AMD An Overview of Clinical Forms
AMD An Overview of Clinical Forms
AN OVERVIEW OF
CLINICAL FORMS
Oudy Semoun
Alexandra Miere
Mayer Srour
Eric Souied
and the Ophthalmology Créteil team
PHARMA
AMD: AN OVERVIEW OF CLINICAL FORMS • 3
4 • AMD: AN OVERVIEW OF CLINICAL FORMS
Dear readers,
Recent years have seen an increase in our understanding of the importance of the
genetic origins of AMD, along with the impact of environmental and nutritional factors.
Macular imaging has become multimodal, with colour and Multicolor® photographs,
autofluorescence and infrared photographs, fluorescein angiography, indocyanine
green angiography, OCTA, spectral-domain OCT, EDI-OCT, and wide-field and ultra-
wide-field imaging.
The information obtained from these new forms of imaging has required us to revisit
the group of diseases categorised as “age-related macular degeneration”, resulting
in an even greater refinement of the associated clinical description and changes to
the AMD classification system. Pigment epithelial detachment covers a spectrum of
entities with different semiology and prognosis, including serous PED, fibrovascular
PED, wrinkled PED and fibrous PED, among others. Fibrosis and atrophy underlying
neovascularization are no longer occult, but better described and better detected.
Finally, EDI-OCT has enabled us to differentiate between different pachychoroid
diseases, including retinal pigment epitheliopathy, central serous chorioretinopathy,
polypoidal vasculopathy and adult-onset foveomacular vitelliform dystrophy. Atrophic
AMD is starting to appear in therapeutic considerations of the disease, and the clinical
forms of the various entities included under this term are gradually being described.
The final piece of the puzzle is longitudinal patient monitoring, both prospective and
retrospective. The technique of eye tracking has been hugely significant in optimising
patient monitoring, both in terms of the natural history of the disease and after
treatment, primarily via OCT.
Oudy Semoun
Alexandra Miere
Mayer Srour
Eric Souied
2
Multimodal imaging for age-related macular
degeneration ............................................................................................................................................................................................................................................... 21
Part 1: Non-invasive imaging .................................................................................................................................................................. 22
Setha Vo Kim, Francesca Amoroso
Age-related macular
degeneration:
epidemiology,
environmental and
genetic risk factors
Alexandra Mouallem-Bézière, Jean-Louis Bacquet
Age-related macular degeneration (AMD) is a common eye disease among older people, causing a
severe visual impairment. The pathophysiology of AMD remains largely unknown: it is a complex,
multi-factorial disease involving aging of the retina and a combination of environmental and
genetic factors.
1. Epidemiology
The majority of the large epidemiological studies on age-related macular degeneration (AMD) have been
conducted in the United States1-3. The Beaver Dam Eye Study3 estimated the prevalence of advanced
forms of the disease at 1.6% of the general population and 7.1% in people over 75 years old. In Europe,
the data from the EUREYE Study4 are consistent with the American data and indicate a prevalence of
1.2% for geographic atrophy and 2.3% for exudative forms. In France, the disease affects an estimated
1.5 million people4. It is estimated that by 2020, around 196 million people worldwide will be suffering
from a form of AMD5.
Several constitutional factors have been studied in order to identify their role in the onset of the disease.
Age is obviously a factor, with prevalence increasing with age. In addition, the prevalence of the disease
varies depending on the ethnic origin of the studied populations6. Advanced forms of AMD are more
common in Caucasian populations than in dark-skinned populations7. In terms of sex, some meta-
analyses have not found any differences in disease prevalence between men and women7. However, the
Beaver Dam Study3 found a higher incidence of age-related maculopathy in women over 75 compared
to men of the same age.
Multiple studies have established correlations between certain cardiovascular risk factors and AMD. Klein
et al.8 found a correlation between patients with neovascular AMD and the presence of cardiovascular
disease. Data from the AREDS study reveal increased cardiovascular mortality among patients with
AMD9. However, other studies, such as AREDS Report No. 19, have not found any correlation between
AMD and angina pectoris10. High blood pressure has been studied independently. The results are similarly
divergent, with some studies finding an association11 and others finding none12.
2) Environmental factors
Lutein and zeaxanthin are the two beta-carotenes found in the retina, with maximum concentration in the
macula. These pigments absorb at least 40% of blue light. Several studies have found that consuming
lutein and zeaxanthin is beneficial in warding off the disease20, but these results are contradicted by other
studies3,21.
Antioxidants have also been studied. AREDS Report No. 822 established a significant association between
antioxidant (vitamins C and E and beta-carotene) and zinc intake and a reduced risk of exudative AMD,
compared to placebo.
The NAT 2 study attempted to evaluate the efficacy of DHA and EPA supplementation in the prevention
of choroidal neovascularization at three years in patients with age-related maculopathy (ARM)23. This
study found a significant reduction (68%) in the risk of developing an exudative form of the disease in
patients with the highest level of DHA and EPA. Genetic analysis of this cohort also demonstrated that
the protective effect of the supplementation was maximum in patients who did not have the C-allele of
the CFH Y402H polymorphism24.
Genetic susceptibility to early, intermediate and advanced AMD was first hypothesised by Gass in
197325. Gass had observed several families with cases of AMD and hypothesised that the disease was
an autosomal dominant disorder. Other studies of family aggregation have identified a higher frequency
of the disease in the relatives of existing patients26-28.
The existence of a genetic component has also been suggested in light of the high phenotypic concordance
between monozygotic twins29-31. Monozygotic twin studies can also be used to assess the “heritability”
of a disease, i.e. the proportion of the phenotype attributable to the genotype. The heritability of AMD is
estimated to be between 45% and 70%, based on the studies conducted32,33.
Genetic linkage studies conducted in the families of people suffering from the disease have identified
various chromosomal loci of susceptibility. In 1998, Klein et al.34 published the first linkage study, which
mapped a locus of susceptibility to 1q. Several years later, this locus was found to contain one of the
major genes for susceptibility: CFH35.
All of these family-based studies have clear limitations due to certain characteristics of AMD. The
advanced age of patients makes it harder to find families for analysis, and non-Mendelian inheritance
makes the disease difficult to analyse in a limited number of patients. Researchers therefore changed tack
and began comparing populations with AMD to control populations, in order to study the distribution of
genetic susceptibility markers.
These genes can also be identified by sequencing candidate regions highlighted in segregation studies or
through positional cloning. The HTRA1/ARMS2 region has been identified using this approach.
In terms of the severity of the disease, genotype-phenotype correlation studies have found an
association with the single-nucleotide polymorphism rs10490924 in ARMS2/HTRA1 in patients with
advanced bilateral forms of AMD49, as well as severe and early forms50. rs1061170 in CFH51 is another
polymorphism associated with bilateral forms. Additionally, a recent study identified a correlation between
a polymorphism in C3 and large vascularised pigment epithelial detachments52.
These genetic data can be combined to calculate disease prediction scores that aim to take into
account the various factors influencing the disease. For example, the score developed by Seddon et al.63
combines 10 single-nucleotide polymorphisms in different loci (ARMS2/HTRA1, CFB, C3, C2, COL8A1,
RAD51B, C3) with age, sex, education level, BMI, smoking history and clinical examination data (drusen,
contralateral involvement) to assess the risk of AMD after 10 years. The score can be calculated at:
https: //www.seddonamdriskscore.org/. The contribution of environmental and genetic factors has been
assessed independently, and the authors conclude that combining both gives a better prediction.
Conclusion
The pathophysiology of AMD remains largely unknown and genetics is thought to
be responsible for around 70%.
References
1. Leibowitz HM, Krueger DE, Maunder LR, et al. The Framingham Eye Study monograph: An ophthalmological and epidemiological study of cataract, glaucoma, diabetic retinopathy,
macular degeneration, and visual acuity in a general population of 2631 adults, 1973-1975. Surv Ophthalmol. 1980;24 (Suppl): 335-610.
2. Bressler NM, Bressler SB, West SK, et al. The grading and prevalence of macular degeneration in Chesapeake Bay watermen. Arch Ophthalmol Chic Ill 1960. 1989;107(6):847-852.
3. Klein R, Klein BE, Linton KL. Prevalence of age-related maculopathy. The Beaver Dam Eye Study. Ophthalmology. 1992;99 (6): 933-943.
4. Augood CA, Vingerling JR, de Jong PTVM, et al. Prevalence of age-related maculopathy in older Europeans: the European Eye Study (EUREYE). Arch Ophthalmol Chic Ill 1960. 2006;
124(4):529-535.
5. Wong WL, Su X, Li X, et al. Global prevalence of age-related macular degeneration and disease burden projection for 2020 and 2040: a systematic review and meta-analysis. Lancet
Glob Health. 2014;2 (2):e106-116. doi:10.1016/S2214-109X(13)70145-1.
6. Klein R, Peto T, Bird A, Vannewkirk MR. The epidemiology of age-related macular degeneration. Am J Ophthalmol. 2004;137(3):486-495.
7. Klein R, Klein BE, Jensen SC, et al. Age-related maculopathy in a multiracial United States population: the National Health and Nutrition Examination Survey III. Ophthalmology. 1999;
106(6):1056-1065.
8. Klein R, Klein BEK, Tomany SC, Cruickshanks KJ. The association of cardiovascular disease with the long-term incidence of age-related maculopathy: the Beaver Dam Eye Study.
Ophthalmology. 2003;110(6):1273-1280.
9. Clemons TE, Kurinij N, Sperduto RD, AREDS Research Group. Associations of mortality with ocular disorders and an intervention of high-dose antioxidants and zinc in the Age-Related
Eye Disease Study: AREDS Report No. 13. Arch Ophthalmol Chic Ill 1960. 2004;122(5):716-726.
10. Clemons TE, Milton RC, Klein R, et al. Age-Related Eye Disease Study Research Group. Risk factors for the incidence of Advanced Age-Related Macular Degeneration in the Age-
Related Eye Disease Study. AREDS report no. 19. Ophthalmology. 2005;112(4):533-539.
11. Hogg RE, Woodside JV, Gilchrist SECM, et al. Cardiovascular disease and hypertension are strong risk factors for choroidal neovascularization. Ophthalmology. 2008;115(6):1046-
1052.e2.
12. Tan JSL, Mitchell P, Smith W, Wang JJ. Cardiovascular risk factors and the long-term incidence of age-related macular degeneration: the Blue Mountains Eye Study. Ophthalmology.
2007;114(6):1143-1150.
13. Souied EH, Benlian P, Amouyel P, et al. The epsilon4 allele of the apolipoprotein E gene as a potential protective factor for exudative age-related macular degeneration. Am J Ophthalmol.
1998;125(3):353-359.
14. Klaver CC, Kliffen M, van Duijn CM, et al. Genetic association of apolipoprotein E with age-related macular degeneration. Am J Hum Genet. 1998;63(1):200-206.
15. Thornton J, Edwards R, Mitchell P, et al. Smoking and age-related macular degeneration: a review of association. Eye Lond Engl. 2005;19(9):935-944.
16. Chong EW-T, Kreis AJ, Wong TY, et al. Alcohol Consumption and the Risk of Age-Related Macular Degeneration: A Systematic Review and Meta-Analysis. Am J Ophthalmol. 2008;
145(4):707-715.e2.
17. Knudtson MD, Klein R, Klein BEK. Alcohol consumption and the 15-year cumulative incidence of age-related macular degeneration. Am J Ophthalmol. 2007;143(6):1026-1029.
18. Peeters A, Magliano DJ, Stevens J, et al. Changes in abdominal obesity and age-related macular degeneration: the Atherosclerosis Risk in Communities Study. Arch Ophthalmol Chic
Ill 1960. 2008;126(11):1554-1560.
19. Chong EW-T, Kreis AJ, Wong TY, et al. Dietary omega-3 fatty acid and fish intake in the primary prevention of age-related macular degeneration: a systematic review and meta-analysis.
Arch Ophthalmol Chic Ill 1960. 2008;126(6):826-833.
Multimodal imaging
for AMD
PART 1:
Non-invasive imaging
Setha Vo Kim, Francesca Amoroso
1. Fundus photography
1) Conventional fundus cameras
Colour fundus photographs are acquired using a white light fundus camera. Various monochrome filters
can be used, highlighting particular structures in the fundus and therefore making it easier to assess them.
This procedure complements dilated fundus examination (Figure 1) and is a highly effective screening
and diagnostic tool, particularly when combined with the remote transmission of digitised data. It can
also be used to objectively monitor retinal diseases over time. Its efficacy in terms of screening for and
monitoring AMD has been demonstrated in the literature1,2 (Figures 2, 3 and 4).
Fig. 2: Soft drusen in the posterior pole. Fig. 3: Adult-onset foveomacular vitelliform dystrophy:
subfoveal material deposit associated with fine drusen.
Fig. 5: Infrared photographs showing epiretinal membrane Fig. 6: Infrared photograph showing a pigment epithelial
folds. detachment.
Fig. 7(A): Red free photograph showing a subretinal Fig. 7(B): Red free photograph of discrete intraretinal
haemorrhage secondary to type 2 choroidal neovascularization. haemorrhage suggesting a type 3 neovascularization,
associated with drusen.
Fig. 8(A): Fundus photograph of adult-onset foveomacular Fig. 8(B): The vitelliform material disappears under blue light.
vitelliform dystrophy associated with reticular pseudodrusen.
The reticular pseudodrusen can be seen in the blue photograph.
Fig. 9(A): Fundus photograph showing retinal pigment Fig. 9(B): The pigment changes are highlighted in the red filter
epithelium changes. image.
Unlike with conventional fundus cameras, the addition of a confocal filter prevents reflected rays from
structures located in sections of the eye other than the posterior pole from reaching the sensor, resulting
in a clear, reliable image even in the presence of media opacities. Images can be generated by associating
monochrome laser sources (red, green and blue lasers), as with the “pseudocolour” MultiColor Imaging
Spectralis® images (Heidelberg, Germany) (Figures 10 and 11), or using white light, as with the Eidon®
system (Centrevue SpA, Padova, Italy).
Fig. 10: Posterior pole Drusen, Multicolor® mode. Fig. 11: Geographic atrophy of the posterior pole, Multicolor®
mode: the choroidal vessels are visible.
The accumulation of lipofuscin and A2-E observed in AMD6 causes pigment epithelium dysfunction7.
Images can be acquired using a conventional fundus camera or a scanning laser ophthalmoscope (SLO)8.
In the presence of retinal atrophy, the disappearance of the retinal pigment epithelium results in a dark,
hypoautofluorescent area. As the disease progresses, a hyperautofluorescent outline can be seen,
reflecting the accumulation of A2-E where the pigment epithelium is becoming damaged (Figure 13).
This indicates that the atrophy is likely to spread in future9.
Fig. 12: Fundus autofluoresnce imaging of the posterior pole in Fig. 13: Fundus autofluorescence imaging of progressive
a healthy subject. geographic atrophy (dry AMD): the hyperautofluorescent
border around the atrophy reveals a ring of pigment epithelium
alteration surrounding the atrophy.
At the end of the 1990s, the first devices were based on analysing the time between the emitted and
reflected beams using a mobile reference mirror. The acquisition speed was 400 A-scan per second, with
a resolution of 10-15 μm.
Since 2007, OCT devices technology is based on the frequency of the refracted rays using the Fourier
transform. This improves the acquisition speed (up to 85,000 A-scan per second) and the image definition.
The axial resolution of SD-OCT is around 3–5 µm and its longitudinal resolution is 15–20 µm, enabling
quasi-histological resolution of the retinal layers (Figure 15). AMD can be characterised by one or more
retinal pigment epithelium (RPE) elevations (Figure 16), the appearance of intraretinal or subretinal
exudative signs14 (Figure 17) and/or areas of retinal atrophy15 (Figure 18). From a qualitative perspective,
treated pathological areas can be analysed and compared precisely over time, using the eye tracking
system. Macular changes can be measured quantitatively: the volume or surface area of drusen16 and
neovascular membranes can be estimated, enabling more precise monitoring.
