imaging the fluorescence properties of lipofuscin (and possibly other molecules with a range of absorption and emission spectra close to that of this fluorophore) at the level of the RPE.

Using a commercially available cSLO, the distribution of FAF in the normal eye can be seen in Figure 3. It is characterized by a uniform grayish signal in the fundus and a marked dark appearance in the optic nerve (absence of RPE) and retinal vessels (absorption of fluorescence by hemoglobin and other blood contents).

Figure 3 - Normal fundus autofluorescence: uniform grayish signal in the fundus and a marked dark appearance in the optic nerve (absence of RPE) and retinal vessels (absorption of fluorescence by hemoglobin and other blood contents)

 

The macular area shows progressive diminished signal intensity towards the fovea because of absorption phenomena of short wavelengths A decreased signal is commonly due to RPE atrophy (absence of lipofuscin), an increase in RPE melanin content and absorption from extracellular material anterior to the RPE (intraretinal fluid, fibrosis and media opacities, being cataract a common cause of decreased FAF intensity and poor image quality, specially with cSLO).

On the other hand, an increased signal may be due to lipofuscin accumulation in the RPE (which is the main fluorophore in FAF imaging), presence of other fluorophores not in the RPE (drusen -including those in the optic nerve head-, older hemorrhages), lack of absorbing material and artifacts⁽¹⁰⁾.

FAF imaging in patients with GA is characterized by a decreased signal with sharp borders corresponding to the area of atrophy on conventional retinography (Figure 4).

Figure 4 - Fundus autofluorecence of patients with GA secondary to dry AMD.
Atrophic areas show hipoautofluorescence. Borders of the lesion show hyperautofluorescence due to the accumulation of lipofuscin

 

However, many patients show increased FAF at the borders of atrophy (Figure 5), which has been histopathologically confirmed as areas of increased lipofuscin-filled RPE cells between atrophic and normal retina.

Figure 5 - Surrounding areas of accumulation of lipofuscin at the junction area show increased autofluorescence.
In pseudocolor 3D areas of higher accumulation of lipofuscin are seen as yellow spikes

 

This finding has not been identified by any other imaging modality.

Holz et al. showed in a longitudinal study that atrophy developed selectively in the junction areas of increased FAF⁽¹¹⁾ but not elsewhere, a finding that could not be confirmed on a small sample by Hwang et al.⁽¹²⁾.

Based on this information, the FAM (Fundus autofluorescence in age-related macular degeneration) study initially described 8 patterns of FAF in the junction zone of GA⁽¹³⁾, which were later modified to incorporate a nineth pattern (Figure 6)⁽¹⁴⁾.

”Reproduced with permission from Rightslink (Elsevier)”.

Figure 6 - FAM-Study classification of fundus autofluorescence patterns of geographic atrophy in age-related macular degeneration⁽¹⁴⁾.

 

According to FAF in the junction zone of atrophy, eyes are classified as none (when there is no increased FAF at the borders of the GA), localized (focal, banded, patchy) and diffuse (fine granular, branching, trickling, reticular and fine granular with punctuated spots)⁽¹⁴⁾.

The relevance of these patterns relays in the fact that they may represent different phenotypic manifestations of the disease.

It has been shown in natural history studies that rates of growth differ between subtypes of FAF and that a strong correlation exists between FAF pattern and progression of atrophy in GA^(14,15).

In the FAM study⁽¹⁴⁾ 195 eyes of 129 patients with GA were followed a median of 1.80 years and classified according to FAF pattern at baseline.

Those without abnormal FAF at the borders of the lesion experienced the slowest progression over time (0.38 mm2 /year, n = 17) compared to those with the focal (0.81 mm2 / year, n = 14) and diffuse (1.77 mm2 / year, n = 112) subtype (p<0.0001).

FAF was more strongly associated with GA growth than other classic risk factors, such as size of baseline atrophy, smoking, age or family history.

Nowadays current devices of autofluorescence equipped with semiautomated software allow quantification of total area of GA and measurement of its progression in time (Figure 7).

Figure 7 - Semiautomated software allows quantification of total area of GA and measurement of its progression in time (Spectralis Heidelberg Retinal Angiograph/OCT; Heidelberg Engineering, Heidelberg, Germany)

 

Furthermore, studies using fine-matrix mapping⁽¹⁷⁾ and SLO microperimetry⁽¹⁸⁾ have found impaired rod photoreceptor function and photopic sensitivity respectively in areas of increased FAF in the junction zone, which underscores abnormalities associated with increased fundus autofluorescence.

Taken together, these results suggest that presence of increased FAF at the borders of GA is associated with a greater rate of progression of atrophy and that different patterns of FAF may reflect differences at the cellular and molecular level that may explain the different evolution of the disease process.

This information is relevant for understanding its physiopathology, natural history and to evaluate future therapeutic strategies.

 

Optical coherence tomography

Recent development of high resolution, high speed spectral domain optical coherence tomography (SD OCT) improves the visualization of the RPE, outer and inner segment of the photoreceptors and external limiting membrane over previous time-domain based technology.

Given the aforementioned reasons, simultaneous imaging techniques that combine SD-OCT and FAF are highly desirable for the evaluation of the atrophic and junction areas of patients with GA (Figure 8).

Figure 8 - High resolution optical coherence tomography correlated with the autofluorescence image.

 

Currently there are instruments that fulfill these criteria and allow the study of the correlation between areas of increased or decreased autofluorescence with the morphologic changes detected in the external retina by SD OCT.

SD OCT in atrophic areas of patients with GA show retinal thinning due to atrophy of the RPE and disappearance of the external retina, that includes inner and outer segment of the photoreceptors and, very frequently, the external limiting membrane⁽¹⁹⁾; in the severe forms, the outer nuclear layer may be no longer identifiable and therefore the outer plexiform layer may be in direct contact with Bruch´s membrane.

The thinned retina permits the deeper penetration of light and a corresponding increased signal from the choroid within the atrophic area (Figure 9).

SD OCT is also useful to identify absence of exudative signs (intraretinal or subretinal pockets of fluid, RPE detachments, maintenance of the continuity of Bruch´s membrane)⁽²⁰⁾.

Several abnormalities have been found with SD OCT in the junction zone^(19,21,22), such as disruption of external retinal layers with different shape of band endings, disappearance of the external limiting membrane and/or of the retina encompassing inner segments of the photoreceptor layer to Bruch´s membrane at the same or at a different transverse planes, small elevations of RPE thought to represent subRPE deposits or increased distance between inner and outer photoreceptors segment and RPE, presumably due to debris between these layers.

Smooth margins with no structural changes from normal to abnormal retina exist in the junction zone when there is no FAF abnormality, which underscores the significance of abnormal FAF⁽²²⁾ (Figure 9).

Figure 9 - SD OCT in normal retina (up left ). In the lower images, SD OCT of the junction area between preserved retina and the geographic atrophy. Atrophic areas show retinal thinning due to atrophy of the RPE and disappearance of the external retina that includes inner and outer segment of the photoreceptors and, very frequently, the external limiting membrane(16); The outer nuclear layer may be no longer identifiable and therefore the outer plexiform layer may be in direct contact with Bruch´s membrane. The thinned retina permits the deeper penetration of light and a corresponding increased signal from the choroid within the atrophic area.

 

In summary, precise quantification of the GA and its progression by fundus autofluorescence imaging and the detailed morphologic study of the external retina by the high resolution SD OCT allow to show the relative slow progressing abnormalities of the outer retinal layers in dry AMD.

This may prove essential for prognostic and interventional strategies, in order to detect potential benefit by slowing down the degenerative process or perhaps detect signs of RPE and/or photoreceptors rescue.

 

>> References

Introduction

Over 100 years ago, Oeller described for the first time the presence of anastomoses between the retinal and choroidal circulations in eyes with disciform scars⁽¹⁾.

Later, these were recognized in association with laser photocoagulation⁽²⁾, radiotherapy⁽³⁾, chorioretinal inflammatory diseases⁽⁴⁾ and parafoveal telangiectasias⁽⁵⁾.

The interest in this condition even led to anatomopathological studies in disciform scars from late age-related macular degeneration (AMD)⁽⁶⁾.

In 1992, Hartnett et al.⁽⁷⁾ described nine cases of retinal neovascularization, to which they referred as “deep retinal vascular anomalous complex”.

In 2000, Slakter et al.⁽⁸⁾ described chorioretinal anastomosis in eyes with pigment epithelial detachment (PED) and indocyanine green angiography (ICG-A) hot spots but it was Yannuzzi et al.⁽⁹⁾ who coined the term retinal angiomatous proliferation (RAP) when in 2001 the authors described chorioretinal anastomoses as a neovascular proliferation originating deeply in the retina.

In 2001, Yannuzzi et al. described chorioretinal anastomosis as neovascular proliferation with origin in the retina, and proposed the designation of RAP – retinal angiomatous proliferation⁽⁹⁾.

Still, several authors^(10-13) maintained the designation of chorioretinal anastomosis, proposing a choroidal origin for this clinical entity.

In 2008, Yannuzzi et al.⁽¹⁴⁾ escribed 5 cases of RAP with the neovascular complex originating in the choroid instead of the retina, and proposed that RAP should be called type 3 neovascularization instead. The two terms have been used interchangeably ever since⁽¹⁵⁾.

However, RAP is still the most common designation.

 

Classification

Based on the presumed origin and evolution of the neovascular process, a three-stage classification was proposed by Yannuzzi et al.⁽⁹⁾, to characterize the clinical manifestations and progressive changes observed in RAP. However, it is often clinically difficult, if not impossible, to determine exactly when progression occurs from one stage to another.

 

Stage I – Intraretinal neovascularization.

The presence of small intraretinal hemorrhages is a hallmark of RAP and a very useful sign for its early clinical diagnosis. A small elevation of the inner/intermediate retina caused by angiomatous tissue may be observed under the slit lamp; this elevation may extend tangentially, assuming a telangiectatic appearance. Dilated retinal vessels may perfuse and drain the intraretinal neovascularization and form retino-retinal anastomoses (RRA)⁽⁹⁾.

The fluorescein angiography (FA) of the intraretinal neovascular complex shows a focal hyperfluorescent area in front of the retinal pigment epithelium (RPE), mimicking classic choroidal or, more frequently, occult neovascularization. ICG-A may reveal a hot spot with staining and leakage.

 

Stage II – Subretinal Neovascularization

ovascular complex shows a focal hyperfluorescent area in front of the retinal pigment epithelium (RPE), mimicking classic choroidal or, more frequently, occult neovascularization. ICG-A may reveal a hot spot with staining and leakage.

