Fernanda Vaz, MD
Ophthalmology Department, Hospital de Egas Moniz New University of Lisbon, Lisbon, Portugal
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(2,6,7).
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(1).
In Age-related Eye Disease Study (AREDS), GA is included in category 3 or intermediate form
(if not involving the center of the macula), and 4 or advanced form (involving the center of the macula)(4).
One eye with GA and choroidal new vessels (CNV) is considered as having exudative form(5).
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(7,8,10,11).
The most consistent risk factors are age, familiar history and tobacco smoking like in other forms of AMD(2,7,10). 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(11,12).
The prevalence of GA is lower in blacks than in whites (2.1% vs 4.8%) like in exudative forms(14,15).
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 Study(16).
In this study, diabetes and the ratio total/HDL cholesterol were linked to increased risk of GA(18). Genetic risk factors have been described in association with AMD.
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, LOC HTrA1, C2 and C3, have been implicated in the development of both forms of AMD(18-20).
Some studies have linked specifically 5p region and 4q 32 region with GA(21,22).
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(3,23,24).
Less frequently, GA can follow a drusenoid RPE detachment, regression of a CNV membrane or a RPE rupture(5,25,26).
In some eyes, atrophy was related to a micro reticular pigment pattern distributed around the perimeter of the fovea(26).
Most histopathologic studies suggest that RPE cells are the primary target in GA and its death results in choriocapillaris atrophy(29,30).
Autofluorescent pigments as lipofuscin that accumulate in RPE cells, contribute to a decline of the cell function and degeneration conducting to GA(27).
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(31,32).
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(33,34,35,36).
Macrophages are often seen in areas of GA, apparently phagocytosing pigment and debris resulting of normal cells deletion(37).
Fundoscopy in GA typically shows a well-circumscribed oval or round area of pigment epithelium atrophy, usually sparing the fovea until late stages (Fig. 2).
All precursor lesions of this final appearance can also be present: large drusen (>125 microns), focal pigmentation changes and refractile deposits(3,23).
On fluorescein angiography, GA appears as a sharply delineated window defect due to atrophy of overlying layers of RPE (Fig. 3).
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 geographic atrophy(24).
Despite these aspects in GA, fluorescein angiography may be indicated only in atypical cases, in order to allow the correct diagnosis(38).
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 (Fig. 4)(39,40).
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(41).
Fundus autofluorescence (FAF)
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(42,43,44).
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(45).
Increased FAF precedes development of GA(46,47).
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(48,49).
Despite the works of Holz and Valkenberg, the prognostic value of FAF remains controverse.
In a recent study, FAF was not a strong risk factor for development or extension of GA(50).
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(51).
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(51,53,58).
On the other hand there is a high interindividual variability(54).
The mean overall enlargement rate of atrophic area is 2,6 mm2/year and eyes with larger areas of atrophy at baseline tend to have larger enlargement rates(53).
GA often first develops surrounding the fovea, sparing the central area(55,56,57).
Because of that, the correlation between visual acuity (VA) and area of atrophy can be complex.
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.
The risk of acuity loss was higher in eyes with better visual acuity at baseline and with lightest iris color(57).
Usually vision loss is bilateral because of lesion symmetry, and evolution since first signs to legal blindness is quite variable(53,58).
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(52,55,57).
The low luminance visual dysfunction and the maximum reading rate seem to be significant risk factors for subsequent VA loss(55).
In AREDS, dietary supplements as zinc and anti-oxidants vitamins C, 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(4,12).
Macular xantophylls and polyunsaturated fatty acids seem to be associated with a lower risk of advanced age-related macular degeneration(59,60).
Because of that antioxidant effect of macular pigments as lutein, zeaxanthin and omega-3 fatty acids has been tested in the Age-related Eye Disease Study 2 (AREDS 2)(59,61).
Low dietary glycemic index also seems to reduce the risk of evolution to advanced AMD(61).
Other behavioral factors such as stop smoking and control of BIM may play an important role on prevention(13).
So far there is no proven drug treatment for dry AMD.
However, several trials are investigating many strategies regarding three major targets: Neuroprotection, oxidative stress protection and suppression of inflammation(62).
Ciliary neurotrophic factor (CNTF) is a potential new treatment protecting photoreceptors of degeneration.
CNTF is released from an encapsulated cell device implanted in vitreous cavity(63,64).
ACU-4429 is a modulator of visual cycle, inhibiting the generation of lipofuscin precursors and is selective to rod system(62).
OT-551 is a new topical anti-inflammatory, antiangiogenic and antioxidant agent, that protects photoreceptor cells by inhibiting lipid peroxidation(64,65).
POT-4 is a C3 inhibitor administered as an intravitreal gel and suppresses local inflammation(66).
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.