Fig. 18: Dry AMD. Atrophy of the RPE and outer layers (area between the arrows), with choroidal hyperreflectivity due to the
window effect.
Fig. 19: Polyp located between the optic disc and the maculaassociated with pachychoroid.
References
1. Age-Related Eye Disease Study Research Group. The Age-Related Eye Disease Study System for Classifying Age-Related Macular
Degeneration from Stereoscopic Color Fundus Photographs: The Age-Related Eye Disease Study Report Number 6. American Journal of
Ophthalmology.2001;132(5):668-81.
2. A Pirbhai, T Sheidow, et P Hooper. Prospective Evaluation of Digital Non-Stereo Color Fundus Photography as a Screening Tool in Age-Related
Macular Degeneration. American Journal of Ophthalmology.2005.139(3):455-61.
3. A Ly, L Nivison‐Smith, N Assaad, et M Kalloniatis. Infrared reflectance imaging in age‐related macular degeneration. Ophthalmic & Physiological
Optics. 2016;36(3):303-16.
4. Mimoun, G., G. Soubrane, et G. Coscas. [Macular drusen]. Journal Francais D’ophtalmologie. 1990;13(10):511-30.
5. NM Pumariega, R. T Smith, MA Sohrab, et al. A PROSPECTIVE STUDY OF RETICULAR MACULAR DISEASE. Ophthalmology.2011(8):1619–1625.
7. F Schütt, S Davies, J Kopitz, et al. Photodamage to Human RPE Cells by A2-E, a Retinoid Component of Lipofuscin. Investigative Ophthalmology
& Visual Science.2000;41(8):2303-8.
8. S Schmitz-Valckenberg, M Fleckenstein, AP Göbel, et al. Evaluation of Autofluorescence Imaging with the Scanning Laser Ophthalmoscope and
the Fundus Camera in Age-Related Geographic Atrophy. American Journal of Ophthalmology.2008;146(2):183-92.
9. FG Holz, A Bindewald-Wittich, M Fleckenstein, et al. et FAM-Study Group. Progression of Geographic Atrophy and Impact of Fundus
Autofluorescence Patterns in Age-Related Macular Degeneration. American Journal of Ophthalmology 2007;143(3):463-72.
10. MR Hee, JA Izatt, EA Swanson, et al. Optical Coherence Tomography of the Human Retina. Archives of Ophthalmology.1995;113(3):325-32
11. CA Puliafito, MR Hee, CP Lin, et al. Imaging of Macular Diseases with Optical Coherence Tomography. Ophthalmology.1995;102(2):217-29.
12. Hee, M. R., C. R. Baumal, C. A. Puliafito, J. S. Duker, E. Reichel, J. R. Wilkins, J. G. Coker, J. S. Schuman, E. A. Swanson, et J. G. Fujimoto. Optical
Coherence Tomography of Age-Related Macular Degeneration and Choroidal Neovascularization. Ophthalmology. 1996;103(8): 1260-70.
13. D Huang, EA Swanson, CP Lin, J., et al. Optical Coherence Tomography. Science.1991;254(5035): 1178-81.
14. MM Castillo, G. Mowatt, N. Lois, et al. Optical Coherence Tomography for the Diagnosis of Neovascular Age-Related Macular Degeneration: A
Systematic Review. Eye.2014;28(12):1399-1406.
15. K Sleiman, M Veerappan, KP. Winter, et al. Age-Related Eye Disease Study 2 Ancillary Spectral Domain Optical Coherence Tomography Study
Group. Optical Coherence Tomography Predictors of Risk for Progression to Non-Neovascular Atrophic Age-Related Macular Degeneration.
Ophthalmology. 2017;124(12): 1764-77.
16. G Gregori, F Wang, PJ Rosenfeld, Z Yehoshua, et al. Spectral Domain Optical Coherence Tomography Imaging of Drusen in Non-Exudative Age-
Related Macular Degeneration. Ophthalmology.2011; 118(7): 1373-79.
17. RF Spaide, H Koizumi, MC Pozzoni. Enhanced Depth Imaging Spectral-Domain Optical Coherence Tomography. American Journal of
Ophthalmology. 2008 146(4): 496-500.
PART 2:
Optical coherence
tomography angiography
(OCTA)
Alexandra Miere, Roxane Bunod, Eric Souied
1. Principle
OCTA is a new retinal imaging technique that generates images of the retinal and choroidal micro-
vasculature using sequential B-scans of the same retinal scan, in order to detect motion contrast1. Unlike
moving objects, stationary objects do not change from one image to the next and are therefore interpreted
as having no flow. Blood circulating in the retinal microvasculature, however, creates fluctuations in
reflectivity from one image to the next, which is interpreted as flow. The specific area in the retina can be
assessed by repeating the analysis on several adjacent slices. A “split-spectrum amplitude decorrelation
angiography” (SSADA) algorithm is used to calculate the difference in reflectivity between several
sequential acquisitions from the same section of the retina2.
2. OCTA images
OCTA data are viewed by creating segmented B-scan scans from different layers in the retina. The layers
of the retina between two segmentation lines are then projected for en-face visualisation1. Most of the
currently available OCTA devices offer several automatic segmentation options (Figures 1 and 2):
- Superficial capillary plexus
- Intermediate capillary plexus
- Deep capillary plexus
- Outer retina
- Choriocapillaris
- Custom segmentation
A structural B-scan is obtained simultaneously, providing real-time verification of the exact location and
depth of the various structures. Figure 3 shows the multimodal imaging of an adult-onset foveomacular
vitelliform dystrophy case complicated by choroidal neovascularization visualized on OCTA.
In addition to qualitative analysis based on the structural B-scan and en-face projections, quantitative
markers for (neo)vascular disease have been developed, including vessel density, area of the foveal
avascular zone, non-perfusion area and neovascular area.
Several OCTA devices using different algorithms and wavelengths are currently available. Swept-source
OCTA devices have a wavelength of 1050 nm, which is higher than that of spectral-domain machines
(840 nm). Devices that use swept-source technology therefore offer better penetration through the
choroid than spectral-domain devices3.
A B C
Fig. 2: OCTA wide-field montage (PlexElite, Zeiss) from a male patient with proliferative diabetic retinopathy.
(A): Vitreoretinal interface slab. Multiple areas of pre-retinal neovascularization can be seen.
(B): Superficial capillary plexus slab. Areas of ischaemia are visible.
(C): Deep capillary plexus slab.
A B C
D E
3. Artefacts
There are several different kinds of artefact that can impact the OCTA signal, regardless of which device
is used. The most common type of artefacts are projection artefacts. Projection artefacts aoriginates in
the hyperreflective RPE; variations of the reflected light off the hyperreflective RPE would be interpreted
as movement by the OCTA device, therefore generating the projection artefact4. Projection artefacts have
been described in chorioretinal atrophy, drusen and drusenoid pigment epithelial detachment.
Other OCTA artefacts have been described, including artefacts linked to image acquisition, the intrinsic
properties of the eye and eye disease (opacity of the media, vitreous haemorrhage), eye motion, as well
as image processing and display strategies. The latter category includes automatic segmentation errors,
thresholding and saturation artefacts.4
4. In practice
When using OCTA, it is therefore important to:
• Carefully identify any false positives and false negatives generated by artefacts
• Acquire good quality images (> 5/10)
• Analyse en-face flow images with the corresponding B-scan.
References
1. Spaide RF, Fujimoto JG, Waheed NK, et al. Optical coherence tomography angiography. Prog Retin Eye Res. 2018;64:1-55.
2. Huang D, Jia Y, Gao SS, Lumbroso B, Rispoli M. Optical coherence tomography angiography using the Optovue device. Dev Ophthalmol. 2016;56:
6-12.
3. Novais EA, Adhi M, Moult EM, Louzada RN, Cole ED, Husvogt L, Lee B, Dang S, Regatieri CV, Witkin AJ, Baumal CR, et al. Choroidal
Neovascularization Analyzed on Ultrahigh-Speed Swept-Source Optical Coherence Tomography Angiography Compared to Spectral-Domain
Optical Coherence Tomography Angiography. Am J Ophthalmol. 2016;164:80-8.
4. Spaide RF, Fujimoto JG, Waheed NK. Image artifacts in optical coherence tomography angiography. Retina 2015; 35: 2163‑2180.
Multimodal imaging
for AMD
PART 3:
Invasive imaging
Setha Vo Kim, Francesca Amoroso
1. Fluorescein angiography
Although two students at the University of Indiana, Novotny and Alvis, had initially described the
technique of fluorescein angiography (FA) in 19611, it was Donald Gass, in 1967, who really introduced it
as a key retinal imaging technique2. As technology developed, the utility of the procedure was confirmed.
Fluorescein angiography requires the use of a fundus camera equipped with excitation and barrier filters.
Fluorescein is injected intravenously, generally into the antecubital area. A blue excitation filter allows
white light to pass through; the unbound fluorescein molecule then absorbs the blue light (465–
490 nm) and emits light with a longer wavelength in the yellow-green spectrum (520–530 nm).
A barrier filter blocks out any other reflected wavelengths to ensure that the images capture only the light
emitted by the fluorescein.
The images are acquired immediately after the injection and over a total time of 6 to 15 minutes,
depending on the disease. They are recorded digitally or on 35 mm film. The complications described
after the injection are temporary nausea (3 to 15% of patients), vomiting (7%)3 and pruritus. More severe
reactions are rare and include urticaria, fever, thrombophlebitis and syncope. Local tissue necrosis can
occur in the event of extravasation of the dye, but mild pain and redness are more common. Anaphylaxis,
cardiac arrest and bronchospasm can occur, but are extremely rare.
Fig. 1: Example of hypofluorescence in fluorescein angiography: drusenoid PED. Multiple areas of hyperfluorescence around the
macula in the early frame (A), increasing during the examination (B), with no late leakage(C). The central hypofluorescence is due to
a blocking effect caused by the overlying drusenoid PED, and the focal inferonasal hyperfluorescence is linked to a window effect
caused by retinal pigment epithelium atrophy.
Fig. 2: Example of leakage or diffusion: type 2 choroidal neovascularization secondary to exudative AMD. Fluorescein angiography
typically clearly shows the neovascular membrane very clearly from the early frame. Here, it is accompanied by cystoid spaces.
Note how the hyperfluorescent neovascular membrane contrasts with the surrounding hypofluorescent ring (A). The fluorescence
increases over the course of the angiogram (B) and, in the late frame, intense leakage can be seen as the dye passes into the
subretinal space (C).
Fig. 3: Example of pooling: serous PED. During the early phase, the PED appears hypofluorescent (A). Over the course of the
angiogram, the fluorescence gradually increases (B), with hyperfluorescence in the late phase, due to gradual pooling of the PED.
Note that the edges of the PED are well marked and the staining is uneven, due to blockage caused by the presence of the pattern
disposition of vitteliform material (C).
Hyperfluorescence occurs as a result of leakage, impregnation (or staining), accumulation (or pooling),
or increased fluorescein transmission (window effect).
The phenomenon of fluorescein leakage is generally caused by immature blood vessels (type 2 choroidal
neovascularisation secondary to AMD, pre-retinal neovascularization in the context of proliferative
diabetic retinopathy) or in a retinal pigment epithelium tear. The areas of leakage are characterised by
fluorescence that increases in intensity and size over the course of the examination, with blurred edges
(Figure 2).
Staining is an increase in fluorescence over the course of the angiogram, with edges that remain sharp.
Normal structures such as the optic nerve head and the sclera show physiological fluorescein staining.
Pooling occurs when the fluorescein gradually fills a fluid-filled space (umbrella shaped, leakage point in
central serous chorioretinopathy, serous PED (Figure 3).
Transmission or a window effect occurs when the RPE is damaged and choroidal fluorescence is
observed during the early frame of the examination. The intensity of the fluorescence does not change
during the angiogram and the edges remain sharp (Figure 4).
ICG angiography uses a diode laser illumination system (805 nm) and barrier filters at 500 nm and
810 nm. Modern tools have enabled high-speed ICG angiography, which can generate up to 30 images
per second with continuous recording of the angiogram. This system is very useful for visualising
structures that appear only briefly, such as feeder vessels of choroidal neovascularization (Figure 5).
Due to its longer wavelength, ICG angiography can sometimes offer a better visualization compared to
fluorescein angiography of the choroidal neovascularization associated with haemorrhage or hyperplasia
of the RPE. It is also useful for pigment epithelial detachments and occult choroidal neovascularization
(Figure 6).
In the differential diagnosis of neovascular AMD, early ICGA hyperfluorescence with a “hot spot” in the
late frames of the angiogram is highly suggestive of chorioretinal anastomosis (type 3 neovascularization)
(Figure 7)8. In contrast, early hyperfluorescence with a late “wash-out” of the neovascular lesion suggests
the presence of polypoidal choroidal vasculopathy (Figure 8)9.
Fig. 6: Indocyanine green angiography of a male patient with type 1 choroidal neovascularization secondary to exudative AMD.
In the early frame, ICGA shows a hyperfluorescent neovascular membrane (A), which persists into the intermediate frame (B) and
increases in fluorescence in the late frame, generating a typical “late plaque” appearance (C).
Fig. 7: Indocyanine green angiography of a female patient with type 3 neovascularization (retinal angiomatous proliferation,
chorioretinal anastomosis). The early ICGA reveals a focal hyperfluorescence (A), its fluorescence increasing in the intermediate
frame (B) and with a visible ‘hot spot’ in the late frame (C).
Fig. 8: Indocyanine green angiography of a male patient with polypoidal choroidal vasculopathy.
The ICGA shows early pooling of the polypoidal lesions (red circle) and the abnormal branching vascular network (yellow arrow) (A),
persisting into the intermediate frame (B), with a wash-out in the late frames (C).
References
1. Novotny HR, Alvis DL. A method of photographing fluorescence in circulating blood in the human retina. Circulation. 1961;24:82-86.
2. Gass JDM, Sever, RJ, Sparks D, Goren J. A combined technique of fluorescein fundoscopy and angiography of the eye. Arch Ophthalmol.
1967;78:455-461.
3. Yannuzzi LA, Rohrer, MA, Tindel LJ, et al. Fluorescein angiography complication survey. Ophthalmology. 1986;93:611-7.
4. Schatz H, Burton TC, Yanuzzi LA, Rabb MF. Interpretation of fundus fluorescein angiography. St. Louis:Mosby-Year Book: 1978.
5. Gass JDM. Stereoscopic atlas of macular diseases: diagnosis and treatment, 4th edition. St. Louis:Mosby-Yearbook; 1997.
6. Gelisken F, Inhoffen W, Schneider U, et al. Indocyanine green angiography in classic choroidal neovascularization. Jpn J Ophthalmol 1998.
42:300-303.
7. Grossniklaus HE, Gass JD. Clinicopathologic correlations of surgically excised type 1 and type 2 submacular choroidal neovascular membranes.
Am J Ophthalmol 1998. 126:59-69.
8. Kuhn D, Meunier I, Soubrane G, Coscas G. Imaging of chorioretinal anastomoses in vascularized retinal pigment epithelium detachments. Arch
Ophthalmol. 1995. 113:1392-1398.
9. Spaide RF, Yannuzzi LA, Slakter JS, et al. Indocyanine green videoangiography of idiopathic polypoidal choroidal vasculopathy. 1995. Retina
15:100-10.
1. Definition
“Age-related maculopathy” (ARM), also known as early and intermediate AMD, covers all age-
related changes to the fundus preceding neovascular or atrophic AMD (i.e. late AMD). These changes
generally occur from fifty years of age.
ARM manifests as changes to the retinal pigment epithelium (RPE): RPE depigmentation or
hyperpigmentation (pigment clumps or migration). Drusen, including hard drusen, isolated or confluent
soft drusen, reticular pseudodrusen, cuticular drusen, refractile drusen and ghost drusen are hallmarks
of ARM. The presence of these lesions in the general population varies considerably by age1 (25% after
52 years to 87% after 80 years).