 

Stage II – Subretinal Neovascularization

Involvement of the subretinal space with localized neurosensory retina detachment, edema and retinal hemorrhages at the edges may already be observed in colour fundus photography. Usually, a clear RRA can be seen, with a perfusing retina arteriole and draining venule communicating within the core of the subretinal neovascularization (Figure 1).

Stage III – Choroidal Neovascularization

Figure 1. RAP lesion – Colour fundus photography of a RRA in a patient with RAP.

 

In stage II, an associated serous PED can be observed in 94% of the eyes⁽⁹⁾.

FA may reveal a well-limited hyperfluorescence area enclosed in the diffuse leakage area associated with the PED, identical to that occurring on a minimally classic choroidal neovascular membrane. ICG-A clearly establishes stage II as the presence of a hot spot with leakage, as well as a hyperfluorescent area associated with serous PED.

Leakage may not be revealed by FA, probably because fibrin from retinal exudates cannot be impregnated with fluorescein, opposing to what occurs with ICG-A⁽¹⁶⁾.

In fact, FA is rarely useful in the differential diagnosis between classic, occult or minimally classic membranes and early RAP lesions (stages I and II)⁽⁹⁾.

On the other hand, ICG-A reveals a slow-growing extra, juxta or subfoveal hot spot, even in asymptomatic patients with RAP stages I and II.

 

Stage III – Choroidal Neovascularization

Stage III is established when the clinical and angiographic examinations can clearly demonstrate the presence of choroidal neovascularization (CNV), sometimes with the appearance of a vascularized PED.

According to Gass⁽¹¹⁾, RAP would progress in 5 stages instead of 3, easily identifiable by FA and ICG-A:

I) Pre-clinical stage - Atrophy of the outer retina, with retinal capillaries moving closer to a choroidal neovascular complex located below the RPE - type 1 CNV – and no clinical signs of chorioretinal anastomosis. ICG-A would be necessary to identify type 1 neovascularization.

II) Early clinical signs - Involvement of the subretinal space with localized neurosensory detachment of the retina, oedema and retinal haemorrhages at the edges may already be observed in fundus colour photography.

Anastomosis between dilated capillaries of the deep retina and the choroidal neovascular complex is associated with small intraretinal hemorrhages, which would constitute the first clinical sign of chorioretinal anastomosis.

This stage may occur weeks or months before stage III, where subretinal CNV is already observed.

III) Proliferation of CNV over the RPE - subretinal neovascularization – type 2 CNV.

IV) Appearance of serous PED - caused by activation of newly formed subepithelial vessels.

V) Mixed neovascularization - piggyback-type neovascularization, with two levels – type 1 and type 2 with cicatricial disciform lesion, making chorioretinal anastomosis visible.

Stage III in the Gass classification⁽¹¹⁾ corresponds to stage I in the Yannuzzi classification⁽⁹⁾, with stages IV and V in the Gass classification⁽¹¹⁾ corresponding to stages II and III, respectively. Currently, the Yannuzzi classification⁽⁹⁾ is the most widely used.

Recently, an optical coherence tomography (OCT) based classification was developed by Su et al.⁽¹⁷⁾ According to the authors, a precursor stage consisting of punctate hyperreflective foci in the outer retina is the first OCT finding:

Stage 1 consists of a larger intraretinal hyperreflective lesion associated with cystoid macular edema but without outer retinal disruption.

Stage 2 is notable for outer retinal disruption that occurs with RPE disruption in most of the cases. Cystoid macular edema is generally present.

Stage 3 is defined by an intraretinal hyperreflective lesion that extends through the RPE to vascularize a drusenoid/serous PED. Cystoid macular edema is generally present.

 

Clinical Findings

The presence of small hemorrhages, soft drusen, and pigmentary changes is well documented in eyes with RAP and in unaffected fellow eyes, both in caucasian and in asian populations^(9,18-21).

Marques et al.⁽¹⁹⁾ conducted a quantitative evaluation of the fundoscopic features of fellow eyes of RAP and compared it with a cohort of fellow eyes of typical exudative AMD.

The authors reported that the total number and area of drusen are significantly smaller in fellow eyes of RAP than in fellow eyes of typical AMD.

Aside from soft drusen, a high frequency of reticular pseudodrusen in eyes with RAP has also been described⁽¹⁸⁾.

The presence of telangiectasias, microaneurysms or punctiform retinal hemorrhages and RRA associated with intraretinal fluid is highly suggestive of RAP (Figure 2), as is the sudden disappearance of a retinal vessel that appears to have moved deeper^(8-14,19,22).

Hard exudates usually accumulate around the central retinal neovascular complex, bordering the PED in stage III RAP lesions.

The advent of OCT, fundus autofluorescence (FAF) and infrared (IR) imaging allowed for a better characterization of the structural changes associated with the disease⁽²³⁾. Figures 2, 3, 4, 5 and 6 show clinical examples of patients with RAP who underwent multimodal retinal imaging.

Figure 2. Multimodal retinal imaging of a RAP lesion. (A) Colour fundus photography displaying intraretinal hemorrhages, hard exudates and neurosensory retinal detachment. (B) FA shows a retinal hiperfluorescent spot apparently at the end of one retinal vessel, an intraretinal hemorrhage, neurosensory retinal detachment and PED. (C) Early phase ICG-A with an intraretinal hot spot (angiomatous proliferation) over diffuse choroidal hyperfluorescence. (D) Late phase ICG-A reveals a subfoveal hiperfluorescent plaque (subretinal neovascularization).

 

Figure 3. ICG-A of a RAP lesion. A RAP lesion with RRA, an apparently intraretinal angiomatous mass and a serous PED, imaged with ICG-A.

 

Figure 4. Multimodal retinal imaging of a RAP lesion. Colour fundus photography (A) with intraretinal hemorrhages, hard exudates and neurosensory detachment. Red-free imaging (B) with two juxtafoveal and extrafoveal small hemorrhages. FA (C) shows two juxtafoveal hyperfluorescent spots, neurosensory detachment and PED. Late ICG-A (D) reveals two juxtafoveal hyperfluorescent hot spots. FA (E) and ICG-A (F) after treatment with laser photocoagulation showing resolution of the exudation.

 

Figure 5. Bilateral RAP lesion: Colour fundus photography (A and B) and red-free images (C and D) clearly show small retinal hemorrhages and lipid exudation. (E and F) Late ICG-A of both eyes with a hot spot and subfoveal hypofluorescence (serous PED). Stratus OCT (G and H) reveals serous PED, neurosensory retinal detachment and intraretinal fluid in both eyes.

 

Figure 6. RAP lesion imaged with spectral-domain (SD)-OCT. SD-OCT of a RAP lesion with an area of intraretinal neovascularization (green arrow) in the deep retina with a RPE detachment (white arrow). Intact and ruptured portions of the Bruch’s membrane and subretinal fluid (yellow arrow) can also be seen.

 

The recent introduction of OCT angiography (OCTA) in clinical practice allowed for an unmet morphological characterization of RAP lesions.

Qualitative and quantitative analyses of type 3 neovascular complexes can be performed using OCTA^(24,25).

Querques et al.⁽²⁶⁾ evaluated the features of RAP with this new imaging modality and found lesions emerging from the deep capillary plexus, forming a clear, tuft-shaped, high-flow network in the outer retinal segment in all eyes, abutting in the sub-RPE space.

The authors also reported a small, clew-like lesion present in the choriocapillaris and that, in some cases, this clew-like lesion seemed to be connected to the choroid through a small-caliber vessel.

Compared with conventional imaging, OCTA may improve the detection and delineation of vascular changes occurring in RAP⁽²⁵⁾.

Further longitudinal studies with OCTA may bring further insights on the pathophysiology of RAP and establish OCTA imaging as an additional biomarker to guide the treatment and monitoring of these lesions.

 

Differential diagnosis

Differential diagnosis is mandatory for:

• Other forms of CNV, namely occult CNV
• Polypoidal choroidal vasculopathy (PCV)
• Macular telangiectasia (MacTel)

Other forms of CNV with ICG hot spots (occult CNV) and PCV are two important diagnoses that can be easily confounded with RAP.

Small intraretinal hemorrhages, sometimes punctiform, in patients with soft drusen, are very typical in RAP, as are telangiectasias and RRA.

On the contrary, retinal hemorrhages in PCV are normally larger, with round reddish-orange nodules usually being observed in the fundus⁽²⁷⁾.

OCT is a useful tool in the differential diagnosis between RAP, PCV and occult CNV^(28,29).

In RAP, intraretinal hyperreflectivity may be observed, corresponding to angiomatous proliferation associated with intraretinal fluid and/or RPE detachment^(30,31).

In PCV, polyps appear in OCT as abrupt protrusions from the RPE/Bruch’s membrane band, often associated with neurosensory detachment^(27,28).

The double layer sign, RPE irregularity and a notched PED are hallmarks of PCV on OCT⁽³²⁾.

MacTel is a condition involving dilation of retinal capillaries located near the fovea, in one or both eyes.

RPE hyperplasia may also occur, with refractive punctiform deposits and macular leakage being observed in FA.

Migration of one or more venules to the deep retina can also be found⁽⁵⁾.

Also, anastomoses between retinal vessels and the choroidal circulation have been described, as well as choroidal new vessels.

The most significant differences are the fact that MacTel are not associated with serous PED, the RPE is healthier and CNV occurs less frequently^(5,22).

 

Natural history

Natural history may be highly variable and in many aspects similar to that of other neovascular AMD lesions.

RAP has been coined a bilaterally aggressive disease with predictable symmetry^(9,33,34).

In fact, the vasogenic potential associated with RAP has been highlighted by several group^(18,33,35), reporting annual and cumulative rates of neovascularization in fellow eyes far exceeding the typical forms of exudative AMD.

Yannuzzi et al.⁽⁹⁾ reported an average of 15 months until the development of neovascularization in their cohort of fellow eyes of RAP.

In a cohort of 52 fellow eyes of RAP, Gross et al.⁽³⁴⁾ reported a cumulative incidence of neovascularization of 40% at 12 months, 56% at 24 months and 100% at 36 months.

Campa et al.⁽³³⁾ have also evaluated the cumulative incidence of neovascularization in a population of 37 fellow eyes of RAP but found a significantly inferior conversion rate (36.4% at 36 months).

Recently, a small cohort (20 fellow eyes of RAP) in a Japanese population(18) identified an incidence of neovascularization of 50% at ~49 months of follow-up (min 24; max 108 months).