While these lesions do not always progress to late AMD, their presence is a risk factor that merits careful
analysis and precise monitoring. A score has been developed to predict exudative AMD based on an
analysis of the fundus. Each eye is scored between 0 and 2 based on the presence or absence of drusen
measuring over 125 microns or RPE changes, giving the patient a total score of between 0 and 4. Five-
year and ten-year neovascular risk was assessed in this way in the AREDS study2 (Tables 1 and 2).
Today, ARM is split into two entities: early AMD and intermediate AMD. Early AMD is characterised by
the presence of numerous small drusen (< 63 microns, known as “hard” drusen) or intermediate drusen
(≥ 63 microns but < 125 microns, known as “soft” drusen). Intermediate AMD is defined by the presence of
extensive small or medium-sized drusen, or by large drusen (≥ 125 microns). Small drusen are commonly
seen in people aged 50 years and over, and can be an epiphenomenon of ageing. Intermediate drusen
are therefore more specific and more likely to be a marker of age-related macular degeneration than of
normal ageing.
RPE changes are considered to be characteristic of early AMD and are defined as hyperpigmentation
and/or hypopigmentation in the macular area, two disc diameters from the fovea, with or without drusen
and with no other known retinal disease. Numerous studies7,8 show that there is a high risk of these
RPE changes progressing towards late AMD, therefore these RPE changes require a precise, regular
monitoring in order to identifiy complications as early as possible (Figure 1).
A B
Colour photograph (A): pigment migration visible in the form of a pigmented lesion (RPE clumps) in the central area, associated
with soft drusen. SD-OCT with corresponding infrared image, vertical B-scan (B, C), showing a focal hyperreflective visibility of the
RPE, associated with a posterior shadowing corresponding to the hyperpigmentation area (red arrow). The B-scan also reveals
dome-shaped elevations of the RPE corresponding to soft drusen.
2) Hard drusen
Hard drusen are found in 25 to 50% of people aged over 50. On fundus examination, these hard drusen
are round, small (diameter < 63 μm) and yellowish, with distinct edges. There are often numerous and
they are usually found in the temporal foveal area.
In fluorescein angiography they appear hyperfluorescent from the early frames of the angiogram,
due to a window effect. In SD-OCT, they create very small elevations in the RPE (Figure 2). They can
merge together and develop into larger drusen, and may be associated with soft drusen or reticular
pseudodrusen. Complications within time are relatively rare and happen at a late stage9 when the hard
drusen are isolated; they are generally stable.
A B
Colour photograph (A) showing round, regular, small (< 63 μm), yellowish hard drusen (white arrow). Fluorescein angiography (B):
the hard drusen appear hyperfluorescent from the beginning of the sequence and are easier to see than on the colour photograph
(white arrow). SD-OCT (C): slight RPE elevations indicating hard drusen (white circle).
Soft drusen are larger (diameter > 63 μm), with blurred edges and an irregular shape, and are paler than
hard drusen. They are often associated with pigment migration.
In fluorescein angiography, they are hypofluorescent during the early phase, filling throughout the
examination and becoming hyperfluorescent in the late frames. They remain hypofluorescent in all
phases of ICG angiography. OCT reveals multiple dome-shaped elevations in the RPE. Soft drusen are
moderately reflective and have a bumpy aspect. The external limiting membrane and the ellipsoid line
often remain visible (Figure 3A).
Confluent sof drusen result in drusenoid pigment epithelial detachments (PED), with a hyperreflective
content. Over several years, drusenoid PEDs can develop into extensive areas of atrophy, as the PED
gradually subsides (Figure 3B).
Soft drusen eventually progress to late AMD in around 30% of cases9. In rare cases, soft drusen can
regress or even calcify, resulting in ghost drusen associated with underlying atrophy.
A B C
Colour photograph (A) showing numerous lesions with blurred edges, paler than the hard drusen, measuring over 125 μm and
located in the macular area, corresponding to soft drusen (white circle). Fluorescein angiography (B), intermediate frame, showing
progressive and delayed staining of the soft drusen (white circle), with no leakage. Late frame of ICGA (C): the soft drusen remain
hypofluorescent (white circle). SD-OCT scan through the soft drusen (D): multiple dome-shaped RPE elevations (white arrow)
with homogeneous hyperreflective content. The external limiting membrane and ellipsoid line remain visible in this case, with no
associated exudation.
4) Reticular pseudodrusen
Reticular pseudodrusen, or blue drusen, were first described by the Ophthalmology Créteil team10. These
drusen are yellowish in colour and generally found on the superior temporal arcades. However, they may
also appear in the macula and retinal periphery and are “cerebroid” in appearance (Figure 4).
They are referred to as “blue” drusen because they are easily visible on fundus photographs with blue
illumination. In SD-OCT, they are hyperreflective. They can affect the ellipsoid zone and are associated
with thinning of the choroid in EDI-OCT.
These deposits are found above the RPE, unlike soft drusen, which are located below it. In fluorescein
angiography, they are either slightly visible or hypofluorescent. In the late phase of ICG angiography, they
appear as a hypofluorescent network.
Four different stages have been described in SD-OCT11. During the early stages (stages 1 and 2), the
hyperreflective material located above the RPE thickens over time, and during the advanced stages
(stages 3 and 4), the material breaks up, resulting in discontinuity or total absence of the ellipsoid line.
Several studies have found that reticular pseudodrusen were associated to a high risk of progression to
atrophic or exudative AMD12,13, and in particular to type 3 neovascularization.
A B C
Colour photograph (A): Lesion measuring approximately one disc diameter causing a drusenoid PED (red arrow) due to confluent
drusen, with pigment migration (white arrow). Intermediate frame of fluorescein angiography (B) showing a blocking effect caused
by the pigment migration (white arrow), contrasting with the progressive hyperfluorescence of the drusenoid PED (red arrow).
Infrared image and corresponding SD-OCT scan (C, D) showing considerable RPE elevation, resulting in a drusenoid PED (red
arrow) with pigment migration (white arrow). Two years later (E), the drusenoid PED has subsided and atrophy is starting to
develop (blue arrow), marked by the posterior hyperreflectivity.
The colour photograph (A) shows yellowish cerebroid pseudodrusen located on the superior temporal arcades (white star).
The pseudodrusen are easier to see in the blue-light photograph (B) (white star). In ICGA (C), reticular pseudodrusen appear
hypofluorescent during the late frame (white star). SD-OCT (D): the pseudodrusen are hyperreflective (white star), dense, irregular,
fusiform in appearance and located above the RPE, meaning they could impact on the ellipsoid zone and the photoreceptors. Blue
drusen are associated with thinning of the choroid in EDI-OCT.
D
A B C
The colour photograph (A) shows yellowish cerebroid reticular pseudodrusen, predominantly in the temporal region (blue arrow)
of the macula, associated with larger, round, yellowish deposits corresponding to soft drusen (red arrow). The pseudodrusen are
easier to see in the blue-light photo (B) (blue arrow). In the intermediate frame of fluorescein angiography (C), the soft drusen are
hyperfluorescent while the reticular pseudodrusen drusen are hypofluorescent (red arrow). In SD-OCT (D), the pseudodrusen are
dense, hyperreflective, irregular and fusiform in appearance, and located above the RPE (blue arrow), in contrast to the soft drusen
that appear as dome-shaped RPE elevations with homogeneous hyperreflective content below the RPE (red arrow).
5) Cuticular drusen
Cuticular drusen were first described in 1977 by D. Gass (Figure 6). These drusen are round, numerous,
small (25–75 microns) deposits concentrated primarily in the posterior pole, occasionally appearing in
the retinal periphery. They appear in younger patients than those who have AMD, with an average age
of 57 years. They carry a risk of developing into late AMD after the age of 70.
In angiography images, they have a characteristic “stars in the sky” appearance. Cuticular drusen are
located between the RPE and Bruch’s membrane. In OCT images, they have a very typical sawtooth
appearance, with a slight or mound-shaped elevation. The choroid displays typical “bar code” reflectivity14,
with alternating areas of hyper- and hyporeflectivity corresponding to the drusen.
They may be associated to vitelliform material, deposits or choroidal neovascularization (usually type 1),
as well as atrophy.
While their ultrastructural characteristics are similar to those of hard drusen, they are closer to soft drusen
in terms of their life cycle and macular complications.
A B
Colour photograph (A): Numerous round, yellowish deposits (25–75 microns) concentrated in the macula, corresponding to cuticular
drusen (blue arrow). Fluorescein angiography (B) shows the characteristic “stars in the sky” appearance, with a greater number of
lesions than in the colour photograph (blue arrow). SD-OCT (C): saw-tooth pattern with slight or mound-shaped elevation (blue
arrow) resulting in a “bar code” appearance (red parenthesis), with alternating areas of hyper- and hyporeflectivity corresponding
to the drusen.
Calcified drusen were first described by Gass in 1973 and were initially called “refractile” due to their
appearance in the fundus. More recently, the term “calcified drusen” has been replaced by “regressing
drusen”15.
Drusen are deposits of extracellular material rich in lipids (cholesterol) found between the basal
membrane of the RPE and the internal layer of Bruch’s membrane. Several phagocytic processes have
been observed, resulting in regression of the drusen in terms of size. Regressing drusen are deposits of
calcified lipid material (cholesterol) that has not been ingested by macrophages, characterized by the
presence of an intensely hyperreflective band in the choroid in SD-OCT images (Figure 7).
Multicolor® photograph showing a calcified drusen (white arrow) with a refringent appearance, with the OCT B-scan through the
lesion. The SD-OCT shows a band of intense hyperreflectivity corresponding to lipid material (cholesterol) that has not undergone
phagocytosis (white arrow).
7) Ghost drusen
Ghost drusen16 were first described by the Ophthalmology Créteil team and appear in OCT images as
dense, hyperreflective pyramidal structures located above the RPE. They are associated with atrophic AMD
(geographic atrophy). They are hard to visualize in fluorescein angiography and appear isofluorescent or
hyperfluorescent on a background of atrophy (Figure 8). In a recent study17 analysing the origins of ghost
drusen, the authors found that soft drusen were initially present in a small number of cases.
A B C
Multicolor® photograph (A): ghost drusen located primarily around the areas of atrophy (white arrow). Infrared photograph (B):
hyporeflective lesion surrounded by a hyperreflective area with several small hyperreflective dots indicating the location of the ghost
drusen (white arrow). Fundus autofluorescence image (C): the ghost drusen (white arrow) appear relatively hyperautofluorescent
compared to the superior adjacent atrophy. SD-OCT (D): hyperreflective pyramid structure characterised by a hyporeflective centre
(white arrow).
A B C
D E F
Fig. 9: Multimodal imaging of a male patient with soft drusen and a vascularised drusen.
ICG angiography, early frame (A) and late frame (B). A hyperfluorescent lesion can be seen in the early frame (green arrowhead),
corresponding to a plaque in the late frame of the examination. Multiple soft drusen can be seen in the form of early and late
hypofluorescence (white arrow). SD-OCT (C) shows drusenoid RPE elevations corresponding to the two observed lesions. OCTA
(D) reveals a discrete high flow lesion in the outer retina slab (green arrowhead). In the choriocapillaris slab with projection artefacts
(E) and without projection artefacts (F), the high flow lesion is visible and corresponds to a vascularised drusen. However, the
choriocapillaris signal is attenuated by the presence of soft drusen (white arrow).
Conclusion
Screening for ARM is an essential part of any eye examination in patients aged 55
years and over, based on dilated fundus examination or color fundus photography.
There are several types of drusen that carry a risk of progression to late AMD. The
risk of neovascular complication is very low with hard drusen. However, soft drusen
are more likely to develop into choroidal neovascularization.
OCT is an essential examination for confirming a diagnosis of ARM and ruling out
neovascular activity. If progression to exudative AMD is suspected, angiography
becomes less useful, as OCTA can be carried out to provide valuable information on
the whether neovascularization is present.
2. Ferris FL, Davis MD, Clemons TE, et al. A simplified severity scale for age-related macular degeneration: AREDS Report No. 18. Arch Ophthalmol
Chic Ill 1960. 2005;123(11):1570–4.
3. Li C-M, Clark ME, Chimento MF, Curcio CA. Apolipoprotein localization in isolated drusen and retinal apolipoprotein gene expression. Invest
Ophthalmol Vis Sci. 2006;47(7):3119-3128.
4. Haimovici R, Gantz DL, Rumelt S, et al. The lipid composition of drusen, Bruch’s membrane, and sclera by hot stage polarizing light microscopy.
Invest Ophthalmol Vis Sci. 2001;42(7):1592-1599.
5. Dentchev T, Milam AH, Lee VM-Y, et al. Amyloid-beta is found in drusen from some age-related macular degeneration retinas, but not in drusen
from normal retinas. Mol Vis. 2003;9:184-190.
6. Curcio CA, Messinger JD, Sloan KR, et al. Subretinal Drusenoid Deposits In Non-Neovascular Age-Related Macular Degeneration: Morphology,
Prevalence, Topography, And Biogenesis Model. Retina. 2013;33(2) : 265-76.
7. Klein R, Klein BEk, Linton KLP. Prevalence of Age-related Maculopathy-The Beaver Dam Eye Study. Ophthalmology. 1992 ;99(6): 933-943.
8. A Randomized, Placebo-Controlled, Clinical Trial of High-Dose Supplementation With Vitamins C and E, Beta Carotene, and Zinc for Age-Related
Macular Degeneration and Vision Loss. Arch Ophthalmol. 2001;119(10):1417-1436.
9. Klein R, Klein BEK, Knudtson MD, et al. Fifteen-year cumulative incidence of age-related macular degeneration: the Beaver Dam Eye Study.
Ophthalmology. 2007;114(2):253-262.
11. Querques G, Canouï-Poitrine F, Coscas F, et al. Analysis of progression of reticular pseudo-drusen by spectral domain-optical coherence
tomography. Invest Ophthalmol Vis Sci. 2012;53(3):1264-1270.
12. Zweifel SA, Imamura Y, Spaide TC, et al. Prevalence and significance of subretinal drusenoid deposits (reticular pseudo-drusen) in age-related
macular degeneration. Ophthalmol. 2010;117(9):1775-1781.
13. Klein R, Meuer SM, Knudtson MD, et al. The epidemiology of retinal reticular drusen. Am J Ophthalmol. 2008;145(2):317-326.
14. Balaratnasingam C, Cherepanoff S, Dolz-Marco R, et al. Cuticular Drusen: Clinical Phenotypes and Natural History Defined Using Multimodal
Imaging. Ophthalmology. 2018;125(1):100-118.
15. Querques G, Georges A, Ben Moussa N, et al. Appearance of regressing drusen on optical coherence tomography in age-related macular
degeneration. Ophthalmol. 2014;121(1):173-179.
16. Bonnet C, Querques G, Zerbib J, et al. Hyperreflective pyramidal structures on optical coherence tomography in geographic atrophy areas. Retina.
2014;34(8):1524-30.
17. Bottin C., Zambrowski O., Querques G, et al. Origins of Ghost Drusen: A follow-up Analysis. Ophthalmology @ Point of Care.
18. Alten F, Lauermann JL, Clemens CR, et al. Signal reduction in choriocapillaris and segmentation errors in spectral domain OCT angiography
caused by soft drusen. Graefes Arch Clin Exp Ophthalmol. 2017;255(12):2347-2355.
19. Alten F, Heiduschka P, Clemens CR, Eter N. Exploring choriocapillaris under reticular pseudo-drusen using OCT-Angiography. Graefes Arch Clin
Exp Ophthalmol. 2016;254(11):2165-2173.
20. Nesper PL, Soetikno BT, Fawzi AA. Choriocapillaris Nonperfusion is Associated With Poor Visual Acuity in Eyes With Reticular Pseudo-drusen.
Am J Ophthalmol. 2017;174:42-55.
21. Querques G, Souied EH. Vascularized Drusen: Slowly Progressive Type 1 Neovascularization Mimicking Drusenoid Retinal Pigment Epithelium
Elevation. Retina. 2015;35(12):2433-9.
22. Chris Or, Jeffrey Heier, Namrata Saroj, A. Yasin Alibhai, Nadia Waheed. Incidence of Vascularized Drusen in Non-Exudative Age-related Macular
Degeneration using Spectral Domain Optical Coherence Tomography (OCT) Angiography. Poster B0328 ARVO 2018.
Type 1
neovascularization
Type 1 neovascularization
PART 1:
Type 1 choroidal
neovascularization
Jean-Louis Bacquet, Alexandra Mouallem-Bézière
1. Definition
Type 1 or sub-RPE neovascularization is by far the most prevalent subtype in exudative AMD, ranging
from 60% to 85% at diagnosis depending on the series. It is defined as neovascular growth below the
retinal pigment epithelium (RPE). First described in 1987 and classified by Gass in 1994, it differs from
type 2 neovascularization, which occurs above the RPE (“visible” neovascularization), and from type 3
neovascularization, which occurs in the retina itself (from the deep retinal capillary plexus), meaning the
natural history and origin of type 3 is highly debated as to its occurence through chorioretinal anastomosis,
retinochoroidal anastomosis or intraretinal proliferation.
Some authors add to type 1 neovascularization the polypoidal choroidal vasculopathy (PCV) class, under
the term “aneurysmal type 1 CNV” (Freund et al.). Since type 1 neovascularization was first defined
based on its appearance in angiography images, the term “occult neovascularization” is sometimes used
instead, in contrast to “visible” or “classic” neovascularization located above the RPE (type 2).
Type 1 neovascularization, given its aspect in fluorescein angiography, is also frequently referred to in the
literature as “ill-defined”.
2. Epidemiology
According to a study conducted on a US population by Jung et al., involving 374 patients (two thirds of
whom 2/3 were female, 95% Caucasian), type 1 neovascularization accounts for 50% of all clinical forms
of exudative AMD, compared to 12% for type 2 (visible) neovascularization and 28% type 3 intraretinal
neovascularization. The remaining 10% comprises mixed forms that combine several subtypes. This was
the first time an epidemiological study found such a low frequency of type 1 CNV and such a high
proportion of type 3.
3. Clinical presentation
Type 1 neovascularization should be the first diagnosis considered when an older patient presents with
an acute or subacute macular syndrome. The signs and symptoms may include:
- Reduced visual acuity, primarily relating to near vision
- Metamorphopsia
- Scotoma/microscotoma
- Micropsia
- Dyschromatopsia
- Reduced contrast sensitivity.
These signs may be present to varying degrees and some may be absent. The patient may have very few
symptoms or be asymptomatic. Compared to type 2 (visible) neovascularization, the functional impact is
more insidious, developing gradually, and the patient may be unable to pinpoint when it started.
It is possible to observe signs of neovascularization in the fundus, including retinal oedema, serous retinal
detachment, macular haemorrhage and exudates. These are more common than with type 2 (visible)
neovascularization. However, these signs are often inconspicuous, and nonspecific. Cystoid macular
oedema or white retinal oedema is unusual in the early stages, but may be found if the diagnosis is made
later on.
Precursors to all forms of AMD, such as drusen and/or reticular pseudodrusen, will of course need to be
monitored for, along with retinal pigment epithelium defects (migrations or changes).
It should be noted that in France, we tend to separate “vascularised pigment epithelial detachment”
(v-PED) from other forms of neovascularization. This is not generally the case in English-language
publications, since “vascularised PED” simply refers to a vascularised elevation of the RPE. However,
vascularised PED and type 1 CNV can have a different prognosis and therapeutic response: v-PED
responds somewhat unpredictably to anti-VEGF treatment, particularly in terms of the risk of RPE tears.
Under biomicroscopy, vascularised PED appears in relief as a yellow-orange lesion. This lesion is typically
hyperfluorescent in fluorescein angiography, while OCT displays heterogeneous reflectivity below the
RPE, which can be associated with hyporeflective subretinal fluid.
These images enable us to identify multiple subtypes of type 1 CNV, which will be described in the next
few chapters (Figure 1).
“Pachychoroid
spectrum”
PCV, type 1
aneurysmal
Active CNV
Multilayer
Type 1 CNV PED
Vascularised
PED “Onion sign”
Quiescent
Wrinkled
PED
Tear -
opening
in PE
5. Physiopathology
This form of neovascularization is choroidal in origin. The currently accepted theory is that vascular
proliferation (angiogenesis) causes the Bruch’s membrane to break in a patient who has drusen and is
genetically predisposed.
The role of intraretinal inflammation (potentially visible in images as hyperreflective dots) and macrophages
is increasingly being demonstrated, particularly by Sennlaub et al. The strong genetic association with
mutations in certain complement factors reinforces this hypothesis.
6. Imaging
First described using angiography and histology, and later by OCT B-scans and, most recently, by OCTA,
the various clinical forms and types of neovascularization now play a crucial role in characterizing patients
suffering from exudative AMD (“neovascular phenotype”).
The main imaging characteristics of type 1 choroidal neovascularization are summarised below. Since
it is important to know where the CNV is located in order to interpret imaging examinations, note that
the biological substrate in the images is the location of the neovascularization: the new vessels emerge
from the choroid and extend under the RPE, detaching the RPE and the overlying neurosensory
retina (PED, SRD) (Figure 2).
A B
C D E
This clinical form most commonly appears as a hyperfluorescent late plaque in late frame of ICG angiography.
2) SD-OCT
Depending on which stage of neovascularisation the patient is experiencing, OCT-SD shows RPE
elevation or detachment in 98% of cases. This elevation is initially moderate, separated from Bruch’s
membrane by a hyporeflective space.
However, the PED can become very significant. Occult neovascularisation should be suspected if the
RPE appears irregular, scalloped, fragmented or thickened.
This type of neovascularisation is often associated with signs of exudative activity, including:
• Subretinal fluid (or serous retinal detachment, SRD)
• Hyperreflective intraretinal dots (according to certain authors)
• Pre-epithelial hyperreflectivity
• Cystoid spaces in the neurosensory retina
• Diffuse or localised increase in retinal thickness.
3) OCT-angiography
OCTA now plays a major role in characterising type 1 CNV. Numerous studies evidence its sensitivity and
call into question the pre-eminence of systematic angiography. The COFT-1 study found a sensitivity level
of 85.7% for OCTA combined with structural B-mode OCT compared to the gold standard (angiography).
OCTA alone was insufficiently sensitive (66.7%). The leading cause of false negatives is the neovascular
signal being blocked by a PED, which can disrupt the automatic segmentation (Figure 3). OCTA therefore
has a place within multimodal imaging, but is insufficient on its own to form a reliable diagnosis.
A B C
D E
Fig. 3: A, B, C: Attempt to visualise a neovascular network within a very high v-PED using manual segmentation. The
neovascularisation remains invisible.
D: Signs of exudative activity evident in structural OCT (intraretinal spaces, exudates, retinal thickening) with no neovascularisation
visible in OCTA.
E: Part of the neovascular network is unblocked and the image becomes clearer following manual correction of the segmentation
(left: automatic, right: manual).
7. Treatment
Patients with neovascular AMD and type 1 neovascularisation are today treated with intravitreal
injections of angiogenesis inhibitors, preferably using the “Initiation – Observation – Individualisation”
strategy recommended in the Fédération France Macula guidelines.
Comparative studies are still being conducted into the different treatment protocols (“pro re nata” or
“PRN”, “treat and extend”, “observe and plan”) and drugs used (anti-VEGF therapy). There is consensus
around the use of three anti-VEGF injections at one-month intervals to initiate treatment for type 1 CNV,
as per the French Society of Ophthalmology guidelines.
The prognosis for type 1 neovascularisation appears to be good, with a positive response to anti-VEGF
therapy in general and less atrophy at follow-up compared to type 3 neovascularisation. Only a minority
of patients (around 5%) require very few IVIs to maintain a “dry” retina (no exudative relapse). A larger
proportion, with more severe forms, require monthly injections, sometimes for several years (around 10%
of cases). Between these two extremes, the majority of patients with type 1 CNV require between six
and eight IVIs on average, administered to an individualised schedule.
Dietary supplements rich in antioxidants, lutein, zeaxanthin and omega-3 have proven effective in terms
of prevention (see dedicated chapter), particularly with regard to the contralateral eye.
8. Future developments
The treatment of type 1 CNV will certainly develop in the future, with:
• Ever more precise characterisation of each patient’s neovascular phenotype
• Quantitative imaging analysis (in particular of the isolated neovascular network) to determine
therapeutic response anatomically within the neovessels and not solely based on the retina
• The establishment of phenotype/genotype correlations
• Therapies adapted to the genetic profile and vascular phenotype of each patient.
References
Dansingani KK, Gal-Or O, Sadda SR, et al. Understanding aneurysmal Type 1 neovascularization (polypoidal choroidal vasculopathy): a lesson in the
taxonomy of ‘expanded spectra’ - a review. Clin Exp Ophthalmol. 2017;46(2):189-200.
Jung J., Chen C., Mrejen S., et al. The Incidence of Neovascular Subtypes in Newly Diagnosed Neovascular Age-Related Macular Degeneration.
American Journal of Ophthalmology, 2014, Pages 769-779.e2
Coscas G, De Benedetto U, Coscas F, et al. (2013) Hyperreflective dots: a new spectral-domain optical coherence tomography entity for follow-up and
prognosis in exudative age-related macular degeneration. Int J Ophthalmol 229:32–7.
Ores R, Puche N, Querques G, et al. (2014) Gray hyper-reflective subretinal exudative lesions in exudative age-related macular degeneration. Am J
Ophthalmol 158:354–61.
Kuehlewein L, Bansal M, Lenis LT, et al. (2015) Optical Coherence Tomography Angiography of Type 1 Neovascularization in Age-Related Macular
Degeneration Am J Ophthalmol 160: 739–748.
Mrejen S, Sarraf D, Mukkamala SK, Freund KB (2013). Multimodal imaging of pigment epithelial detachment: a guide to evaluation. Retina 33:1735–
1762.
Pang CE, Messinger JD, Zanzottera EC et al. (2015). The Onion Sign in Neovascular Age-Related Macular Degeneration Represents Cholesterol
Crystals. Ophthalmology 122:2316-26.
Querques G, Capuano V, Costanzo E, et al. 2016. Retinal pigment epithelium aperture : A Previously Unreported Finding in the Evolution of Avascular
Pigment Epithelium Detachment. Retina 36 Suppl 1:S65-S72.
Lam D, Semoun O, Blanco-Garavito R, et al. 2017, Wrinkled vascularized retinal pigment epithelium detachment prognosis after intravitreal anti-VEGF
therapy. [Epub ahead of print]
Querques G, Srour M, Massmaba N, et al. 2013, Functional characterization and multimodal imaging of treatment-naïve quiescent choroidal
neovascularization. Invest Ophthalmol Vis Sci 21;54:6886-92.
Semoun O, Cohen SY, Srour M, et al. Comité scientifique de la Fédération France Macula. Individualized management of patients with exudative AMD,
IOI protocol: Injection-observational-individualization. J Fr Ophtalmol. 2017;40(3):169-176.
Inoue M, Jung JJ, Balaratnasingam C, et al. COFT-1 Study Group, 2016. A comparison between optical coherence tomography angiography and
fluorescein angiography for the imaging of Type 1 neovascularization. Investig. Opthalmology Vis. Sci. 2016 ;57,314-323.
Type 1 neovascularisation
PART 2:
Quiescent type 1 choroidal
neovascularisation
Vittorio Capuano
1. Diagnosis
In 2013, Giuseppe Querques and the Ophthalmology Créteil team1 described the functional and imaging
characteristics and natural progression of quiescent type 1 choroidal neovascularisation (CNV) in the
context of intermediate age-related macular degeneration (AMD) (Figure 1).
1. In fluorescein angiography (FA), quiescent CNV appears as hyperfluorescent lesions with distinct
edges and no late-phase leakage
3. In optical coherence tomography (OCT), quiescent CNV appears as a slight pigment epithelial
detachment (PED) with a horizontal main axis, with no signs of subretinal or intraretinal exudation.
This PED displays moderate reflectivity, allowing Bruch’s membrane to be seen.
A D J G
B E K H
C F L I
Fig. 1: Multimodal imaging of quiescent type 1 choroidal neovascularisation and progression over time.
In fluorescein angiography (A, F), the quiescent CNV appears as hyperfluorescent lesions with distinct edges and no late-phase
leakage (A, B, C). In indocyanine green angiography (ICG), the quiescent CNV is characterised by a hyperfluorescent network in the
early and intermediate phases (D, E, F), and a hyperfluorescent plaque in the late phases (G, H, I). In optical coherence tomography
(OCT), the quiescent CNV appears as a slight pigment epithelial detachment (PED) with a horizontal main axis, with no signs of
subretinal or intraretinal exudation. This PED displays moderate reflectivity, allowing Bruch’s membrane to be seen (J, K, L) (HRA,
Heidelberg Engineering).
1. No patient symptomatology
1. No exudative signs for at least six months. However, certain authors3 refer to quiescent CNV from the
first observation, without waiting for this time to elapse
Three years later, Palejwala et al.4 published the first OCTA image of quiescent CNV. This non-invasive
examination quickly became the gold standard for diagnosing and monitoring quiescent CNV. In 2016,
Carnevali5 calculated the sensitivity and specificity of OCTA (non-invasive examination) versus standard
imaging techniques (FA, ICG, OCT) (invasive examination) used during the initial diagnosis, finding a
detection rate of 81.8% and 100%, respectively. Quiescent CNV typically appears in OCTA as regular
lesions with well defined edges and no peripheral arcades. A feeder vessel is a common finding (Figure 2).
Quiescent CNV is not exclusive to “intermediate” AMD; it has also been described with the late atrophic
form6 (Figures 3 and 4) and with pachychoroid diseases7. More recently, a case was reported with angioid
streaks8. In these cases, the main and secondary characteristics are similar to those defined for AMD, but
with certain peculiarities. In particular, with atrophic AMD, OCTA is less sensitive than with intermediate
AMD (68% versus 81%), but has the same specificity (100%).
In the only longitudinal study6 conducted on patients with quiescent CNV and atrophic AMD, the patients
retained their visual acuity longer than patients without foveal involvement. The hypothesis here is that
quiescent CNV could be a protective factor against atrophic progression, due to the blood flow feeding
the choriocapillaris and photoreceptors.
In patients with a pachychoroid and quiescent CNV, the most common characteristic in OCTA is a
hypersignal with distinct edges and a fairly irregular shape within a PED. In late-phase ICG angiography,
hypofluorescence due to wash-out of the lesion has also been described7.
A B
Fig. 2: Intermediate-phase indocyanine green angiography (ICG) and optical coherence tomography angiography (OCTA) of
quiescent type 1 choroidal neovascularisation (CNV) (multimodal imaging of the patient in Figure 1).
In ICG angiography, the quiescent type 1 CNV is characterised by a hyperfluorescent network in the early and intermediate
phases (A: HRA, Heidelberg Engineering). In OCTA, the same lesion is characterised by a regular shape with well defined edges
and a feeder vessel, with no peripheral arcades (B: PLEXelite, Carl Zeiss) (C: rtvue, Optovue).
2. Monitoring
Bimonthly monitoring from the first visit onwards and for at least six months is strongly recommended.
If exudative signs appear during this monitoring period, “early detection of CNV” is the preferred term.
From the sixth month onwards, the diagnostic criteria for quiescent CNV have been fulfilled. Quiescent
CNV typically remains stable (with no exudation over the long term), while the size of the neovascular
membrane increases. Quarterly monitoring is nevertheless indicated. The complication rate varies from
6.6% for exudation associated with intermediate AMD to 26% with atrophic AMD. If exudation occurs
after the sixth month, the preferred term is “conversion of quiescent CNV to active (or exudative)”.