RAP was historically associated with a poor natural history and high relapse rates. Kuhn et al.⁽¹⁰⁾ studied 22 eyes and observed structural evolution towards classic membranes, signs of RPE rupture and fibrous scars in 36.4%, 4.6% and 31.8% of cases, respectively. The authors described a decrease in visual acuity (VA) in 77% of cases. Similarly, Hartnett et al.⁽³⁶⁾ reported that 9/11 patients they evaluated developed legal blindness.

A study by Viola et al.⁽³⁷⁾, where the authors investigated the natural history and visual outcome in 14 eyes with untreated RAP for an average of 20 months, found that, by the time of the final examination, VA had decreased to ≤20/200 in 11 eyes (69%), and 5/14 patients (36%) were legally blind. However, while several treatment approaches proved unsuccessful in the past^(13,38), the blockade of vascular endothelium growth factor (VEGF) has considerably changed the paradigm of this condition, with recent studies showing similar functional results to typical exudative AMD^(39,40).

An important feature of the natural history of RAP lesions is geographic atrophy (GA).

It was previously thought that GA in RAP was secondary to the treatments used^(41,42).

However, in a recent study using FAF, McBain et al.⁽⁴³⁾ found high rates of GA in RAP, independently of treatment modalities (phtotodynamic therapy or anti-VEGFs) and other authors have reported similar results^(39,44).

This finding may be a consequence of the reduced choroidal perfusion and reduced choroidal thickness in RAP, which can be the reason behind the high vasogenic potential reported with this disease^(21,43,45).

 

Epidemiology

RAP is now considered a well-established neovascular phenotype of AMD, with distinct clinical and epidemiologic features, most likely due to genetic differences still not completely understood⁽⁴⁶⁾ .

Environmental risk factors such as age and arterial hypertension have been associated with the development of both RAP and typical exudative AMD⁽⁴⁷⁾ .

Even though the role of arterial hypertension is still matter of debate, it is now widely accepted that RAP patients are significantly older than typical exudative AMD patients – average age of 79 vs. 76 years^{(6,8,9,14,47)}.

The average and median ages of a series of 108 patients studied by Yannuzzi et al.⁽⁹⁾ were 80 and 81 years, respectively.

Women are more frequently diagnosed with RAP, comprising 64.7% to 71% of the affected individuals^(9-10,47).

In caucasian populations, RAP is estimated to represent 5-28% of wet AMD cases^{(9,15,34,40,48,49)}.

Using the ImageNet 1024 videoangiography system, our team observed a prevalence of 9.4% in a consecutive series of 563 patients⁽¹³⁾.

Studies in asians suggest a considerably lower prevalence^(35,50) and there are no known reports of RAP in blacks⁽¹⁴⁾.

Prevalence appears to be greater in hyperpigmented eyes^(8,9,51).

Rapid progress in identifying genetic risk factors for AMD susceptibility has been made over the last few years, including the identification of two major loci at chromosome 1q32 and 10q26^(46,52).

Even though several studies have evaluated genetic risk factors for exudative AMD, very few have specifically addressed the genetic alterations found in RAP^(46,53,54).

Wegscheider et al.⁽⁵³⁾ and Caramoy et al.⁽⁴⁷⁾ suggest that CFH gene variants are less frequent in patients with RAP than in other forms of exudative AMD, while ARMS2 variants, namely the single nucleotide polymorphism (SNP) A69S (rs10499024), appears to have a stronger association with this neovascular phenotype^(46,54,55).

These findings stress the importance of detailed phenotyping in AMD in order to identify genetic biomarkers for the distinct AMD subtypes.

 

Anatomopathology

In 2000, Lafaut et al.⁽¹⁶⁾ studied six “deep retinal vascular anomalous complex” lesion specimens from retinal translocation surgery in eyes evaluated with FA and ICG-A.

The deep vascular anomalous complex was located in front of the RPE, thus not representing CNV. Nodular fibrovascular complexes surrounded by a ring of diffuse drusen – basal linear deposits under electron microscopy –, pigment epithelium and an amorphous fibrous material containing photoreceptor outer segment debris were observed.

Even though the lesion was not covered by RPE, the latter was preserved around the lesion. The amorphous material did not cover the retinal surface of the membrane, which adhered directly to the outer nuclear layer in 50% of cases.

Anastomoses between fibrovascular nodules and the choroidal circulation were only observed in cases showing isolated disciform scars.

Avascular fibrocellular membranes were observed on the inside of Bruch’s membrane and, in three cases, on the choroidal side of diffuse drusen.

Fibrin was present in affected retinas; choroidal anastomoses were not observed. Still, the presence of retino-choroidal anastomoses could not be entirely excluded, since the specimens may not have included the entire lesion.

In a histopathological study performed in nine neovascular lesions classified as RAP, Shimada et al.⁽⁵⁶⁾

found only intraretinal neovascularization in stage II cases.

In stage III cases, these authors found choroidal and intraretinal neovascularization; they concluded that their findings were in agreement with the classification proposed by Yannuzzi et al.⁽²⁹⁾.

Furthermore, the authors identified the expression of VEGF (in intraretinal neovascularization and in the RPE), CD68-positive macrophages (in the neovascularization area) and expression of hypoxia-inducing factors (HIF) alpha 1 and alpha 2 in neovascular endothelial cells.

They hypothesized that intraretinal neovascularization would appear before the occurrence of CNV and be associated with ischemia and increased expression of VEGF and inflammatory factors.

Monson et al.⁽⁵⁷⁾described the histopathological characteristics of RAP in an 87-year-old woman.

The images obtained by fundus examination and FA were histopathologically consistent with a neovascular intraretinal angiomatous complex without subretinal pigment epithelial neovascularization.

Gass et al.⁽¹¹⁾ described chorioretinal anastomoses and atrophy of the outer nuclear layer in a pre-clinical case, with outer retinal capillaries moving close to a CNV focus, having proposed a choroidal origin for RAP, instead of the retinal origin initially proposed by Yannuzzi et al.⁽⁹⁾.

The origin of the neovascularization process (choroidal or retinal) in RAP remains a controversial issue.

Regardless, the anatomopathological studies helped elucidate that RAP lesions are associated with ischemia, age-related macular alterations, increased VEGF production and macrophage expression, similar to other forms of AMD-related CNV.

 

Pathogenesis

Kuhn et al.⁽¹²⁾studied the evolution of RAP in the second eye of two patients.

The authors assumed the existence of two asymmetrical membranes, a smaller retinal membrane – the angiomatous anomaly – over a larger membrane – the choroidal membrane, secondary to failure of a diseased RPE, unable to modulate and inhibit neovascular factors.

Even though the exact mechanism and origin of RAP remains elusive, the important role of VEGF cannot be overemphasized. The presence of serous PED may increase hypoxia by moving the retina even further from the choroid⁽³⁶⁾.

RPE decompensation is probably behind the deposition of amorphous materials in the subretinal space (under the basal lamina) and consequent thickening of Bruch’s membrane, all contributing to a decrease in inner retinal oxygenation.

Choriocapillaris atrophy and hypoxia may lead to intense neovascular activity, with reactive synthesis of growth factors, such as VEGF⁽⁵⁸⁾.

Overexpression of VEGF is sufficient to produce intraretinal neovascularization and subretinal neovascular membranes in animal and human models^(58,59).

The relatively good response of these lesions to ranibizumab, bevacizumab and aflibercept confirms the central role of VEGF in the pathogenesis of RAP lesions.

As described above, Yannuzzi et al.⁽⁹⁾ initially hypothesized that the initial neovascular process in RAP occurs in the deep retina and consists of 3 stages. Gass⁽¹¹⁾ contested a retinal origin for anastomosis, based on the following facts:

1. no communication between retinal vessels and the choroid could be found in surgical specimens;
2. the location of vascular anomalous complexes on the outer nuclear layer of atrophic retinas is more compatible with a choroidal origin hypothesis;
3. although the macular retina is not excised in surgery, the retinal neovascular complex disappears and no macular hole is left, which contradicts its retinal origin.

It is clear that no definitive sequential histopathological evidence exists to support an intraretinal versus a choroidal origin for RAP lesions.

In 2008, Yannuzzi et al.⁽¹⁴⁾ re-evaluated the early stages of RAP lesions in five eyes, using FA, ICG-A, time-domain-OCT and frequency-domain -OCT, having concluded that the initial lesion may have its origin not only on the deep retinal capillaries but also in the choroid.

They described RAP as “type 3 neovascularization”, a type of neovascularization with preference for the retina, displaying the following manifestations:

1. Focal neovascular proliferation from the deep retinal layer (originally RAP);
2. Intraretinal neovascular extension from underlying occult type I CNV (originally occult chorioretinal anastomosis);
3. De novo breaks in Bruch’s membrane with neovascular infiltration into the retina.

The term type 3 neovascularization helps to resolve the various conflicting theories and descriptions behind the origin of RAP lesions: neovascularization in RAP may originate not only from the deep retinal capillaries but also from the choroid.

However, the main issue is not the intraretinal or choroidal origin of neovascularization but its unique characteristic of presenting two neovascular foci: one located in the deep retina and the other at a choroidal level.

 

Treatment

Before the anti-VEGF era, several treatment methods that included direct laser photocoagulation of the vascular lesion, laser photocoagulation of the feeder retinal arteriole, surgical ablation, scatter grid-like laser photocoagulation, photodynamic therapy (PDT), transpupillary thermotherapy, intravitreal triamcinolone acetonide (IVTA) and combined regimens were used, yielding at most VA stabilization, marginally better VA or short-term VA improvements^{(12,13,60,61)}.

 

Thermal laser Photocoagulation

Kuhn et al.⁽¹⁰⁾ photocoagulated hot spots in 28 eyes, reporting “occlusion” in 25% of cases, including one eye with recurrence.

No occlusion was observed in the remaining 75% of cases, despite multiple treatment sessions.

Treatment caused tearing of the RPE in 45% of cases and VA decreased in 86% of eyes.

The low success rate might have been related to inadequate laser penetration – given the serous PED height or absorption of radiation by the sub-RPE fluid –, incorrect location or the presence of multiple afferent vessels.

Slakter et al.⁽⁸⁾ confirmed this poor prognosis, particularly in serous PED cases. Occlusion was not possible in 86% of eyes and progressive evolution to a disciform scar was noticed.

Clinical experience showed that some extrafoveal stage I lesions (according to the Yannuzzi classification⁽⁹⁾) are amenable to laser photocoagulation⁽⁶²⁾.

In a series of 108 eyes with RAP, Bottoni et al.⁽⁶⁰⁾ achieved full obliteration of chorioretinal anastomosis in 57.1% of these lesions.

However, the risk of complications needs to be evaluated and careful follow-up is mandatory, given the high rates of persistence and recurrence.

 

Surgical ablation

Borrillo et al.⁽⁶³⁾performed vitrectomy with detachment of the posterior hyaloid face, surgical section of afferent arterioles and draining veins and membrane excision in 4 eyes with RAP and PED – stage II – with resolution of intraretinal edema, collapse of the PED after 7-10 days and an increase in the average VA (20/200 pre-operative vs. 20/70).

Shimada et al.⁽⁵⁶⁾ excised the neovascular complex in 9 eyes from 8 patients – stages II and III.

VA remained stable in the post-operative period; however, significant destruction of the RPE and the choriocapillaris occurred in patients with serous PED.

The authors concluded that surgery may stabilize VA in stage III but is not indicated in stage II.

Surgical section of the afferent vessel associated with intravitreal injection of triamcinolone was performed in one eye, with resolution of exudation and VA of 20/320 at 6 months⁽⁶⁴⁾.

Nakata et al.⁽⁶⁵⁾ found surgical ablation (even combined with PDT) not useful for the treatment of RAP lesions, given the high frequency of reperfusion from retinal inflow vessels associated with this procedure.

Shiragami et al.⁽⁶⁶⁾ observed recurrence of RAP lesions in all 7 cases treated, 2 to 13 months after surgical ablation.

Surgical ablation of RAP is no longer recommended for RAP lesions, considering the better outcomes achieved with other treatment modalities, such as antiangiogenic drugs.

 

Photodynamic therapy (PDT) alone or in combination with intravitreal triamcinolone acetonide (IVTA)

Studies of PDT in monkeys with neovascular complexes and chorioretinal anastomoses, showed occlusion of the neovascular complex but not of the anastomoses⁽⁶⁷⁾.

The authors hypothesized that the anastomoses would be responsible for repermeabilization of the neovascular complex.

Kusserow et al.⁽⁶⁸⁾ treated 6 eyes with predominantly classic membranes and chorioretinal anastomoses without success.

No improvements in VA were observed for any of the treated eyes and membranes continued to grow.

Our group reported stabilization or improvement in VA after PDT in 73.3% of eyes (<3 lines loss) at 12 months⁽¹³⁾, thus representing a better outcome than the natural evolution(36).

However, a significant VA decline was observed in the second year, mainly due to recurrence⁽¹²⁾.

Boscia et al.⁽⁶⁹⁾ treated 13 eyes with PDT, having concluded that PDT would only be useful in cases where serous PED represents less than 50% of the lesion.

In 2006, these authors referred that early treatment of eyes with smaller lesions using PDT with verteporfin potentially led to a beneficial effect on vision, whereas it might worsen the natural progression of larger lesions, with most eyes undergoing enlargement, disciform transformation or RPE tear⁽⁷⁰⁾.

Reported short-term results of non-randomized studies on RAP lesions treated with PDT and IVTA^(41,71,72) revealed apparently better VA outcomes and/or a reduced number of treatment sessions compared to PDT alone.

However, recurrences were also frequent^(42,61).

Krebs et al.⁽⁷³⁾ found no significant differences between the PDT monotherapy group and the combined PDT and IVTA group regarding evolution of distance VA, retinal thickness and lesion size, having concluded that new therapeutic strategies might be required in RAP lesions, probably including therapy with antiangiogenic agents.

 

Antiangiogenic agents

Better visual outcomes can be achieved by treating RAP lesions with intravitreal ranibizumab^{(39,40,44,74,75)}, bevacizumab^(76,77) and most recently aflibercept⁽⁷⁸⁾.

Data from clinical trials suggests that the response of RAP lesions to CNV treatments may be similar to that of other variants of neovascular AMD⁽⁷⁹⁾.

Improved VA and short-term edema reduction or elimination, were also observed in combined treatments using PDT and ranibizumab^(44,80) or bevacizumab^(81.82).

However, no superiority of combined treatment with PDT and ranibizumab or bevacizumab has been demonstrated.

The Comparison of Age-Related Macular Degeneration Treatments Trials (CATT) study followed up a large cohort of patients with treatment-naïve exudative AMD eyes, who received randomly assigned ranibizumab or bevacizumab for 2 years. The cohort included eyes with classic and occult CNV as well as RAP.

The authors recently published a comparison between eyes with and without RAP regarding the 2-year visual and morphologic outcomes⁽³⁹⁾.

This study showed a rapid improvement in VA in eyes with RAP within the first 3 months of intravitreal anti-VEGF therapy that continued to improve and then stabilized during the first year.

However, in the second year, VA began to decline modestly such that there was no statistically significant difference between eyes with or without RAP in overall VA gain at the end of the second year.

The authors noticed that fewer injections were needed to treat RAP than the other types of CNV but the rate of GA was higher in the RAP group⁽³⁹⁾.

A recently published study⁽⁴⁴⁾ on the long-term results (≥36 months of follow up) of PRN intravitreal ranibizumab for the treatment of RAP lesions in clinical practice (n=79), showed that ∼40% of the eyes had stable or improved VA at the end of follow-up, and about one fifth of the individuals preserved reading vision.

Nevertheless, about 60% of the patients presented with some degree of GA at the last visit.

The presence of subretinal fluid at baseline was found to positively correlate with a favorable visual outcome, thus appearing to be an important prognostic factor for functional improvement.

 

>> References

Introduction

Polypoidal choroidal vasculopathy (PCV) was described for the first time in 1982⁽¹⁾.

Different names were proposed like posterior uveal bleeding syndrome⁽²⁾ or multiple recurrent retinal pigment epithelium (RPE) detachments in black women⁽³⁾.

It has a characteristic imaging expression on indocyanine green angiography (ICG), peculiar characteristics in optical coherent tomography (OCT) and apparently different responses to treatments when compared to occult or classic choroidal neovascularization.

Diagnosis is based on ICG and confirmed with characteristic fundus and OCT changes.

The primary abnormality was initially thought to involve the choriocapillaris, with the characteristic lesion consisting of a branching vascular network of vessels terminating, in one or more aneurismal dilatations, known as polyps.

Clinically a reddish orange, nodular or spheroid, polyp-like structure may be observed.

The natural course of the disease often follows a remitting-relapsing course, and clinically, it is associated with chronic, multiple, recurrent serosanguineous detachments of the RPE and neurosensory retina. In the majority of cases, there is significant long-term visual morbidity caused by fibrosis, RPE atrophy and secondary choroidal neovascularisation.

Cheung et al. reported that 50% of patients had reduction of vision to worse than 20/200 after mean follow up of 3 years⁽⁴⁾.

A more recent knowledge has expanded the spectrum of PCV allowing us a clearer characterization of PCV. However, the classification of PCV as distinct subtype of exudative age-related macular degeneration (AMD) easily differentiated from other diseases and other subtypes of choroidal neovascularization associated with AMD is still controversial.

Polypoidal choroidal vasculopathy has also been described in different pathological conditions including central serous chorioretinopathy⁽⁵⁾, circumscribed choroidal hemangioma⁽⁶⁾; melanocytoma of the optic nerve⁽⁷⁾; pathological myopia and staphyloma⁽⁸⁾; or choroidal osteoma⁽⁹⁾. PCV seems to behave more like a neovasculopathy occurring in a variety of different diagnoses than a distinct abnormality of the inner choroidal vasculature⁽¹⁰⁾

Most recently, Freund and others have described PCV as part of the pachychoroid spectrum, characterised by the following clinical and OCT features: thick choroid on enhanced depth imaging (EDI-OCT), dilated choroidal vessels in Haller layer, attenuation of the choriocapillaris and inner choroid (Sattler’s layer) and loss of fundus tessellation.

While it is plausible that some patients are predisposed to development of PCV due to underlying pachychoroid, it is unclear if the mechanism responsible for inner choroidal thinning and ischemia is largely mechanical (due to compression from the enlarged choroidal vessels) or primarily driven by hypoxic and inflammatory changes in the choriocapillaris.

 

Pathogenesis

The pathogenesis of PCV is not completely understood. It is widely accepted to be originated at the inner choroidal level.

Polyps may develop from a choroidal vascular network or from a plaque of occult new vessels^(11,14).

Yuzawa et al.⁽¹³⁾ described the filling of PCV lesion simultaneously with the surrounding choroidal arteries suggesting that PCV lesions grow from inner choroidal vessels.

Few clinicopathological studies have been reported.

MacCumber et al.⁽¹⁴⁾ examined an enucleated high myopic eye with rubeosis and vitreous haemorrhage from a diabetic patient without diabetic retinopathy, with high blood pressure and history of multiple, bilateral, recurrent neurosensory and pigment epithelium detachments (PED).

Bruch’s membrane was crossed by choroidal vessels and an extensive fibrovascular proliferation was disclosed within Bruch’s membrane and the inner retinal space. They did not observe choroidal saccular dilatations except that some choroidal veins were quite large.

Inflammation was expressed by the presence of B and T lymphocytes at the level of choroid and fibrovascular tissue and the expression of intercellular adhesion cytokines like ICAM-1 was shown.

Lafaut et al.⁽¹⁵⁾ reported the histopathologic features of surgically removed submacular tissue from an elderly patient with a pattern of PCV on ICG.

A thick fibrovascular membrane located on the choroidal side of the RPE was described. The RPE layer was discontinuous whereas on its choroidal side an almost intact layer of diffuse drusen was observed.

A group of dilated thin-walled vessels were found and located directly under diffuse drusen within a sub-RPE, intra-Bruch’s fibrovascular membrane.

Dilatations appeared to be of venular rather than arteriolar origin and some lesions were associated with lymphocytic infiltration.

The presence of choroidal infiltration by inflammatory cells was also referred by Rosa et al.⁽¹⁶⁾

Okubo et al.⁽¹⁷⁾ described unusually dilated venules adjacent to an arteriole with marked sclerotic changes and newly formed capillaries within the wall of the degenerate arteriole and near the dilated venule.

Therasaki et al.⁽¹⁸⁾ described clusters of dilated thin-walled vessels surrounded by macrophages and fibrinous material in neovascular membranes obtained from submacular surgery for PCV.

Hyalinization of choroidal vessels and massive exudation of fibrin and blood plasma were observed in all the five specimens of PCV lesions studied by Nakashizuka et al.⁽¹⁹⁾.

They also found some blood vessels located above the RPE in two of the five eyes. Immunohistochemically, CD68-positive cells were described by them around the hyalinized vessels.

There were no alpha-SMA-positive cells in the vessels of PCV. CD34 staining showed endothelial discontinuity.

Vascular endothelial cells within the PCV specimens were negative for VEGF. HIF-1alpha positive inflammatory cells were located in the stroma of specimens⁽¹⁹⁾.