However, this nomenclature is not accepted by certain authors3.
A B
Fig. 3: Multimodal imaging of quiescent type 1 choroidal neovascularisation (CNV) associated with atrophic age-related macular
degeneration.
In autofluorescence, the quiescent type 1 CNV is not visible (A). In fluorescein angiography (FA), the quiescent CNV appears as
hyperfluorescent lesions with distinct edges and no late-phase leakage. Hyperfluorescence caused by atrophy of the retinal pigment
epithelium can also be seen (B). In indocyanine green angiography (ICG), the quiescent CNV is characterised by a hypercyanescent
plaque in the late phases (C). In optical coherence tomography (OCT), the quiescent CNV appears as a slight pigment epithelial
detachment (PED) with a horizontal main axis, with no signs of subretinal or intraretinal exudation. This PED displays moderate
reflectivity, allowing Bruch’s membrane to be seen. Around the membrane, the retina has thinned and light transmission has
increased due to atrophy of the retinal pigment epithelium (HRA, Heidelberg Engineering).
2015 2016
2017 2018
Fig. 4: Optical coherence tomography angiography (OCTA) of quiescent type 1 choroidal neovascularisation (CNV) associated with
atrophic age-related macular degeneration (AMD) (multimodal imaging of the patient in Figure 3).
In OCTA, the quiescent type 1 CNV associated with AMD is characterised by a fairly irregular shape with well defined edges and
no feeder vessel or peripheral arcades. Its flow appears “filiform”. This lesion will grow over time without changing in appearance
(C: Rtvue, Optovue).
3. Treatment
The gold standard treatment for “activated” quiescent CNV is a series of anti-VEGF intravitreal injections,
which has been shown to be effective in reducing intraretinal and subretinal signs of exudation1,6,7.
Conclusion
Quiescent CNV is a new entity described in conjunction with intermediate and
late AMD, as well as with pachychoroid diseases and angioid streaks. It refers to
treatment-naive CNV (visible in ICG angiography and OCTA) that shows no signs
of exudation in OCT or leakage in FA.
References
1. Querques G et al. Functional characterization and multimodal imaging of treatment-naive «quiescent» choroidal neovascularization. Invest
Ophthalmol Vis Sci 2013; 54(10):6886-6892.
2. Gass JDM. Stereoscopic Atlas of Macular Diseases. Diagnosis and Treatment. Mosby, St Louis; 1997.
3. de Oliveira Dias JR. et al. Natural History of Subclinical Neovascularization in Nonexudative Age-Related Macular Degeneration Using Swept-
Source OCT Angiography. Ophthalmology. 2018 Feb;125(2):255-266.
4. Palejwala NW et al. Detection of non exudative choroidal neovascularization in age-related macular degeneration with optical coherence
tomography angiography. Retina. 2015;35(11):2204-11.
5. Carnevali A et al. Optical Coherence Tomography Angiography: A Useful Tool for Diagnosis of Treatment-Naïve Quiescent Choroidal
Neovascularization. Am J Ophthalmol. 2016;169:189-198.
6. Capuano V. et al. Treatment-Naïve Quiescent Choroidal Neovascularization in Geographic Atrophy Secondary to Nonexudative Age-Related
Macular Degeneration. Am J Ophthalmol. 2017;182:45-55.
7. Carnevali, A. et al. OCT Angiography of Treatment-Naïve Quiescent Choroidal Neovascularization in Pachychoroid Neovasculopathy.
Ophthalmology Retina, 1(4), 328–332.
8. Mentes J, et al. Multimodal imaging characteristics of quiescent Type 1 neovascularization in an eye with angioid streaks. Am J Ophthalmol Case
Rep. 2018 24;10:132-136.
Type 1 neovascularisation
PART 3:
Retinal pigment
epithelial tears
Alexandra Mouallem-Bézière
Gass also describes the typical ophthalmoscopy results, in which an RPE tear appears as a single, clearly
defined area of absent RPE associated with a pigmented mound corresponding to the retracted RPE. The
tear can progress and become larger, with significant loss of vision.
In 1990, Coscas et al.3 identified the clinical characteristics that precede the appearance of a tear. Before
a tear appears, the PED was found to increase in size and change shape. The authors also described
the angiographic characteristics: a hyperfluorescent area appearing on the edges of the PED in the early
phase, spreading gradually over time (Figure 1). They concluded that laser treatment should be considered
carefully in the presence of these characteristics, given the high risk of a tear as a result of thermal
contraction of the neovascular lesion. In addition to laser photocoagulation, several treatments have been
reported as possible causes of RPE tears secondary to AMD, including photodynamic therapy4,5 and the
injection of various anti-VEGF drugs6,7. Anti-VEGF injections can cause contraction and fibrosis of the
neovascular lesion, associated with fibrovascular PED, which correlates directly with breaks in the RPE8.
Recently, Nagiel et al.9 used spectral-domain optical coherence tomography (SD-OCT) to identify
contractile neovascular tissue on the undersurface of the PED as a causal mechanism of RPE tears
following anti-VEGF injections for AMD.
In 2016, Mouallem et al.10 described a case series with double tears of the RPE. These double tears
on either side of the neovascularised PED can occur simultaneously or sequentially. The advanced
physiopathology is the development of choroidal neovascularisation in the centre of the PED, exerting
tangential pressure on both sides, which leads to the tear10,14.
A B C
Fig. 1: Multimodal imaging (autofluorescence photograph, fluorescein angiography and SD-OCT) of an 83-year-old treatment-
naive female patient who presented with an RPE tear in her right eye at the initiation of anti-VEGF treatment for exudative AMD.
(A): Autofluorescence photo of the posterior pole prior to the RPE tear.
(B): Fluorescein angiography image (intermediate phase, 3 minutes) showing hyperfluorescence along the edge of the PED
corresponding to a fragile area of the RPE.
(C): Autofluorescence photograph of the posterior pole revealing a large, grade-4, hypoautofluorescent tear in the geographic edge
of the RPE. The RPE appears hyperautofluorescent where it has rolled up at the edge of the tear.
(D): Pre-tear SD-OCT image revealing bridge-shaped choroidal neovascularisation under the entire RPE, which appears wrinkled
due to the horizontal pressure exerted by the neovascularisation.
(E): Post-tear SD-OCT image showing an increase in the size of the PED, with accordion-like retraction of the RPE at the edge of
the detachment.
Conclusion
Retinal pigment epithelium tears are complications of neovascularised PED.
The prognosis for these tears is directly linked to their severity grade, and their
appearance has no impact on the usual retreatment criteria for exudative AMD.
References
1. Hoskin A, Bird AC, Sehmi K. Tears of detached retinal pigment epithelium. Br J Ophthalmol 1981; 65:417–422.
2. Gass JD. Pathogenesis of tears of the retinal pigment epithelium. Br J Ophthamol 1984; 68:513-519
3. G. Coscas, F. Koenig, G. Soubrane. The pretear characteristics of pigment epithelial detachments. A study of 40 eyes. Arch Ophthalmol 1990;108:1687–1693
4. Gelisken F, Indhofen W, Partsch M, et al. Retinal pigment epithelial tear after photodynamic therapy for choroidal neovascularization. Am J Ophthalmol 2001;131:518–520.
5. Pece A, Introini U, Bottoni F, et al. Acute retinal pigment epithelial tear after photodynamic therapy. Retina 2001;21:661–665.
6. Dhalla MD, Blinder KJ, Tewari A, et al. Retinal epithelial pigment tear following intravitreal pegaptanib sodium. Am J Ophthalmol 2006;141:751–753.
7. Bakri SJ, Kitzmann AS. Retinal pigment epithelial tear after intravitreal ranibizumab. Am J Ophthalmol 2007;143:505–507.
8. Spaide RF. Enhanced depth imaging optical coherence tomography of retinal pigment epithelial detachment in age-related macular degeneration. Am J Ophthalmol 2009;
147:644–652.
9. Nagiel A, Freund KB, Spaide RF et al. Mechanism of retinal pigment epithelium tear formation following intravitreal anti-vascular endothelial growth factor therapy revealed
by spectral-domain optical coherence tomography. Am J Ophthalmol 2013; 156:981-988.
10. Mouallem A, Sarraf D, Chen X et al. Double retinal pigment epithelium tears in neo vascular age related macular degeneration. Retina 2016;36(11):2197-2204
11. Sarraf D, Reddy S, Chiang A et al. A new grading system for retinal pigment epithelial tears. Retina 2010;30:1039-45.
12. Gutfleisch M, Heimes B, Schumacher M et al. Long-term visual outcome of pigment epithelial tears in association with anti-VEGF therapy of pigment epithelial detachment
in AMD. Eye (Lond). 2011;25:1181-6.
13. Durkin SR, Farmer LD, Kulasekara S, Gilhotra J. Change in vision after retinal pigment epithelium tear following the use of anti-VEGF therapy for age-related macular
degeneration Graefes Arch Clin Exp Ophthalmol. 2015 Mar 8.
14. Mukai R, Sato T, Kishi S. Repair mechanism of retinal pigment epithelial tears in age-related macular degeneration. Retina. 2015;35:473-80.
Type 1 neovascularisation
PART 4:
Wrinkled pigment
epithelial detachment
Gérard Mimoun
Vascularised pigment epithelial detachment (PED) associated with AMD is still considered to be a
factor in poor visual prognosis, with a gradual or sometimes sudden reduction in visual acuity linked
to macular haemorrhage or a retinal pigment epithelium tear.
However, we have observed a unique form of vascularised PED, characterised by small, jagged
folds of the retinal pigment epithelium (RPE) in spectral-domain OCT that we refer to as “wrinkled
PED”, which could be a characteristic of good visual prognosis, with fewer anti-VEGF intravitreal
injections and a long relapse-free period.
1. Definition
Wrinkled pigment epithelial detachment is defined as a neovascularised pigment epithelial detachment
(nPED) with a height in SD-OCT of over 200 µm and generally less than 500 µm, with at least four small,
jagged folds of the retinal pigment epithelium (Figure 1).
Its appearance differs from that of unwrinkled neovascularised pigment epithelial detachment, where the
RPE is raised or bulging but stretched out and smooth.
2. Description
1) Fundus examination
The macula appears raised and slightly blurry, sometimes with several visible folds, giving it a scalloped
appearance. This is much more visible in en-face OCT imaging.
3) Fluorescein angiography
Wrinkled pigment epithelial detachment can be observed in the initial phase prior to treatment (Figure 2).
• In early-phase fluorescein angiography, progressive, inhomogeneous staining is visible, with in this
particular case (Figure 2) a hyperfluorescent lesion located nasal to the macula and pinpoints in the
temporal section of the macula.
• In late-phase fluorescein angiography, leakage from the lesion can be seen in the nasal section of the
macula, with an increase in pinpoints temporal to the macula.
In the early phase, a round, fairly extensive hypercyanescent lesion can be seen, with small folds already
visible. These hypercyanescent folds grow over time and by the late phase are even more visible,
associated with a plaque in the centre of the pigment epithelial detachment.
In Figure 3, the patient has PED complicated by choroidal neovascularisation in the context of exudative
age-related macular degeneration, but OCT is required in order to characterise it.
A slightly bulging, heterogeneous elevation of the retinal pigment epithelium can be seen. In the slice
focused on the section nasal to the lesion that was hyperfluorescent in FA, an irregular, hyperreflective
structure can be seen on the undersurface of the RPE. Above the RPE, the subretinal area is grey and
hyperreflective, with blurred edges (grey SRD), suggestive of exudative choroidal neovascularisation
(Figure 4).
OCTA shows a neovascular flow hypersignal within the pigment epithelial detachment. Moderate
arborisation is present, usually without any bordering arcades.
A B
A B
The number of anti-VEGF IVIs was much lower in the wrinkled PED group, to a statistically significant
extent (p = 0.0011), at 2.93 the first year, 1.93 the second year and 1.88 the third year. In the unwrinkled
PED group, the number of IVIs was much higher, at 5.87, 4.44 and 6.21 respectively.
The average maximum time without relapse for wrinkled PED was significantly longer, at 7.87 months
at 1 year, 13.5 months at 2 years and 14.8 months at 3 years. The times for the unwrinkled PED group
were shorter, at 4.59, 7.83 and 8.57 months respectively.
Wrinkled PED is associated with a statistically significant better visual prognosis at 3 years, a lower
number of IVIs and a long-term reduction in relapse frequency.
The presence of small, jagged folds in the RPE during treatment could therefore be a phenotypic
characteristic and a valuable predictive factor, making their identification important when treating patients
with vascularised PED associated with AMD.
The physiopathology of wrinkled PED remains unknown. However, we know that as the neovascularised
PED is formed and increases in size under the effects of subretinal pigment epithelial exudation linked to
occult choroidal neovascularisation, the RPE itself distends and increases in length.
There are several possible explanations: under the effect of IVIs, the hydrostatic pressure within the
PED decreases, weakening the PED, and the now-distended and lengthened RPE is forced to roll up,
since it has become too long for the surface it covers. This would explain the folds in the RPE and its
wrinkled appearance. Alternatively, it could be a specific characteristic of less active or less exudative
neovascularisation, or linked to the retinal pigment epithelium being less capable of distension in certain
patients.
Fig. 4: Initial OCT before treatment: slightly bulging, heterogeneous elevation of the retinal pigment epithelium with grey SRD above
the RPE.
Fig. 5: OCT one month after IVI: small, jagged folds appear in the RPE and the PED flattens out. The grey SRD has started to
disappear, as have the signs of exudation.
Fig. 6: OCT one month after six IVIs: small, jagged folds in the RPE, the PED has flattened out, no grey SRD or signs of exudation.
Fig. 7: En-face OCT before and after treatment. En-face OCT provides a clearer image of the
RPE folds.
Conclusion
Wrinkled PED is a progressive form of neovascularised PED that occurs during
treatment for AMD. It is characterised by small, jagged folds in the RPE and is
associated with a better visual prognosis, fewer IVIs and a long-term reduction in
relapse frequency—to a statistically significant extent.
The presence of small folds in the RPE could therefore be a valuable predictive
factor, and their identification in clinical practice important for patient treatment.
References
1. Mrejen S, Sarraf D, Mukkamala SK, Freund KB. Multimodal imaging of pigment epithelial detachment: a guide to evaluation. Retina Phila Pa.
2013;33(9):1735-1762. doi:10.1097/IAE.0b013e3182993f66.
2. Hoerster R, Muether PS, Sitnilska V, et al. Fibrovascular pigment epithelial detachment is a risk factor for long-term visual decay in neovascular
age-related macular degeneretion. Retina Phila Pa. 2014;34(9):1767-1773.
3. Poliner LS, Olk RJ, Burgess D, Gordon ME. Natural history of retinal pigment epithelial detachments in age-related macular degeneration.
Ophthalmology. 1986;93(5):543-551.
4. Rofagha S, Bhisitkul RB, Boyer DS, et al. SEVEN-UP Study Group. Seven-year outcomes in ranibizumab-treated patients in ANCHOR, MARINA,
and HORIZON: a multicenter cohort study (SEVEN-UP). Ophthalmology. 2013;120(11):2292-2299.
5. Solomon SD, Lindsley K, Vedula SS, et al. Anti-vascular endothelial growth factor for neovascular age-related macular degeneration. Cochrane
Database Syst Rev. 2014;(8):CD005139.
6. Chan CK, Abraham P, Meyer CH, et al. Optical coherence tomography-measured pigment epithelial detachment height as a predictor for
retinal pigment epithelial tears associated with intravitreal bevacizumab injections. Retina Phila Pa. 2010; 30(2):203-211. doi:10.1097/
IAE.0b013e3181babda5.