Hyalinization of choroidal vessels, like arteriosclerosis, seems to be characteristic of PCV⁽¹⁹⁾.

All these previous histopathological studies identifying a large spectrum features (like dilated choroidal vessels, intra-Bruch’s neovascularization, inflammatory cells, drusen material, thick membranes, single saccular dilatations or clusters of dilated thin walled vessels) may partially being expressing the influence of disease stage⁽²⁰⁾.

For many authors^{(11,15,21-25)} and following the results of clinicopathological studies, PCV may represent a subtype of exudative AMD.

Yannuzzi et al.^(24,25) found a prevalence of 7.8% of PCV in a population with signs of exudative AMD and Laffaut et al.⁽¹⁵⁾ described the presence of late ICG hyperfluorescent plaques in 58% of 45 cases with PCV, proposing that PCV should be considered a subtype of exudative AMD.

Many other authors^(23,24,25) describe PCV cases with subretinal neovascularization.

Ahuja et al.⁽²⁶⁾ described a prevalence of PCV in 85% of a consecutive series with 16 eyes diagnosed as exudative AMD and showing a PED greater than 2 mm of diameter, haemorrhage or retinal neurosensory exudation.

According to Yannuzzi^(24,25)PCV and exudative AMD may differ in mean age of onset (PCV affected patients are younger), presence of exudative peripapillary lesions (more frequent in PCV), prevalence of soft drusen (greater in AMD patients) and ethnicity (PCV more prevalent in non-white population).

PCV also has less tendency to develop fibrous proliferation and a higher incidence of PED and neurosensory detachments.

PCV and AMD share common genetic factors, which suggests that PCV and wet AMD are similar in some pathophysiologic aspects.

A common genetic background may exist between typical exudative AMD and PCV patients. Complement pathway plays a substantial role in the pathogenesis of PCV, like in AMD.

A consistent association of the ARMS2/HTRA1 locus with both neovascular AMD and PCV suggests that the two diseases at least in part share molecular mechanisms⁽²⁷⁾.

The nonsynonymous variant I62V in the complement factor H gene is strongly associated with PCV and may be a plausible candidate for a causal polymorphism leading to the development of PCV, given its potential for functional consequences on the CFH protein⁽²⁸⁾.

The LOC387715/HTRA1 variants are associated with PCV and wet AMD in the Japanese population. The associations are stronger in AMD than in PCV⁽²⁹⁾.

LOC387715 rs10490924 correlates with vitreous hemorrhage and also with the lesion size. Rs10757278 on chromosome 9p21 was shown to be significantly associated with the risk of PCV in Chinese Han patients⁽³⁰⁾.

Among the patients with AMD and PCV, those with a homozygous HTRA1 rs11200638 risk allele seem to have larger CNV lesions⁽³¹⁾.

Another susceptibility gene in PCV, the elastin gene (ELN), has shown to be more associated with PCV than with AMD⁽³²⁾.

With meta-analyses, variants in four genes were found to be significantly associated with PCV: LOC387715 rs10490924 (n=9, allelic odds ratio [OR]=2.27, p<0.00001), HTRA1 rs11200638 (n=4, OR=2.72, p<0.00001), CFH rs1061170 (n=4, OR=1.72, p<0.00001), CFH rs800292 (n=5, OR=2.10, p<0.00001), and C2 rs547154 (n=3, OR=0.56, p=0.01). LOC387715 rs10490924 was the only variant showing a significant difference between PCV and wet AMD (n=5, OR=0.66, p<0.00001)³³.

PCV has been referred to be associated with other ocular disorders like macroaneurysms or inflammatory diseases^(11,34).

However, this association is still inconclusive and deserves further investigation.

A relation between the retinal vascular changes in hypertensive retinopathy, like vascular remodelling, aneurismal dilatations and focal vascular constrictions, and choroidal alterations in PCV was proposed by Ross et al.⁽³⁴⁾.

 

Epidemiology

PCV is usually diagnosed in patients between 50 and 65 years old but the age of diagnoses may range between 20 and more than 80 years.

Most patients with PCV are likely to have AMD signs.

Prevalence of PCV in patients with AMD signs ranged between 4.8% and 23% in different series and different countries^{(3,11-14,21,35)}.

It is considered to be more prevalent in Asian population⁽³⁵⁾ and African American than in Caucasian^(3,11,12,21) as it seems to preferentially affect pigmented individuals.

Preference for female gender is referred in Caucasian^(11,12,21) whilst in Asian population male are more affected^(13,35).

Bilateral involvement is common and may be as high as 86%⁽²¹⁾.

 

Natural evolution

The disease has a remitting-relapsing course and is associated with chronic, multiple recurrent serosanguineous detachment of the neurosensory retina and RPE, with long-term preservation of vision^(11,24,25).

Visual acuity (VA) loss is associated with central macular involvement and may range from mild to severe VA loss or blindness.

Treatment of central macular lesions with photodynamic therapy (PDT) and more recently with antiangiogenic drugs has precluded a better knowledge of natural history of PCV with macular involvement.

Approximately half the patients with PCV lesions in the posterior pole may have a favourable course without treatment⁽³⁶⁾.

In the remaining half the disorder may persist for a long time with occasional repeated bleeding and leakage, resulting in severe macular damage and VA loss.

Eyes with a cluster of grape-like polypoidal dilations of the vessels may have a higher risk for severe visual loss^(37,38).

Choroidal vascular lesions may be located in the peripapillary area, in central macula or in midperiphery.

The analysis of VA outcome must consider location of polyps and/or abnormal vascular network. Most of the PCV natural history series describe lesions in the posterior pole, differentiating macular from extramacular and/or peripapillary polyps and macular involvement ranged from 25% to 94% ^{(12,35,38,39)}.

Kwok et al.⁽³⁸⁾ followed the natural history of nine eyes with macular involvement after a follow-up ranging from 5 to 60 months and found VA improvement of two lines in only one eye (11.1%), VA change of one line in one eye, and VA decrease of two lines in seven eyes (77.7%).

Uhyama et al.⁽³⁶⁾ followed 14 eyes with PCV (13 with macular involvement) for a mean period of 39.9 months and described VA improvement of two lines in five eyes (35.7%) and VA decrease of two lines in four eyes (28.5%).

A favorable course was demonstrated in 50% of the cases with the remaining half of the cases showing recurrent leakage and hemorrhages and progressive VA loss.

Lesions may grow by enlargement with proliferation and hypertrophy of the vascular component but, apparently, not by confluence.

Polyps may bleed, grow, regress or leak and a choroidal neovascularization may appear

A massive spontaneous choroidal hemorrhage is rare but may constitute a severe complication associated with blindness⁽³⁷⁾.

Progression to RPE atrophy is common and may be related with resolution of PED, chronic or recurrent leakage with PED or neurosensory detachment, autoinfarction, regression or flattening of the lesion.

Chronic atrophy and foveal cystic degeneration is associated with severe VA loss^{(2,11,15,23,24,25,39)}.

 

Diagnosis

The current gold standard of diagnosis of PCV is based on ICG imaging (Figure 1) and may be complemented with OCT, fluorescein angiography and fundus findings (Figures 2, 3 and 4).

Clinical examination may show one or more reddish-orange, spheroid, subretinal mass located at the macular or juxtapapillary area (Figure 2).

This mass may correspond to the anteriorly projection of multiple polyps and is very suggestive of PCV.

Also very suggestive of PCV is the serous or serous-hemorrhagic PED and/or neurosensory detachment associated with extensive subretinal haemorrhages and circinate hard exudates (Figures 1,2 and 4).

Fluorescein angiography alone is not useful for PCV diagnosis.

Neurosensory detachment and serous or sero-hemorrhagic PED may suggest the diagnosis but polypoidal lesions are only visualized if the overlying pigment epithelium is atrophic.

Intermediate or late leakage on fluorescein angiography is very often diagnosed as occult choroidal neovascularization with late leakage from undetermined source or may be confused with chronic central serous chorioretinopathy.

The characteristic PCV lesion in ICG is sub-RPE network of vessels ending, in the great majority of cases, in aneurysmatic dilatations, which elevate the overlying RPE, giving rise to the characteristic sharp inverted V-shaped or thumb-like protrusions seen on the OCT.

Other important signs on OCT suggestive of underlying PCV are the irregular undulation of the RPE separated from the Bruch’s membrane (often referred to as the ‘double layer sign’) and adjacent PED.

The lesion may be juxtapapillary, macular or may be rarely located in the midperiphery.

Juxtapapillary lesions often show, in early ICG images, a radial arching pattern, and the vascular channels may be interconnected by smaller spanning branches more numerous at the edge of the lesions⁽³⁹⁾.

Interestingly, juxtapapillary lesions have been reported to be more common in patients of Caucasian and Afro-Caribbean descent compared to macular distribution of polyps being more commonly seen in East Asian patients.

When the polypoidal lesion is located in the macular area, the vascular network often arise in the macula and follows an oval distribution⁽³⁹⁾.

The area surrounding the vascular network is hypofluorescent during early phases of ICG and in late phase ICG often shows a reversal of the pattern: the area surrounding the polypoidal lesion becomes hyperfluorescent and the centre shows hypofluorescence.

In very late phases ICG shows disappearance of the fluorescence (washout) in non-leaking lesions (Figure 4-d)^{(11,22,24,25)}.

OCT and particularly 3-D OCT is very useful for diagnosis confirmation.

Documentation of polyps (number, location, size) and associated features (neurosensory detachment, serous or haemorrhagic PED, haemorrhage) may be assessed by OCT. Particular findings like the undulation of RPE (corresponding to the branch vascular network), double layer sign (the inner layer is RPE and the outer layer is the inner boundary of the Bruch/choroid complex), Bruch membrane thinning, the identification of polyps under a detached RPE and thicker choroid compared with typical neovascular AMD can be assessed as well.

Polypoidal lesion and branching vascular network identified in ICG may be visualized on spectral domain OCT in near 95% of the cases as areas of moderate reflectivity between the clearly delineated abnormal section of RPE and Bruch’s membrane⁽⁴⁰⁾.

Polypoidal lesions are visualized as anterior protrusions of a highly reflective RPE line.

With “en face OCT”⁽⁴¹⁾branching vascular networks were detected as elevations of the RPE.

Serous PED may be seen as round protrusions of the RPE and are often accompanied by adjacent smaller round protrusions of the RPE, consistent with polypoidal lesions.

These protrusions of the RPE are often fused and typically appeared as a ‘snowman’.

Subsequent longitudinal examination reveals the polypoidal lesions to be sharp protrusions of the RPE with moderate inner reflectivity.