7. Lam D, Semoun O, Blanco-Garavito R, et al. WRINKLED VASCULARIZED RETINAL PIGMENT EPITHELIUM DETACHMENT PROGNOSIS AFTER
INTRAVITREAL ANTI-VASCULAR ENDOTHELIAL GROWTH FACTOR THERAPY. Retina. 2018. Jun;38(6):1100-1109.
Type 1 neovascularisation
PART 5:
Unique pre-epithelial type 2
neovascularisation healing
style following anti-VEGF
treatment: “igloos”
Gérard Mimoun
Choroidal neovascularisation (CNV) is very often responsible for a severe reduction in visual acuity,
due to bleeding, exudation or fibrous scarring. CNV was initially categorised using fluorescein
angiography, into classic neovascularisation (well defined) and occult neovascularisation (poorly
defined). With the introduction of spectral-domain OCT (SD-OCT), it began to be categorised into pre-
epithelial type 2 neovascularisation, located above the RPE, and “occult” type 1 neovascularisation,
located below the RPE.
In high myopia and angioid streaks, pre-epithelial type 2 CNV is the most common form of CNV, but
it is rare in AMD, occurring in just 9 to 17% of cases. Subepithelial type 1 or occult CNV is the most
common form found in AMD.
Thanks to anti-VEGF therapy—particularly when administered in the early stages of the disease—
visual acuity after IVI in high myopia, angioid streaks and certain cases of AMD is currently well
controlled and remains acceptable.
However, we have observed a specific progressive sign of healing arising after pre-epithelial type 2
CNV is treated with anti-VEGF IVI: a dense, relatively homogeneous, localised subretinal elevation
with a characteristic hyperreflective dome appearance in SD-OCT, which we refer to as an “igloo”
(Figure 1).
When pre-epithelial type 2 CNV heals in this way, the prognosis is generally favourable, with fairly
good visual acuity, fewer intravitreal injections and a long period without relapse.
1. Description
The igloo is a way in which pre-epithelial type 2 CNV heals after IVI treatment. In the initial pre-treatment
stage, the signs are therefore the same as with visible type 2 CNV.
1) Fundus examination
The macula is raised and retinal haemorrhage may be visible. The CNV itself can sometimes be detected.
2) Fluorescein angiography
In early-phase fluorescein angiography, a vascular network with a feathered appearance is clearly visible,
exhibiting early hyperfluorescence, with a highly hyperfluorescent wreath and a slightly darker centre.
In the late phase, there is extensive leakage from the lesion: the hyperfluorescence increases in intensity
and size, and its edges become blurred (Figure 2).
A B
C D
Fig. 1: SD-OCT of a 79-year-old woman with exudative AMD, before treatment and three months after IVI.
OCT-angiography reveals a pathological neovascular flow hypersignal, often with a characteristic image
of perfused CNV with a clearly visible surrounding arcade, a dark halo of vascular steal and extensive
arborisation with numerous fine connected networks.
A B
D E
Fig. 2: Angioid streaks complicated by CNV before and four months after IVI.
• Before treatment:
A: Fluorescein angiography (FA) showing visible CNV with leakage.
B: OCT: Pre-epithelial (type 2) CNV with grey SRD and SRD.
C: ICG image showing streaks becoming stained in the late phases and subfoveal CNV.
Locating the igloo against the outline of the RPE is not easy, since the OCT tone along the edges of
the igloo is the same as the RPE, making it hard to precisely follow the line of the RPE. The igloo often
appears as a smooth hyperreflective subretinal elevation above the RPE that seems to correspond to
CNV scarring above the RPE. More rarely, the igloo looks like a PED and is therefore located below
the RPE, indicative of CNV scarring below the RPE that can only be explained by the RPE expanding
over and enveloping the CNV.
Conclusion
The igloo is a specific progressive sign of healing following anti-VEGF IVI treatment
for pre-epithelial type 2 CNV, appearing in OCT as a dense, fairly homogeneous,
hyperreflective, dome-shaped, localised subretinal elevation. It is found in particular
in cases of high myopia and angioid streaks—occurring rarely in AMD—and is
associated with a good visual prognosis.
References
Dolz-Marco R, Phasukkijwatana N, Sarraf D et al. Regression of Type 2 Neovascularization into a Type 1 Pattern after Intravitreal Anti-Vascular
Endothelial Growth Factor Therapy for Neovascular Age-Related Macular Degeneration. Retina 2016.
Hoang Mai Le, Gérard Mimoun, Salomon Y. Cohen et al. Regression of Type 2 neovascularization in Age-related Macular Degeneration, Myopia and
Angioid Streaks: the “Igloo” pattern. (sous presse).
Type 1 neovascularisation
PART 6:
Polypoidal choroidal
vasculopathy
(aneurysmal type 1
choroidal neovascularisation)
Mayer Srour, David Sayag
Polypoidal choroidal vasculopathy (PCV) was first described by Yannuzzi in 19821. The condition
involves abnormal branching vascularisation within the choroid, associated with aneurysmal
vascular dilations that can lead to serous detachment of the retinal pigment epithelium (RPE) and, in
some cases, extensive haemorrhage. PCV is more prevalent in Asian populations than in Caucasian
populations. Age at diagnosis is lower than in AMD.
PCV is usually idiopathic, but can also be secondary to exudative AMD, myopic staphyloma, CSCR/
DRPE (central serous chorioretinopathy/diffuse retinal pigment epitheliopathy) or a naevus2. More
recently, Freund placed PCV on the pachychoroid spectrum (Figure 1) and renamed it “aneurysmal type
1 neovascularisation”3.
Clinic Localised RPEC Inter-PM RPEC Serous PED/ Type 1 CNV Polyps
+ SRD bleb
IR spaces +/-
SRD
Fluo A Non-specific Non-specific Serous PED/ V-PED Polyps
Leakage point
AutoFluo Localised RPEC Localised RPEC +/- Gravitational Non-specific Non-specific
tracks
ICG Hyperpermeability CNV plaque Polyps
EDI-OCT Pachychoroid with pachyvessels
OCTA No flow Type 1 CNV BVN +/- polyps
1. Diagnosis
1) Fundus
In the fundus, PCV may be suspected in the presence of PED with lipid exudates, often associated with
subretinal haemorrhage. One or more round, red-orange lesions may also be observed under the RPE,
varying in size and prominence (Figure 2).
Drusen, RPE changes and areas of geographic atrophy are absent or rare. The location of the PCV can
be subfoveal, juxtafoveal, extrafoveal, peripapillary or even peripheral, and varies from population to
population. In Caucasians, the lesions are often extrafoveal, with high levels of peripapillary involvement
(23 to 68%). In Asian populations, polyp location is variable, with the central subfoveal area affected
in almost 30% of certain Chinese populations. In rare cases, polypoidal dilations may be found on the
retinal periphery and can be classified as exudative haemorrhagic chorioretinopathy. Patients are often
asymptomatic and lesions may be discovered by chance. The condition often improves spontaneously. If
extensive haemorrhage is present, a differential diagnosis may be required to rule out a tumour.
Pigment epithelial detachment (PED) is present (green circle), often associated with
exudates (yellow arrow) and subretinal haemorrhage (red arrow).
Fluorescein angiography has low specificity for PCV. It shows progressive staining of aneurysmal
dilations, with little diffusion during the late phase and, very often, associated hyperfluorescence of the
branching vascular network (BVN), which is difficult to differentiate from polyps (Figure 3). Polypoid
lesions are often masked by bleeding and merged into the PED hyperfluorescence.
ICG is the key examination for diagnosing PCV. Polyps are displayed as round, hyperfluorescent lesions
in the early phases and can remain into the late phase or undergo “wash-out” (Figure 4). The BVN is
visible in the early phases, with polyps located within the network or on its edge, resulting in a late
“hyperfluorescent lesion”. Choroidal hyperpermeability is also evident when using ICG. ICG is essential
when monitoring PCV, since it can show whether the polyp hyperfluorescence has lessened or
disappeared following treatment. Pulsatile filling of polyps can sometimes be seen and is thought to be a
risk factor for rupture and haemorrhage.
A B
Progressive hyperfluorescence of aneurysmal dilations, with little diffusion during the late phase and a hyperfluorescent branching
vascular network that is difficult to differentiate from the polyps.
A B
The polyps appear as round, hyperfluorescent lesions (red arrows) in the early phases and may remain into the late phases or
undergo “wash-out” (red circle). The polyps are accompanied by the BVN (green arrow), which is visible in the early phases with
polyps located within the network or on its edge, resulting in a late hyperfluorescent lesion.
3) SD-OCT
A diagnosis of PCV can be strongly suspected based on OCT findings. Polyps appear as dome-shaped
elevations in the RPE containing one or more hyperreflective rings associated with a hyporeflective
centre. The BVN causes flat, irregular elevation of the RPE (“double layer sign”), visible as two highly
reflective layers (the RPE and a layer beneath the RPE). If exudation is present, serous retinal detachment,
intraretinal cystoid spaces or retinal thickening may be observed. Swept-source imaging or EDI usually
reveals a pachychoroid (> 300 μm) associated with “pachyvessels”2 (Figure 5).
The polyps look like an ogival or “steep slope” PED, sometimes adjoining a serous PED. The double layer sign is also found, along
with a thick choroid.
4) OCTA
OCTA is at present of little use when it comes to imaging polyps. However, the BVN, which is characterised
by linear blood flow, is easily detected by OCTA. Polypoid lesions appear in most cases as hypodense
round structures with no flow or hyperdense round structures with flow (between 17 and 45% of cases,
depending on the series4), occasionally encircled by a hypodense halo. The lack of signal in the polyp
does not mean that there is no blood flowing, rather that the characteristics of the blood flow do not meet
the OCTA detection criteria (Figures 6 and 7). This failure to detect the blood flow may be caused by
signal attenuation linked to the RPE, or by the flow characteristics being undetectable: the flow may be
too weak or circulating only at the edges of the polyp, or there may be hyalinisation of the polyp, blocking
out the light. As the technology currently stands, therefore, OCTA cannot be used in place of ICG to detect
and assess PCV4.
OCTA of the choriocapillaris shows the abnormal choroidal network in the form of a hyperdense
lesion with flow (green arrow). The polypoid lesion appears as a hypodense, round structure
with no flow (red arrow) and the neovascularisation as a hyperdense structure with flow
(yellow arrow).
2. PCV classification
Certain authors have suggested categorising PCV into different subtypes5-6 based on how it appears in
ICG and OCT. Type 1 PCV, or “polypoidal neovascularisation”, is distinguishable by its drainage network
(the BVN) and feeder vessel, a thinner choroid with large lesions, and a positive response to anti-VEGF
(Figure 7). Type 2 PCV, or “idiopathic PCV”, features no BVN, or only a very faint one, a small lesion and a
thicker choroid (Figure 8). Recently, an American team proposed a new term for PCV: “aneurysmal type 1
choroidal neovascularisation”5. However, there is no consensus around these subcategories and terms.
Conclusion
ICG remains the gold standard for diagnosing PCV and can be useful for monitoring
it. OCT can provide highly characteristic images of PCV, while the current resolution
of OCTA is insufficient for it to replace ICG in PCV. Finally, a diagnosis of AMD must
be reassessed by ICG if the therapeutic response is insufficient, as PCV may be
present. Angiography for PCV can also involve OCT B-scan, en-face OCT and OCTA,
since they can distinguish between the BVN and polypoid lesions and display the
neovascular activity, revealing the extent of intraretinal and subretinal exudation.
References
1. Yannuzzi LA. Idiopathic polypoidal choroidal vasculopathy. Macula Society Meeting 1982; Miami, Florida, USA.
2. Gallego-Pinazo R, Dolz-Marco R, Gómez-Ulla F et al. Pachychoroid Diseases of the Macula. Med Hypothesis Discov Innov Ophthalmol. 2014;3(4):111-5.
3. Dansingani KK, Gal-Or O, Sadda SR et al. Understanding aneurysmal Type 1 neovascularization (polypoidal choroidal vasculopathy): a lesson in the
taxonomy of ‘expanded spectra’ - a review. Clin Exp Ophthalmol. 2018;46(2):189-200.
4. Srour M, Querques G, Semoun O, et al. Optical coherence tomography angiography characteristics of polypoidal choroidal vasculopathy. Br J Ophthalmol.
2016 ;100(11):1489-1493
5. Kawamura A, Yuzawa M, Mori R et al. Indocyanine green angiographic and optical coherence tomographic findings support classification of polypoidal
choroidal vasculopathy into two types. Acta Ophthalmol. 2013;91:e474–e481.
6. Tanaka K, Nakayama T, Mori R et al. Associations of complement factor H (CFH) and age-related maculopathy susceptibility 2 (ARMS2) genotypes with
subtypes of polypoidal choroidal vasculopathy. Invest Ophthalmol Vis Sci. 2011;52:7441–7444.
Type 1 neovascularisation
PART 7:
Fibrous choroidal
neovascularisation
Alexandra Miere, Manar Addou-Regnard, Eric Souied
1. Definition
Subretinal fibrosis is the result of complex tissue repair mechanisms, secondary to either the natural
healing process1 or treatment with vascular endothelial growth factor inhibitors (anti-VEGF therapy)2.
Recent studies on animal models3-6 have found that the key factors in the development of fibrosis are
connective tissue growth factor (CTGF), platelet-activating factor (PAF), the platelet-activating factor
receptor (PAF-R) and macrophage-rich peritoneal exudate cells (PEC). In addition, the extensive use of
anti-VEGF therapy to treat exudative age-related macular degeneration (AMD) has been found to carry a
risk of the treated choroidal neovascularisation developing into macular atrophy and/or subretinal fibrosis,
both of which are associated with poor visual outcomes7,8.
2. Diagnosis
During dilated fundus examination, subretinal fibrosis can be identified as a clearly defined accumulation
of raised, yellowish tissue, often with concave edges (Figure 1). In fluorescein angiography (FA), fibrosis is
characterised by impregnation of the lesion in the late phases, with no leakage of the dye7,8 (Figure 2). In
SD-OCT, fibrosis appears as a compact, hyperreflective subretinal lesion, possibly accompanied by loss
or fragmentation of the adjacent retinal pigment epithelium (RPE) and ellipsoid zone (Figure 1).
OCT-angiography (OCTA) is an imaging technique that offers excellent resolution at defined depths9,10,
enabling detailed analysis of the fibrotic scar. OCTA images of subretinal fibrosis almost always reveal
a perfused vascular network, along with collateral architectural changes in the outer retina and the
choriocapillaris (Figures 2 and 3). However, this examination can be impossible to interpret in some
cases, due to very poor visual acuity or an inability to hold the gaze and the need for a high level of patient
cooperation.
The neovascular network inside the fibrotic scar can display several different patterns, the most common
of which is the “pruned tree” or “dead tree” (Figures 2 and 3), defined by an absence of visible fine
capillaries in the slab corresponding to the fibrosis (outer retina and/or choriocapillaris). The accumulation
of fibrous tissue may cause a blocking effect, resulting in a flow/signal void (Figure 3). In OCTA, subretinal
fibrosis associated with neovascular AMD can be identified as distinct abnormal vascular networks within
the fibrotic scar, which were previously undetectable by FA or SD-OCT alone11.
The development of a qualitative classification for these networks would prove particularly valuable from
a clinical and physiopathological perspective.
A B
Fig. 1: Non-invasive multimodal imaging of the right eye of a male patient with a history of exudative AMD.
A: The Multicolor® image shows a clearly defined, yellowish lesion with concave edges.
B: In the infrared image, the lesion is well defined and hyperreflective.
C: The SD-OCT slice through the centre of the lesion shows a compact, uniform hyperreflective lesion. The retinal pigment
epithelium is barely visible.
A B
B’
C D
Fig. 2: Invasive and non-invasive multimodal imaging of the left eye of a female patient with a history of exudative AMD.
A and B: Fluorescein angiography in the early (A) and late (B) phases shows a hyperfluorescent subfoveal lesion, with no leakage
of the dye in the late phases.