Consistent with the location of the branching vascular network, a highly reflective line may be seen often just beneath the slightly elevated reflective line of RPE^(40,41).

Optical coherence tomography angiography is able to detect the BVN in all cases and is located in the space between the RPE and Bruch's membrane. Polypoidal lesions can be detected in up to 85% of the cases, showing high flow signal. They are located just below the top of the pigment epithelial detachment⁴²

Figure 1 - PCV. Red-free (A) image with circinate lipidic exudation. Intermediate phase ICG shows an abnormal choroidal vascular network and multiple polyps in the centre of the circinate exudation (B).

 

Figure 2 - Fundus colour photography with lipidic exudation and a reddish-orange, spheroid, subretinal mass located at the macular area associated with macular edema. Late ICG (top, right) reveals the presence of two polypoidal lesions in the papilomacular bundle. On OCT (bottom) the polyp is well delineated and a sub-foveal neurosensory detachment is observed.

 

Figure 3 - Same eye of Figure 2 after one PDT session. The two polyps show fluorescein staining (A; fluorescein angiography late phase). A complete polyps resolution is observed on late ICG (B). OCT (C): An intermediate reflectivity is now registered inside the polyp limits (no fluid) associated with a complete resolution of the neurosensory detachment.

 

Figure 4 - Fundus colour photography (A) reveals the presence of circinate lipidic exudation surrounding a reddish-orange lesion temporal to the fovea. Late phase fluorescein angiography (B) shows a diffuse leakage with juxtafoveal involvement and two focal areas of staining and leakage. Two polypoidal foci separated by an abnormal choroidal vascular network are observed on ICG intermediate phase (C). Late phase ICG (D) shows an hiperfluorescent plaque, a more temporal hot spot (active polyp) and a more nasal hypofluorescent spot with hyperflluorescent borders (apparently inactive polyp).

 

Differential diagnosis

Differential diagnosis of PCV with central serous chorioretinopathy, AMD with choroidal neovasculazrization, inflammatory conditions and tumors is not always easy and needs ICG for a clear differentiation.

Retinal vascular lesions such as retinal artery macroaneurysms and retinal microaneurysms may be mistaken for polypoidal lesions based on ICG findings, but can be easily differentiated with good stereopairs made available during the ICGA.

Polypoidal choroidal vasculopathy is a primary cause of macular serous retinal detachment without hemorrhage in patients over 50 years of age in Asian population⁽⁴³⁾.

Since clinical and fluorescein angiographic findings are often indistinguishable among central serous chorioretinopathy, PCV, and occult choroidal neovascularization, ICG might help to establish a more definitive diagnosis^(22,44).

Central serous chorioretinopathy shows staining or late leakage but not an abnormal choroidal vascular network neither polyps.

Both these conditions are part of the pachychoroid spectrum of diseases characterised by loss of fundus tessellation, increased choroidal thickness, dilated large choroidal vessels in Hallers layer of the choroid, and attenuation of the choriocapillaris and Sattler’s layer.

The differential diagnosis becomes more challenging when lipid exudation and small PED are associated. ICG may be helpful differentiating PED from polypoidal lesions.

Small PED from central serous chorioretinopathy become hypofluorescent in late phases ICG and hyperfluorescent in late phases fluorescein angiography.

In contrast, polypoidal lesions are usually hyperfluorescent in late phases ICG because of its vascular nature⁽²²⁾.

PCV represents a subtype of type 1 CNV in AMD^{(11,15,23,24,25)}.However, some features distinguish PCV from other subtypes of CNV: eyes with PCV are characterized by a higher incidence of neurosensory detachments, greater neurosensory detachment height, and less intraretinal edema than eyes with occult or predominantly classic CNV⁽⁴⁵⁾.

Non polypoidal lesions in exudative AMD patients tend to produce small calibre vessels that are associated with grayish membranes not easily observed clinically, in contrast with the reddish-orange lesions clinically observed in PCV and corresponding to vascular saccular polypoid lesions^(24,45,47).

Stromal choroidal fibrosis is common in predominantly classic and occult lesions but is quite rare in PCV.

PED associated with CNV in AMD has a poor prognosis whilst PED in PCV lesions virtually never forms fibrotic scars^(22,24,39).

The natural evolution of CNV in AMD eyes to fibrosis and disciform scar is not observed in PCV eyes.

Tumoral lesions like choroidal circumscribed hemangioma, renal cell carcinoma or metastasis from carcinoid syndrome may also be confused with PCV⁽³⁹⁾. Again ICG is essential for differentiation.

Choroidal hemangiomas show, in general, a rapid filling of dye in very early phases and a washout in late phases.

ICG characteristic lesions of PCV are not observed in choroidal or metastatic tumors and ultrasound is also effectiveFigure 4. for characterization of the tumoral mass.

Inflammatory lesions, like posterior scleritis, multifocal choroiditis, panuveitis, acute posterior multifocal placoid pigment epitheliopathy, Harada disease, sympathetic uveitis and birdshot chorioretinopathy may also be confused with PCV.

PCV is not associated with anterior or posterior uveitis neither with leakage nor staining of the optic disc in fluorescein angiography^(4,22,24,39).

Lipid deposition, often observed in PCV is not commonly seen in inflammatory conditions.

Scleral or choroidal thickening and liquid in the subtenon space have never been described in PCV eyes^(24,39).

 

Treatment

Treatment of PCV lesions is only recommended when central vision is being threatened by persistent and progressive exudative changes.

Otherwise, a conservative approach is recommended.

Different treatment modalities like laser photocoagulation, transpupillary thermotherapy, PDT and surgery or intravitreal antiangiogenic drugs have been reported.

To date, only one randomized, controlled clinical trial (RCT) has been published to prove the efficacy and safety of ranibizumab vs PDT or PDT plus ranibizumab (the EVEREST trial).

There are currently two ongoing major RCT which will provide more robust evidence supporting the most optimal therapy for the condition.

Direct laser photocoagulation of leaking polyps has proven short-term safety and efficacy for extrafoveal lesions^(46-49).

Other authors, however, described poor results⁽⁵⁰⁾ and persistence or even increasing exudation in up to 44% of the cases⁽¹⁵⁾, or VA decrease of 2 or more lines in near half of the eyes and legal blindness in up to two third of the eyes⁽³⁵⁾.

Yuzawa et al.⁽⁴⁹⁾ reported good efficacy of laser photocoagulation in near 90% of the eyes if all the polyps and abnormal vascular network were treated.

If the treatment involved only the polyps, more than half of the eyes suffered VA decrease related with exudation, recurrences or foveal scars. Direct laser photocoagulation of feeder vessels, identified in ICG, was also reported as showing VA improvement of 2 or more lines in 8 out of 15 eyes⁽⁵¹⁾.

Considering the possibility of using other treatment modalities, laser photocoagulation should be reserved for well defined extrafoveal active polyps.

Transpupillary thermotherapy has shown to be useful in PCV⁽⁵²⁾.

A large number of reports on PDT in subfoveal PCV has been published(^{(23,26,53-57)}.

Results at one year and even at two years are apparently superior to those of PDT in predominantly classic or occult CNV.

A longer follow-up shows a trend to a progressive but not significant VA decline at two⁽⁵⁷⁾ and 3 years follow-up⁽⁵⁸⁾: 

The rate of eyes gaining a significant amount of vision dropped from 26% in the first year to 15% in the third year, and the rate of eyes with significant VA loss (3 or more ETDRS lines) increased from 17% to 26%.

A high rate of recurrence (44%) occurred during the 3-years follow-up but it was not associated with significant VA decline⁽⁵⁸⁾.

This high rate of recurrences may be associated with a poor response of the vascular network to PDT⁽⁵⁹⁾.

VA decline at 3 years may partially be explained by photoreceptors death due to chronic or recurrent neurosensory detachment, massive hemorrhages and progressive RPE atrophy.

The number of treatments decreases markedly after the first year from an average of 2.0 to 0.4 and 0.5 in the second and third years respectively⁽⁵⁸⁾.

Complications like subretinal haemorrhages, haemorrhagic PED or even massive hemorrhages have been associated to PDT in PCV eyes⁽⁶⁰⁾.

However, all of these complications may also occur without any treatment.

Surgical removal of polypoidal lesions and associated hemorrhages has been reported^(15-19,61) 

with and without macular translocation.

Considering the potential alternative treatments and the high rate of serious complications, macular translocation is no longer being considered for PCV⁽⁶²⁾.

The relatively large vascular lesions of PCV patients needs to be considered if a vitrectomy is planned in cases without associated massive hemorrhage.

Intravitreal antiangiogenic drugs, like bevacizumab⁽⁶³⁾ and ranibizumab⁽⁶⁴⁾ have been used for treating PCV eyes.

Ranibizumab short-term results⁽⁶⁴⁾ seem to be promising in terms of maintenance or improvement of VA and resolution of subretinal fluid and PED.

Polyps were reported to disappear in 69% of the cases at 3 months when using ranibizumab⁽⁶⁴⁾ but not with bevacizumab⁽⁶³⁾.

Intravitreal bevacizumab appeared also to result in stabilisation of vision and reduction of exudative retinal detachment in PCV patients in short-term evaluation.

However, it had limited effectiveness in causing regression of the polypoidal lesions⁽⁶³⁾.

Some exudative AMD eyes refractory to ranibizumab or bevacizumab may be, in fact, PCV cases.

Combined treatments associating PDT and intravitreal ranibizumab or bevacizumab have been reported to show good efficacy in these refractory cases of AMD⁽⁶⁵⁾.

The EVEREST study is the first multicenter, double-masked,ICG-guided randomized controlled trial with an angiographic treatment outcome designed to assess the effect of Visudyne® (PDT) alone or in combination with Lucentis® (ranibizumab) compared with Lucentis® alone in patients with symptomatic macular PCV.

A total of 61 PCV patients of Asian ethnicity from 5 countries (Hong Kong, Taiwan, Korea, Thailand, and Singapore) participated in the study.

The six months EVEREST study results⁽⁶⁶⁾ suggests that in a majority of patients, Visudyne® therapy, with or without Lucentis®, may lead to complete regression of the polyps that can cause vision loss in patients with PCV.

A complete polyp regression (primary endpoint) was achieved in 77.8% of patients who received the Visudyne® – Lucentis® combination, in 71.4% of Visudyne® monotherapy patients and in 28.6% of patients in the Lucentis® monotherapy group (p=0.0018 for combination, p=0.0037 for Visudyne® vs Lucentis®).

Best corrected VA from baseline to month six improved in average in all groups with patients in the combination group achieving the highest gain (+10.9 letters from baseline). Lucentis® monotherapy patients gained +9.2 letters, and Visudyne® monotherapy patients +7.5 letters. Differences between the groups are not statistically significant.

All therapies were well tolerated and the safety findings were consistent with the established safety profiles of Visudyne® and Lucentis®.

Three multicentre randomized clincial trials are beeing runned with Ranibizumab (EVEREST II) and Aflibercept (PLANET AND ATLANTIC) comparing the efficacy of anti-VEGF alone or in combination with PDT.

The results will help to clarify if a combined treratment is necessary for polyps closure, resolution of exudation and visual acuity imporovement. 

The EVEREST II is a phase 4, 2-year randomized clinical trial comparing ranibizumab alone with a combined teraphy of ranibizumab plus verteporfin PDT in 321 Asian patients.^(71,41)

The PLANET randomized clinical trial enrolled 331 patients in Asia and in two European countries. 

It is a phase 3-4, 1-year study comparing aflibercept alone with aflibercept plus verteporfin PDT in patients with PCV.⁽⁷²⁾

The ATLANTIC study is a phase 4, 1-year randomized clinical trial, being runned in Portugal and Spain, comparing intravitreal treat and extend aflibercept monotherapy with aflibercept treat and extend regimen with adjunctive PDT in patients with PCV. ⁽⁷³⁾

 

Combined therapies

The adjunctive use of intravitreal triamcinolone acetonide with PDT did not appear to result in additional benefit in the PDT treatment of PCV.

A retrospective analysis of PCV patients who underwent PDT with or without intravitreal triamcinolone acetonide, with a follow-up of 2 or more years, was reported in 27 eyes and showed to have a non-sustained effect after 1 year.⁽⁶⁷⁾

The combination therapy between PDT and an anti-VEGF showed encouraging results including improved vision, reduced incidence of subretinal hemorrhage and reduced recurrence of polyps, when compared to PDT monotherapy for PCV. Reduced-fluence PDT combined with intravitreous bevacizumab for PCV improved vision and reduced complications.⁽⁶⁸⁾ even at two years.⁽⁶⁹⁾

Combined treatment with PDT plus anti-VEGF should be considered in eyes diagnosed with leakage from the branching vascular network and the polyps, in eyes with visible exudation associated with PED and  when ICGA results do not permit clearly to distinguish between PCV and CNV and when both conditions co-exist.⁽⁷⁰⁾

The results of the three ongoing multicenter and randomized studies will certainly help to clarify the need to use or not combined therapies in the treatment of PCV.

 

Conclusion

PCV may be considered a subtype of exudative AMD.

ICG is always mandatory for the diagnosis of PCV, showing a unique lesion – an abnormal inner choroidal vascular network with polypoidal structures at the borders.

OCT and fundus findings may complement the diagnosis.

PCV needs to be differentiated from other forms of exudative AMD, central serous chorioretinopathy, inflammatory conditions and some choroidal tumors.Photodynamic therapy with Visudyne®, alone or in combination with antiangiogenic drugs, seems to be necessary for a complete resolution of the polypoidal lesions.

 

>> References

Serous PED in AMD

 

Retinal pigment epithelial detachment (PED) is part of age-related macular degeneration (AMD) clinical spectrum.

However, different types of PED have been reported in the literature and they have been either related or not with AMD.

Serous PED is defined as an area of sharply demarcated, dome-shaped serous elevation of the retinal pigment epithelium (RPE). The histopathology of serous PED is consistent with the detachment of the RPE basement membrane, along with the overlying RPE from the remaining Bruch’s membrane due to accumulation of fluid⁽¹⁾.

The presence of this lesion is a negative prognostic factor for AMD in terms of visual acuity outcome.

While no definite therapeutic indications have been set so far, early detection of serous PED is important for the prognosis and the management of patients with AMD.

In AMD, serous PED can be either associated or not with choroidal new vessels (choroidal neovascularization – CNV). However, the vascularised type is by far the most observed.

Several theories regarding the relationship between serous PED and CNV have been proposed.

To explain its pathogenesis, Gass theorized the growth of new vessels from the choroid (Type 1 neovascularization (NV)) inside the Bruch’s membrane thickness, that actively leak, increasing the hydrostatic pressure and causing RPE detachment among the less adherent layers⁽²⁾.

This concept has been later supported by the evidence that the development of CNV comes with inflammatory mechanisms that add more damage to Bruch’s membrane, supporting RPE separation from the inner collagenous layer^(3-5).

When the growth of new vessels starts from the inner retina, more recently described as Type 3 NV and also known as retinal angiomatous proliferation (RAP), it has been hypothesized that the serous PED formation, which is very frequently associated, can be related to RPE invasion by the neovascular complex^6-8).

By contrast, other authors observed that the presence of PED can represent a pre-existing condition that can promote CNV growth through a further Bruch’s membrane damage, expression of the same ongoing disease^(9,10).

Although the pathogenesis of the PED is not completely understood, from these studies the formation of NV seems to be a pivotal moment.

At fundus examination, serous PED appears as a round or oval, distinct dome-shaped area of regular detachment of the RPE and the overlying neurosensory retina, with yellow to orange color and smooth surface. Margins are typically sharply demarcated; and focal RPE atrophy and pigment figures are frequently observed^(9,11).

However, the concomitant presence of NV can generate a variety of associated ophthalmoscopic aspects, such as hemorrhagic and exudative components, areas of irregular elevation of the RPE and serous detachment of the surrounding neuroretina.

Presentation of Type 1 NV located at the margin of the PED may vary,  usually resulting in a reniform or notched aspect, or a flat-sided RPE detachment⁽¹²⁾.

Serous PED may be imaged by fluorescein angiography (FA), indocyanine green angiography (ICGA) and optical coherence tomography (OCT).

FA represents, however, the gold standard for diagnosis of serous PED.

Examined by FA, serous PED classically shows an early uniform hyperfluorescence of the entire lesion, slightly delayed compared to the background fluorescence, that progressively increases in brightness as the examination progresses (pooling).

Serous PED hyperfluorescence typically does not change in size or shape during the angiographic phases.

FA can also demonstrate the presence of NV, usually associated to serous PED as Type 1 NV, like areas of indistinct late subretinal staining, more evident when located at the margin of the RPE detachment or corresponding to the “notch”(11).

The presence of NV can be also deducted by the presence of an hemorrhagic component of the PED, the dark meniscus described by Gass⁽¹²⁾.

However, a more precise localization of the neovascular component can be obtained with digital ICGA. Indocyanine green molecule has biophysical properties that, unlike fluorescein, make it useful to enhance vessels anatomy through RPE, blood and turbid exudation.

In detail, ICGA enables to better delineate the presence and the type of new vessels associated with a serous PED, and for this reason is considered a fundamental tool in the management of this disease^(13-15).

On ICGA, serous PED appears as an hypofluorescent lesion, with sharply delineated margins, that remains constantly hypofluorescent during all the phases of the examination(16).

When the new vessels are not present, no signs of localized hyperfluorescent areas are detectable; the outline of the PED is sharply round and it is therefore considered a pure serous PED.

In AMD patients, Yannuzzi found an incidence of 4% of non-vascularized PED among serous PED⁽¹⁵⁾.

When the neovascular component is present, it has been suggested the term vascularized PED⁽¹⁵⁾,.which accounts to approximately 24% of newly diagnosed exudative AMD⁽¹⁷⁾.

New vessels associated with serous PED are represented in different subtypes.

High-speed videoangiography with scanning laser ophthalmoscope appears as a precious tool that allows the ophthalmologist to identify the new vessel pattern and their angiographic behaviour⁽¹⁸⁾.

Recognize the different types of NV, by distinguishing angiographic findings, is mandatory for the distinct natural course, visual prognosis and different response to the treatment of the three main kinds of new vessels associated to serous PED in AMD.

The most common type of new vessels associated with serous PED are those occurring from the choroid beneath the RPE monolayer^(15-,17).

These new vessels have been recently classified as Type 1 NV and are by far the most common type of NV in AMD⁽¹⁹⁾(Figure 1).

Figure 1 - Vascularized PED with Type 1 NV. (A) FA, (B)ICGA and (C) OCT.

 

In the early phases, ICGA shows the NV feeder artery that arises from the choroidal circulation, and subsequently the draining venule.

At the same time the capillary network of the neovascular membrane can be detected.

Unlike fluorescein, indocyanine green leaks slightly and the NV hyperfluorescence is usually minimal, with the exception of some cases that show an intense leakage, considered as very active new vessels.

Frequently, in the late phases, a well-defined area of mild hyperfluorescence corresponding to the NV network can be appreciable.

The second type of new vessels complicating serous PED are the RAP(7,20-22), also referred as Type 3 NV⁽¹⁹⁾.

These vascular lesions, as reported by various authors, may involve the outer retina and the RPE, through a progression that has been hypothesized to originate from the retinal circulation and/or choroid.

ICGA typically shows the presence of a “hot-spot”, due to the early hyperfluorescence of the intraretinal neovascular complex, that increases during the angiography, with an intense leakage in the late phases.

Its brightness is enhanced by the surrounding hypofluorescence of the underlying PED (Figure 2).

Figure 2 - Vascularized PED with Type 3 NV (RAP). FA (A) and  ICGA early (B) and late phases (C and D).

 

In the late stages of the disease, the choroidal neovascular complex is typically connected with one or more retinal vessels which appear tortuous and dilated^(7,22,23).

The Type 3 NV may be single or multiple, its origin is typically extrafoveal, and an intraretinal hemorrhage in correspondence of the neovascular lesion is frequently observed⁽²⁰⁾.

The third type of new vessels associated with serous PED in AMD is consistent with polypoidal choroidal vasculopathy (PCV)⁽²⁴⁾.

PCV is a peculiar form of CNV, characterized by the presence of orange, aneurismal, polyp-like round dilatations at the border of a branching vascular network of choroidal origin.

Although PCV affects more frequently middle-age black and asian populations, its clinical spectrum is expanded to whites, where it has been found to be present in 8-13% of patients with concomitant AMD lesions.

In these cases, when the manifestations attributable to both PCV and AMD are present, some authors consider PCV as a subtype of CNV in AMD^(24,25).

Hemorrhagic manifestation is common in patients with PCV.

Serous PED associated with PCV shows frequently a blood level in the lower portion of the detachment.
ICGA is the state-of-the-art examination to distinguish the typical features of the two vascular components.

The vascular network is characterized by the presence of one or more aneurismal lesions that show a bright fluorescence since the early phases, followed in the late phases by a clearing of the dye, called “wash-out”, typical of this disease (Figure 3).

Figure 3 - Vascularized PED with PCV: FA (A) and early (B), mid (C) and late (D) phases of ICGA.