B’: The OCTA flow image and the corresponding B-scan reveal, in the central 10° (3 x 3 mm acquisition window), the presence of
a high-flow vascular network in the outer retina slab. Note that there are no fine capillaries, suggesting a “pruned tree” or “dead
tree” pattern.
C and D: Indocyanine green angiography reveals the occult component of the fibrotic neovascular membrane.
A C
B D
Fig. 3: Different OCTA images (6 x 6 mm) of high-flow networks inside fibrotic lesions.
A and B: Flow image (A) and corresponding B-scan (B) of a male patient with subretinal fibrosis secondary to AMD. In the flow
image (A), a high-flow network can be seen in the choriocapillaris slab (white arrow), along with a flow/signal void caused by the
fibrous tissue (dotted line).
C and D: Flow image (C) and corresponding B-scan (D) of a different male patient with subretinal fibrosis secondary to AMD. In the
flow image (C), a high-flow “pruned tree” network can be seen in the slab including the fibrotic lesion, with the main vessels still
visible (white arrow). Note that at either end of the lesion, there is nevertheless some capillary sprouting visible (asterisk).
References
1. Kumar V, Abbans K, Nelson F. Robbins and Cotran pathologic basis of disease. 9th ed. Philadelphia: Elsevier Saunders; 2014.
2. Hwang JC, Del Priore LV, Freund KB et al. Development of subretinal fibrosis after anti-VEGF treatment in neovascular age-related macular
degeneration. Ophthalmic Surg Lasers Imaging 2011;42:6-11.
3. Moussad EE, Brigstock DR. Connective tissue growth factor. What’s in a name? Mol. Genet. Metab. 2000; 71, 276–292.
4. Zhang H, Yang Y, Takeda A et al. ANovel Platelet-Activating Factor Receptor Antagonist Inhibits ChoroidalNeovascularization and Subretinal
Fibrosis. PLoS One. 2013 27;8(6):e68173.
5. Jo YJ, Sonoda KH, Oshima Y et al. Establishment of a new animal model of focalsubretinal fibrosis that resembles disciform lesion in advanced
age-related macular degeneration. Invest Ophthalmol Vis Sci 2011;52:6089-6095.
6. Cui W, Zhang H, Liu ZL. Interleukin-6 receptor blockade suppresses subretinal fibrosis in a mouse model. Int J Ophthalmol2014;7:194-197.
7. Bloch SB, Lund-Andersen H, Sander B, Larsen M. Subfoveal fibrosis in eyes withneovascular age-related macular degeneration treated with
intravitrealranibizumab. Am J Ophthalmol 2013;156:116-124.
8. Channa R, Sophie R, Bagheri S et al. Regression of choroidal neovascularization results in macular atrophy in anti-vascular endothelial growth
factor-treated eyes. Am J Ophthalmol 2015;159:9-19.
9. Jia Y1, Tan O, Tokayer J et al. Split-spectrum amplitude-decorrelation angiography with optical coherence tomography. Opt Express 2012; 20:
4710-4725.
10. Jia Y1, Bailey ST, Wilson DJ et al. Quantitative optical coherence tomography angiography of choroidal neovascularization in age-related macular
degeneration. Ophthalmology 2014;121:1435-1444.
11. Miere A, Semoun O, Cohen SY et al. Optical coherence tomography angiography features of subretinal fibrosis in age-related macular
degeneration. Retina. 2015; 35(11): 2275-84.
Type 2 neovascularization
secondary to AMD
Al’a El Ameen
“Pre-epithelial” or “visible” type 2 choroidal neovascularisation (CNV) is the least common form of
neovascularisation found in exudative age-related macular degeneration (AMD) (17.6% of cases)1.
However, the visible characteristics of this form were the first to be described in the literature, since
it can be seen directly in fluorescein angiography (FA). Like type 1 CNV, it is choroidal in origin,
but expands below the retina after breaking through Bruch’s membrane and the retinal pigment
epithelium (RPE).
During fundus examination, it appears as a slightly raised yellowish macular lesion, sometimes
accompanied by bleeding around the edges (Figure 1).
Studies of infrared photographs (820 nm) of type 2 CNV have revealed the presence of a whitish ring
encircling a dark central core2. The edges of this ring correspond to the leakage zone in late-phase FA and
the black halo encircling the neovascular membrane in the early phase (Figures 2 and 4).
Fluorescein angiography has long been the imaging gold standard for diagnosing type 2 CNV. It appears
as a well defined area of hyperfluorescence corresponding to the neovascular membrane in the early
phases, surrounded by a hypofluorescent area. The late phases are marked by gradual leakage of the dye
from this lesion (Figures 2, 3, 4 and 5).
In indocyanine green angiography, type 2 CNV is often visible in the early phases as a hypercyanescent
neovascular network that generally tends to disappear in the later phases (wash-out) (Figures 2, 3 and
4), but occasionally an area of hypercyanescence remains.
OCT shows both the neovascularisation and its exudative activity. In SD-OCT (spectral-domain OCT),
type 2 neovascularisation is characterised by a pre-epithelial hyperreflective fusiform lesion accompanied
by underlying hyporeflectivity (due to a shadow effect). Within this inhomogeneous hyperreflectivity, there
is an area of abnormality in the appearance of the RPE, which can vary in size. If the neovascularisation
is recent, the hyperreflectivity is not very dense or homogeneous, while older fibrovascular lesions result
in more marked hyperreflectivity. As the neovascularisation progresses, slightly hyperreflective “grey”
exudation gradually appears3, resulting in an accumulation of hyporeflective subretinal fluid (serous
retinal detachment [SRD]) at the edges or on top of the neovascularisation, along with intraretinal spaces
(Figures 1 and 3).
OCT-angiography (OCTA) provides direct images of the neovascular network. The various types of
neovascularisation that occur secondary to AMD are classified based on the slab in which the neovessels
are visible and the presence or absence of flow in the corresponding B-scan. Type 2 neovascularisation is
primarily visible in the outer retina slab, which is normally avascular4 (Figures 2, 3 and 5). In certain cases,
a feeder vessel can be identified. Numerous studies have attempted to assess the sensitivity of OCTA in
detecting neovascularisation, with results ranging from 85 to 100%5-7.
The SD-OCT slice through this lesion (green line) shows that it is hyperreflective, pre-epithelial and accompanied by SRD and
exudative intraretinal spaces.
A C E
B D F
A: The lesion from Figure 1 is hyperfluorescent in the early phase of fluorescein angiography and surrounded by a hypofluorescent
ring.
B: The hyperfluorescence increases in intensity into the late phases of FA, with mild diffusion.
C: The early phase of ICGA reveals a hypercyanescent neovascular network.
D: The cyanescence increases slightly in the late phases, with no marked diffusion.
E: OCTA (Optovue®) clearly shows the type 2 CNV in the outer retina.
F: In infrared imaging, the neovascularisation appears as a whitish, hyperreflective ring (white arrow) encircling a dark central core.
A B E
C D
C D
A: Infrared photograph: the neovascularisation appears as a whitish, hyperreflective ring (white arrow) with a dark central core.
B: Late-phase fluorescein angiography: gradual leakage of the dye.
C: Early-phase ICGA: well defined area of hypercyanescence corresponding to the neovascularisation.
D: Late-phase ICGA: the cyanescence tends to decrease over time, due to wash-out.
A B
C D
2. Semoun O, Guigui B, Tick S et al. Infrared features of classic choroidal neovascularisation in exudative age-related macular degeneration. Br J
Ophthalmol. févr 2009;93(2):182-5.
3. Ores R, Puche N, Querques G et al. Gray hyper-reflective subretinal exudative lesions in exudative age-related macular degeneration. Am J
Ophthalmol. août 2014;158(2):354‑61.
4. El Ameen A, Cohen SY, Semoun O et al. Type 2 neovascularization secondary to age-related macular degeneration imaged by optical coherence
tomography angiography: retina. 2015;35(11):2212‑8.
5. Inoue M, Jung JJ, Balaratnasingam C et al. A Comparison Between Optical Coherence Tomography Angiography and Fluorescein Angiography for
the Imaging of Type 1 Neovascularization. Invest Ophthalmol Vis Sci.2016;57(9):314-323.
6. Faridi A, Jia Y, Gao SS et al. Sensitivity and Specificity of OCT Angiography to Detect Choroidal Neovascularization. Ophthalmol Retina.
2017;1(4):294-303.
7. Souedan V, Souied EH, Caillaux V et al. Sensitivity and specificity of optical coherence tomography angiography (OCT-A) for detection of choroidal
neovascularization in real-life practice and varying retinal expertise level. Int Ophthalmol. 2018;38(3):1051-60.
Type 3
neovascularization
Alexandra Miere, Giuseppe Querques, Roxane Bunod, Eric H. Souied
Type 3 neovascularisation is a clinical form of neovascular AMD that preferentially affects the
neurosensory retina and causes a compensatory telangiectasic neovascular response, associated
with intraretinal proliferation1.
1. History
In 1992, Hartnett described an angiomatous retinal lesion associated with drusen, which caused pigment
epithelial detachment2. The origins of this form of neovascular AMD have been hotly contested over
the years. Several different terms have been used to refer to this kind of lesion, including “chorioretinal
anastomosis” (CRA) and “retinal angiomatous proliferation” (RAP).
In 2001, Yannuzzi described RAP as an angiomatous proliferation of retinal origin extending into
the subretinal space. Several hypotheses were put forward to explain the formation of these lesions,
which originate in the deep capillary plexus, including communication with (pre)existing choroidal
neovascularisation3,4.
Two years later, Gass described CRA as an intraretinal extension of type 1 neovascularisation5.
It was Freund who eventually introduced the term “type 3 neovascularisation”, as a logical extension of
Gass’ categorisation5, to bring together the various different potential origins of this vascular complex.
According to Freund, type 3 neovascularisation can cover all of the following6:
• Focal neovascular proliferation originating in the deep capillary plexus (essentially RAP)1,6,7
• Intraretinal neovascular extension of a type 1 choroidal neovascular membrane (occult choroidal
neovascularisation) (essentially CRA)1,5-7
• De novo ruptures of Bruch’s membrane, with neovascular infiltration of the retina1,6,7.
Type 3 lesion diagnosis and treatment response are determined using various imaging techniques:
fluorescein angiography (FA), indocyanine green angiography (ICGA) and spectral-domain optical
coherence tomography (SD-OCT)3.
In ICGA, type 3 lesions are visible in the early phases in the form of a hyperfluorescent lesion, resulting in
a “hot spot” in the late phases3,8,9 (Figures 1 and 2).
SD-OCT plays a key role in type 3 neovascularisation. Typically, type 3 neovascularisation involves
thinning of the choroid. The various stages in its development were recently categorised by Su et al.10 as
follows:
• Stage 1: intraretinal hyperreflective foci (HRF) associated with cystoid macular oedema, but no outer
retinal disruption
• Stage 2: progression towards outer retinal disruption, with or without disruption of the retinal pigment
epithelium
• Stage 3: posterior progression of the type 3 lesion, resulting in proliferation through the retinal pigment
epithelium and serous pigment epithelial detachment (Figure 2).
A B C D
C
A’
Fig. 1: Multimodal imaging of the left eye of an 83-year-old woman with type 3 neovascularisation.
A and B: Indocyanine green angiography reveals a hypercyanescent lesion in the early phases (A, white arrow on enlarged image
A’) and late phases (“hot spot”) (B, white arrow).
C and D: Early-phase (C, white arrow on enlarged image C’) and late-phase (D) fluorescein angiography shows a hyperfluorescent
lesion below the avascular foveal zone, causing significant leakage.
E: Optical coherence tomography shows a funnel-shaped, hyperreflective intraretinal lesion accompanied by cystoid spaces.
H’
C D
H
I’
E F I
Fig. 2: Multimodal imaging of the right eye of a 78-year-old woman with type 3 neovascularisation.
A and B: The B-scan (B), guided by the infrared photograph (A), through the lesion below the avascular foveal zone reveals a
hyporeflective pigment epithelial detachment, above which is a hyperreflective intraretinal vascular complex.
C and D: Indocyanine green angiography (C: early phase, D: late phase).
E and F: Fluorescein angiography (E: early phase, F: late phase) showing a hyperfluorescent lesion on the edge of the avascular
foveal zone, causing a typical “hot spot” (D) in the late-phase ICGA images.
G and G’: OCTA images: B-scan with flow superimposed (G’) and en-face image of the deep capillary plexus (G) in 3 x 3 mm,
revealing flow in the hyperreflective complex.
H and H’: B-scan with flow superimposed (H’) and en-face image of the outer retina (H) in 3 x 3 mm, revealing a tuft-shaped lesion.
I and I’: B-scan with flow superimposed (I’) and en-face image of the choriocapillaris (I) in 3 x 3 mm, showing no neovascular lesions
in this area.
3. Benefits of OCT-angiography
OCT-angiography (OCTA) has been used for several years to precisely analyse the retinal and choroidal
microcirculation.
Acquisition windows of 3 x 3 mm and correct segmentation of the deep capillary plexus, outer retinal
layers (outer retina) and choriocapillaris provide valuable information on the physiopathology and
progression of type 3 lesions.
In OCTA, type 3 neovascularisation is defined by retino-retinal anastomosis (Figure 3). From the deep
capillary plexus (origin), high-flow vessels extend into the outer retina, forming a high-flow, tuft-shaped
lesion (hatched) (Figures 2 and 3). This lesion may gradually extend further, into the area below the
retinal pigment epithelium and/or the choriocapillaris: a small glomerular lesion (“clew-like” lesion) can be
seen in the OCTA image11.
OCTA provides additional evidence to confirm that, in the majority of cases, type 3 neovascularisation is
characterised by an intraretinal vascular complex originating in the deep capillary plexus (Figures 2 and
3)11,12.
4. Progression
Two distinct lesion progression profiles have been observed when monitoring patients undergoing anti-
VEGF therapy: either progression to atrophy (characterised by the lesion disappearing from the outer
retina slab) or progression into the space below the retinal pigment epithelium (characterised by the
persistence or de novo appearance of the small glomerular lesion in the choriocapillaris layer)13.
It is therefore essential to take good quality images at the first examination and during follow-up in order
to determine the progression profile.
5. Nascent type 3
One of the early signs of type 3 neovascularisation described recently by Sacconi et al.12 is hyperreflective
intraretinal foci located above a drusenoid PED, which can be easily confused with pigment migration.
OCTA evidences early flow in these foci, before the exudative type 3 neovascular lesion appears.
A new entity, “nascent type 3 neovascularisation”, has therefore been created to describe this progression
from a preclinical stage (hyperreflective focus and no exudation) to a clinical stage (progression of
the lesion from the deep capillary plexus into the retinal pigment epithelium and the space below it,
accompanied by exudation)12 (Figure 4).
A B C
A’ B’ C'
A: En-face image of the deep capillary plexus slab, showing a high-flow vessel and a lesion at the intersection of the two navigation
lines.
B: The outer retina slab reveals a high-flow, tuft-shaped lesion (white arrow).
C: Note that in the en-face image of the choriocapillaris slab, the lesion is not visible.
A’, B’ and C’: B-scan with flow superimposed for each automatically segmented slab, showing the flow in this small, hyperreflective
intraretinal complex. The flow is confined to the intraretinal zone (C’, white arrowheads).
April 2016
July 2016
November 2016
March 2017
In April 2016, note the presence of a hyperreflective intraretinal lesion, located above a drusenoid PED, which could easily be
confused with pigment migration. The OCTA B-scan shows that there is flow within it (arrow). Over time, the lesion gets closer to
the PED, eventually taking the typical form of type 3 neovascularisation. This is therefore a case of nascent type 3 (images by Prof.
Giuseppe Querques).
References
1. Freund KB, Ho IV, Barbazetto IA, et al. Type 3 neovascularization: the expanded spectrum of retinal angiomatous proliferation. Retina
2008;28:201–211.