 

Nevertheless, some polyp-like structures can actively leak showing late staining of their walls and surrounding exudation.

The polypoidal lesions are usually located at the margin of the serous PED⁽²⁶⁾.

Recognition of these lesions is critical because of their different clinical course, prognosis and treatment response as compared to the other neovascular AMD subtypes.

OCT provides images that allow an exact correlation with the angiographic findings.

In cross-sectional OCT scans, serous PED appears as an optically empty dome-shaped elevation of the external high reflective band – the RPE, that steeply detaches from Bruch’s membrane⁽²⁶⁾.

The overlying retina, usually adherent to the bullous PED, at the margins of the lesion, can be slightly detached from the underlying RPE.

More additional information can be provided by OCT in vascularized PED⁽²⁸⁾.

The tomographic sections, guided by FA and ICGA in the area corresponding to the CNV, show a smoother elevation of the RPE, continuous with the serous detachment, with a deeper backscattering, due to the presence of the fibrovascular tissue.

Hyporeflective areas of homogeneous optically empty spaces referable to fluid accumulation are frequently present in the intraretinal and subretinal spaces⁽²⁹⁾.

Intraretinal optically empty spaces are more pronounced when the serous PED is associated with a Type 3 NV, especially with cystic shape (Figure 4).

By positioning the scan line corresponding to the “hot-spot”, the neovascular abnormality is represented as a dense or hyperreflective pre-epithelial zone in the inner retinal layers, where the outer hyperreflectant layers are no longer detectable⁽³⁰⁾.

The RPE close to that lesion shows frequently effractions or interruptions in its hyperreflective layer⁽³¹⁾.

The retinal topographic measurement sustains an increased retinal thickness.

In eyes with serous PED and PCV, the polypoidal lesions show a sharp protrusion of RPE, similar to the PED but steeply sloped.

The polyps cavity, usually optically empty, is contiguous to irregular RPE elevation, expression of the occult neovascular component of the lesion^(32,33).

Subretinal and intraretinal fluid, observed as hypofluorescent optical empty areas, are related to the PCV activity (Figure 4-B).

Figure 4 - Type 3 NV (RAP) (left) and PCV (right): OCT and ICGA patterns.

 

The recent introduction of OCT angiography (OCT-A) has made it possible to image perfusion of the different retinal layers, without injection of the dye and by utilizing endoluminal flow as intrinsic contrast.

Figure 5 provides an example of how OCT-A image the choroidal network in a case of PCV.

Figure 5 - Serous PED and recurrent PCV (FA, OCT, ICGA and OCT-A). On the top, OCT scan acquired simultaneously with FA; On the bottom, appearance of the polyps profile on OCT-A (left) and ICGA (right).

 

However, the static nature of this examination and the presence of possible artefacts are important limitations that should be acknowledged.

With regard to serous PED, shadowing due to loss of signal transmission in correspondence of the PED may make it difficult to detect the CNV complex on OCT-A (Figure 6).

Figure 6 - Vascularized PED and serous PED with Type 1 NV (FA, ICGA and OCT-A) . On the top, FA (left) shows dye pooling due to presence of a serous PED; ICGA (top middle and right) shows a neovascular network on the edge of the PED (notch); On the bottom, OCT-A does not provide a definite imaging of the neovascular net.

 

Serous PED natural course depends on the presence or not of the neovascular component⁽³⁴⁾.

In pure serous PED there is generally a slow enlargement of the lesion, with a minimal progression of visual loss over a long period (months or years).

However, many can subsequently develop neovascularization, which make it worse⁽³⁵⁾.

Natural course in vascularized PED may vary, and it is related to the type of new vessels associated.

The most common acute complication of PED is the tearing of the RPE^(36-39).

It usually occurs at the edge of the PED, at the intersection of the detached and attached RPE.

Clinically, RPE tear or rip appears as a well-defined area of bare choroid, contiguous to a darker hyperpigmented rugate area, that corresponds to the mound of the RPE that has torn away^(40,41).

The ripped RPE usually rolls towards the CNV, and its propensity to tear can be predicted by the observation of pre-tear characteristics, such as an increase in the size and a modification in the shape, the presence of small holes at the PED margins, the presence of hemorrhages or subretinal fluid, but the most noteworthy aspects are the irregular filling of the PED visible at the FA, height of PED > 580 nm, duration > 4.5 months, hyperreflective radial lines on near reflectance imaging, smaller ratio of vascularized PED and anti-vascular endothelial growth factor (anti-VEGF) therapy^(42-46).

RPE tears occurs either spontaneouslyor after a treatment, formally laser photocoagulation, photodynamic therapy and intravitreal injection of steroids or anti-VEGF agents^{(47- 58)}.

The exact pathogenesis of RPE tears is poorly understood.

Concerning PEDs’ natural course, it has been hypothesized that tangential shearing forces in the PED can cause the break of the RPE basement membrane at the edge of the detachment; however, it is more likely the result of several variables, where the presence of a CNV plays a major role.

Several causal relationships have been reported for RPE tears occurring after treatment, including the heat generated by photocoagulation, the abrupt increase of intra-PED fluid, a contraction of the associated CNV and the concomitant sudden resolution of the sub RPE fluid.

The combining presence of vitreomacular traction and the deformation of the globe due to the mechanical trauma by the needle have also been reported as causative agents⁽⁵⁹⁾.

After RPE rips, the majority of patients complain of a sudden severe visual decrease.

In a small percentage of eyes, where the tear spares the fovea, patients can experience a temporary preservation of good visual function⁽⁶⁰⁾.

However, in the long term, the progression of a subretinal scar leads to a severe visual decrease. In the prognosis of serous PED, it must be also considered the high risk of bilateral involvement⁽⁶¹⁾.

Treatment of serous PED, associated or not with CNV, has always been a challenge and so far there are no recommended guidelines for their management.

Pure serous PEDs have been treated in the past with laser grid or scattered photocoagulation, nevertheless with disappointing results⁽⁶¹⁾.

Treatment of serous PED, associated or not with CNV, has always been challenging and there are no recommended guidelines for their management so far. Pure serous PEDs have been treated in the past with laser grid or scattered photocoagulation, with disappointing results⁽⁶²⁾.

No other approaches have been attempted to treat these lesions.

When a neovascular net is present, treatment of serous PED has been focused on CNV management. However, given that vascularized PEDshave never been included in the major randomized controlled trials, we need to make the treatment decision based on small series published, which are often retrospective and do involve different therapeutic approaches.

Now, in the anti-VEGF therapy era, all the previous employed treatments appear unsatisfactory.

Laser photocoagulation has been widely employed and might still have a limited indication when an ICGA-well defined CNV lies remote to the detached RPE⁽⁶³⁾.

Verteporfin photodynamic therapy (PDT) alone has been proved to be harmful, increasing the risk of RPE tear, hemorrhages and sudden visual acuity decrease^{(39,48,49,64)} (Figure 7).

Figure 7 - Vascularized PED with CNV (Type 1 NV) before (left) and after (right) PDT: RPE tear (FA and OCT).

 

However, PDT combined with intravitreal triamcinolone acetonide injection (IVTA) has been demonstrated as potentially capable of visual acuity stabilization and recurrences reduction⁽⁶⁵⁾.

Nevertheless, the high rate complications (cataract and glaucoma) has reduced the use of intravitreal triamcinolone.

After the encouraging results obtained with the anti-VEGF intravitreal therapy in the treatment of occult CNV, the use of anti-VEGF treatment have been extended to vascularized PED with disappointing results^(66-69).

Both acute complications and poor anatomical response to the treatment frequently invalidate our attempts to heal the lesion. RPE tears and subretinal hemorrhages have been reported to complicate intravitreal ranibizumab and bevacizumab treatments^(51-57).

Moreover, the sub-RPE fluid hardly responds to the anti-VEGF therapy, possibly due to the hydroconductivity changes of Bruch’s membrane⁽⁶⁹⁾.

In a retrospective case series of 328 patients treated with bevacizumab, ranibizumab, pegaptanib and PDT+IVTA respectively, after a mean follow-up of 42.4 weeks, the authors reported a significant stabilization of visual acuity in each group, better in the bevacizumab and ranibizumab ones compared to the other two, and an overall RPE tears frequency of 12.5%.

However, they conclude that with these treatments, only a partial regression of the lesions can be obtained, and the risk of RPE tears is not avoided⁽⁶⁹⁾.

Another retrospective study⁽⁵⁸⁾. reviewed patients outcome of vascularized PED treated with PDT alone, PDT combined with IVTA or intravitreal anti-VEGF injections alone (bevacizumab or ranibizumab) and showed better functional results for the anti-VEGF treatment group.

Moreover, in this series, Type 1 NV with vascularized PED as compared to Type 3 NV, along with a better visual acuity at baseline, showed greater risk of acute RPE tear after treatment⁽⁵⁸⁾.

In a recent prospective study⁽¹⁹⁾, treatment of PED associated with subfoveal Type 1 NV with intravitreal ranibizumab injection and with a 3-monthly loading phase and a pro re nata strategy led to only partial results over a 24-month follow-up.

Recently several studies have investigated the efficacy of intravitreal aflibercept therapy of PED in AMD showing good anatomical response with improvement or no significant change in visual acuity^(70-73).

Moreover, intravitreal aflibercept has been shown to be a promising treatment in PED resistant to intravitreal ranibizumab treatment^(72-74)..

Several studies have identified factors which might influence response of PED to anti-VEGF treatment^(75-78).

Dirani et al.⁽⁷⁵⁾ showed that better visual improvement was associated with lower baseline visual acuity, presence of subretinal fluid and RAP. Moreover, in their series, PED reduction was associated with higher PED at baseline, predominantly serous PED and use of aflibercept.

Cho et al.⁽⁷⁶⁾, in a recent case series, found that lower PED height at baseline, PCV or RAP compared to typical neovascular AMD, serous PED compared to fibrovascular PED, and aflibercept compared to ranibizumab, have an higher chance of PED resolution during anti-VEGF treatment of PEDs.

However, in a recent post hoc analysis of a phase III randomized controlled trial⁽⁷⁷⁾, it has been shown that, at 24 months after initiation of anti-VEGF treatment, around half of the patients presenting PED at baseline showed complete resolution of PED regardless of PED status and height at baseline.

Recently Chen et al.⁽⁷⁸⁾ outlined the importance of differentiating eyes presenting RAP as they havebetter anatomical and functional outcomes with fewer injections as compared to PED with Type 1 NV.

In the future, new combination therapies and new therapeutic strategies, along with identification of new clinical biomarkers of response to treatment, will help to improve the prognosis of the patients affected by vascularized PED.
 

>> References