2. Hartnett ME, Weiter JJ, Garsd A, Jalkh AE. Classification of retinal pigment epithelial detachments associated with drusen. Graefes Arch Clin Exp
Ophthalmol. 1992;230:11e19.
3. Querques G, Souied EH, Freund KB. How has high-resolution multimodal imaging refined our understanding of the vasogenic process in Type 3
neovascularization? Retina 2015;35:603-13.
4. Yannuzzi LA, Negrao S, Iida T, et al. Retinal angiomatous proliferation in age-related macular degeneration. Retina 2001;21:416–434.
5. Gass JD. Biomicroscopic and histopathologic considerations regarding the feasibility of surgical excision of subfoveal neovascular membranes.
Am J Ophthalmol 1994;118:258–298.
6. Freund KB, Ho IV, Barbazetto IA, et al. Type 3 neovascularization: the expanded spectrum of retinal angiomatous proliferation. Retina.
2008;28:201–211.
7. Yannuzzi LA, Freund KB, Takahashi BS. Review of retinal angiomatous proliferation or Type 3 neovascularization. Retina 2008;28:375–384.
8. Kuhn D, Meunier I, Soubrane G, Coscas G. Imaging of chorioretinal anastomoses in vascularized retinal pigment epithelium detachments. Arch
Ophthalmol 1995;113:1392–1398.
9. Jackson TL, Danis RP, Goldbaum M, et al. Retinal vascular abnormalities in neo-vascular age-related macular degeneration. Retina 2014;34:568–
575.
10. Su D, Lin S, Phasukkijwatana N, et al. An updated staging system of type 3 neovascularization using spectral domain optical coherence
tomography. Retina. 2016; 36 suppl 1: s40-s49.
11. Miere A, Querques G, Semoun O, et al. Optical coherence tomography angiography in early type 3 neovascu-larization. Retina. 2015 nov; 35(11):
2236-41.
12. Sacconi R, Sarraf D, Garrity S, et al. Nascent Type 3 Neovascularization in Age-Related Macular Degeneration. Ophthalmology Retina. (2018).
10.1016/j.oret. 2018.04.016.13. Miere A, Querques G, Semoun O, et al. Optical coherence tomography angiography changes in early type 3
neovascularization after anti-vascular endothelial growth factor treatment. Retina. 2017;37(10):1873-1879.
Adult-onset
foveomacular vitelliform
dystrophy (AFVD)
Jean-Louis Bacquet, Agnès Glacet-Bernard
1. Introduction
Adult-onset foveomacular vitelliform dystrophy (AFVD) is an eye disease found in middle aged to elderly
patients. It is fairly common and there is disagreement as to the best way to diagnose and manage it
(additional examinations, risk of complications, treatment).
It is commonly confused with other diseases, and certain aspects of its diagnosis and treatment are
the same as for age-related macular degeneration. However, its natural history, clinical and paraclinical
presentation and, in some cases, its genetic origin mark it out from AMD.
AFVD is defined by a deposit of vitelliform (“resembling an egg yolk”) material in the macula of a patient
aged over 50 years.
a) Historic definition
The first description of AFVD is in a 1974 work by Donald Gass, in which he provides a clinicopathologic
description of a “peculiar” macular dystrophy in nine patients. He links these clinical observations to the
angiographic behaviour of the lesion and concludes that the macular dystrophy is presumably genetic in
origin and autosomal dominant.
b) Contemporary definition
The precise, current definition of AFVD is summarised by Chowers et al.:
- It is a clinically and genetically heterogeneous disorder
- The term "dystrophy” is usually reserved for monogenic diseases and therefore should only be applied
to genetic cases of AFVD, which are fairly rare
- The link with AMD and pattern dystrophy is not clear
- A similar-looking submacular deposit of foveal material can be observed in other pathological situations
(toxicity, immune disease, infectious disease, injury etc.).
Today, AFVD is therefore considered to be a spectrum rather than a specific retinal disease.
The variety of different terms used to refer to AFVD contributes to the confusion:
- Adult vitelliform macular degeneration (according to Glacet-Bernard et al., Epstein et al. and Greaves
et al.)
- Pseudovitelliform macular degeneration (according to Sabates et al.)
- Adult-onset foveomacular pigment epithelial dystrophy (according to Vine et al.)
- Adult pseudovitelliform dystrophy (according to Burgess et al.)
- Adult vitelliform macular dystrophy (according to Bird et al.)
In 1997, Marmor et al. suggested that AFVD falls within the pattern dystrophy group, and it remains as
such in several reference works (e.g. Retinal Atlas, Yannuzzi et al.) (Figure 1).
Numerous authors have focused on genetic AFVD, resulting in information on the mechanisms by which
the material accumulates at molecular level. As with AMD, genetic research has given us a clearer idea of
the cellular and molecular natural history of AFVD.
- AFVD linked to monoallelic BEST1 mutation appears to be due to impaired phagocytosis of the
photoreceptor outer segments.
- AFVD linked to monoallelic PRPH2 mutation seems to be caused by structural aberrations of the
photoreceptor outer segments, making it harder and slower for them to be internalised and broken
down.
- More rarely, another gene, IMPG, described by Isabelle Meunier, Christian Hamel et al., causes
impairment of the interphotoreceptor matrix (IPM). This disturbs the cycle of internalisation/renewal
by the RPE.
In “non-genetic” AFVD, the most likely causal mechanism is local dysfunction of the RPE. The natural
history of AFVD is probably less severe and slower than that of AMD.
2. Clinical presentation
AFVD is described as typically bilateral in effect, but in practice is often asymmetrical. The functional
signs reported by patients with AFVD are those found with AMD (though often not all are present).
However, it is frequently discovered unexpectedly during a routine ophthalmology consultation after the
age of 40 years. A diagnosis may be suspected in the presence of a yellowish deposit of subretinal
material, without the patient reporting any issues, and visual acuity is frequently maintained (Figure 2).
This material is then subjected to as yet unknown factors and ends up breaking down and being
reabsorbed, sometimes with no impact on the photoreceptors or visual acuity, sometimes affecting retinal
function (atrophy). Neovascularisation may occur at any stage of the disease.
3) Macular atrophy
One of the most severe complications is macular atrophy complicating the resorption of the material:
visual acuity is often limited to 20/400 or less (Figure 5).
When it is the only observed ophthalmological sign, this atrophy makes retrospective diagnosis
difficult, if not impossible. The peri-atrophic retina may be free of any kind of lesion associated with
pattern dystrophy. A diagnosis must therefore be determined in the same way as with other causes of
bilateral macular atrophy: AMD, late-onset Stargardt disease, central areolar choroidal dystrophy, cone
dystrophy etc.
This outcome remains rare to our knowledge, but in the literature, geographic atrophy resulting from
AFVD is not usually studied separately to other diagnoses.
Fig. 4: Hyperreflective dots above the subretinal material in a 56-year-old male patient.
Visual acuity 20/20, no pigment migration observed.
4) Neovascular complications
As with atrophic progression, the frequency of choroidal neovascularisation (CNV) is difficult to quantify
not only due to the difficulty of defining AFVD as a clinical entity, but because the literature often does
not distinguish between neovascularisation and macular atrophy when discussing the complications of
AFVD. Furthermore, imaging results can be misleading if an exudative lesion is present, as it may be
mistaken for neovascularisation. According to the literature, CNV can complicate AFVD at any stage of
the disease. Its prevalence is approximately 10%.
3. Imaging
1) Multimodal imaging (Figures 6, 7 and 8)
Fig. 7: Multimodal imaging of a 67-year-old female patient with both AFVD and reticular pseudodrusen.
Left: Colour fundus photograph showing the accumulation of yellowish material. Middle: Infrared (IR) photograph.
Right: Blue-light autofluorescence photograph.
Fig. 8: Multimodal imaging of a 56-year-old male patient with AFVD and bilateral associated pattern dystrophy.
Top: SD-OCT showing the accumulation of material in the form of a regular, homogeneous, hyperreflective lesion above the retinal
pigment epithelium. Bottom left: Infrared (IR) photograph. Bottom right: Autofluorescence photograph showing the accumulation
of material and reticular lesions.
In general, there are several types of artefacts that can impact on the quality of the images. Clinical
experience and an analysis of the literature show that the following artefacts are most common in AFVD:
- Projection of the superficial capillary plexus
- Segmentation errors linked to the deposit of material, which disrupts the automatic recognition of the
anatomy of the normal retinal layers
- Movement artefacts: certain patients with severe forms of the disease (in particular, those with
neovascularisation and atrophy) may have difficulty holding their gaze, and image acquisition can
therefore be difficult.
When diagnosing choroidal neovascularisation using OCTA, it is important to identify any false positives and
false negatives linked to artefacts. For example, projection artefacts—caused by a hyperreflective retinal pigment
epithelium—cause variations in the reflected light, which are then interpreted as movement by the OCTA device,
leading to numerous false positives, such as for drusenoid PED. We have also observed this type of artefact
within the subretinal material, associated with artefacts linked to incorrect automatic segmentation (Figure 9).
The scientific literature concerning imaging for CNV associated with AFVD can be summarised as follows:
- CNV can be assessed qualitatively: it is clearly visible in OCTA, with good sensitivity and specificity
- The sensitivity of OCTA is better than or equal to that of angiography when diagnosing CNV
- Quantitative changes in retinal perfusion can be visualised for all the examined vascular plexuses.
4. Treatment
There is currently no specific treatment that affects the progression of AFVD. Unlike AMD, no studies
have found dietary supplements rich in antioxidants, omega-3, lutein or zeaxanthin to be therapeutically
effective for this indication.
However, the efficacy of anti-VEGF drugs in treating neovascular complications is well demonstrated: In
2013, Mimoun et al. published a case series of 24 patients with AFVD treated with anti-VEGF therapy
using a similar protocol to that used for AMD (three initial anti-VEGF injections in naive patients), with
anatomical and functional success. To our knowledge, no therapeutic studies have compared the different
anti-VEGF drugs in this indication. Dynamic phototherapy and laser treatment for AFVD have been found
to carry a risk of iatrogenic atrophy and were therefore dropped in favour of anti-VEGF drugs.
AFVD can be distinguished from CNV-induced serous retinal detachment through careful examination
of complementary images: numerous patients with AFVD are incorrectly treated with anti-VEGF drugs,
with no clear effect on the subretinal material.
In the future, gene function restoration protocols may be developed for AFVD with established genetic
causes, in particular through viral vector-mediated gene therapy.
References
Chowers I., Tiosano L., Audo I. et al. Adult-onset foveomacular vitelliform dystrophy A fresh perspective. Progress in Retinal and Eye Research;
2015:47:64-85.
Gass J.D., A clinicopathologic study of a peculiar foveomacular dystrophy. Transactions of the American Ophthalmological Society, 1974 :72 :139-56
Renner AB, Tillack H, Kraus H, et al. Morphology and functional characteristics in adult vitelliform macular dystrophy. Retina Phila Pa. 2004;24(6):929-
939.
Meunier I, Sénéchal A, Dhaenens C-M, et al. Systematic Screening of BEST1 and PRPH2 in Juvenile and Adult Vitelliform Macular Dystrophies: A
Rationale for Molecular Analysis. Ophthalmology. 2011;118(6):1130-1136
Meunier I, Manes G, Bocquet B, et al. Frequency and Clinical Pattern of Vitelliform Macular Dystrophy Caused by Mutations of Interphotoreceptor
Matrix IMPG1 and IMPG2 Genes. Ophthalmology. 2014;121(12):2406-2414. doi:10.1016/j.ophtha.2014.06.028
Lupidi M, Coscas G, Cagini C, Coscas F. Optical Coherence Tomography Angiography of a Choroidal Neovascularization in Adult Onset Foveomacular
Vitelliform Dystrophy: Pearls and Pitfalls. Investig Opthalmology Vis Sci. 2015;56(13):7638
Mimoun G, Caillaux V, Querques G, Rothschild P-R, Puche N, Souied EH. Ranibizumab for choroidal neovascularization associated with adult-onset
foveomacular vitelliform dystrophy: one-year results. Retina Phila Pa. 2013;33(3):513-521
Battaglia Parodi M, Rabiolo A, Cicinelli MV, Iacono P, Romano F, Bandello F. Quantitative analysis of Optical Coherence Tomography Angiography in
Adult-Onset Foveomacular Vitelliform Dystrophy: Retina. January 2017:1.
Geographic atrophy:
a new classification
Vittorio Capuano, Joel Uzzan
1. Definition
Geographic atrophy (GA) is the atrophic or “dry” form of age-related macular degeneration (AMD). It is
characterised by atrophy of the outer layers of the retina (retinal pigment epithelium and photoreceptors),
leading to more or less total loss of retinal sensitivity (scotoma)1.
GA occurs in the late phase of age-related maculopathy (ARM), which is characterised by the presence
of drusen and/or pseudodrusen associated with pigment epithelium changes (hypopigmentation or
hyperpigmentation).
2. Risk factors
From what we currently understand, GA appears to be a multifactorial disease.
Several studies have highlighted a number of risk factors for the development and progression of atrophic
AMD (other than age), including certain diets and smoking.
In terms of genetics, two loci associated with the disease have been identified in 1q32 and 10q262. The
first is a locus that also includes the gene coding for complement factor H (CFH), which is involved in
the alternative complement pathway (an inflammation factor). The second locus includes “age-related
maculopathy susceptibility 2” (ARMS2)3. However, genetic counselling is only called upon in exceptional
cases.
3. Diagnosis
Diagnosis is clinical and paraclinical: a visual acuity test, patient interview and multimodal imaging are the
three key stages. The use of the ETDRS score for visual acuity is strongly recommended, since it is better
suited to poor vision. Family history and clinical history are very important in establishing a diagnosis—in
particular, the patient’s age and the age at which the first symptoms appeared (potentially in the form
of nyctalopia). In terms of multimodal imaging, we are going to look at the three key examinations for
diagnosing GA.
The atrophic areas appear as round or polylobed whitish lesions, often multifocal, within which the large
choroidal vessels are more visible than usual (Figure 1). The edges are distinct and sometimes marked
by slight hyperpigmentation. These atrophic lesions vary greatly in position, number and size. A size of
0.05 mm2 was defined as the lower limit for diagnosing GA by a team of researchers in 19954.
More recently, the 2013 AREDS study amended this limit: the atrophic lesion must be a minimum of
0.145 mm2 (i.e. 1/4 of a disc diameter)5.
At present, several “wide-field” and “ultra-wide-field” fundus cameras are available on the market. The
value of this type of imaging lies in its ability to visualise lesions outside the arcades and more on the
periphery. While atrophic AMD always spares the middle periphery and rarely extends beyond the
arcades, the quality of wide-field images means they can usually be used to characterise macular lesions
(Figure 2).
Nevertheless, fundus photography alone is insufficient, in particular where the edges of the GA are
not easily visible. Pseudodrusen are often hard to see or not visible at all. In addition, the quality of
fundus photographs can be affected by milieu opacity or mediocre dilatation, and remains very operator-
dependent.
In 2017, a group of international retinal imaging experts put forward a new classification for retinal
atrophy based on OCT6,7.
According to the authors, OCT is the best tool for this classification, which employs characteristics specific
to this type of imaging:
- OCT displays every layer of the retina
- En-face reconstruction can highlight the cell loss in each layer
- OCT can detect areas of atrophy at an earlier stage
- “Eye tracker” mode allows a specific area of the retina to be monitored over time automatically
- OCT is the gold standard for diagnosing and monitoring exudative AMD and is a widely used tool
- OCT images are very similar to histological slice images
- OCT provides information on whether the fovea is affected by the disease.
The authors based their new terminology on the presence of complete or incomplete atrophy of two
structures visible in OCT: the outer retina and the retinal pigment epithelium (RPE). The combinations
thereof form four phenotypes:
• c-RORA (complete RPE and Outer Retinal Atrophy)
• i-RORA (incomplete RPE and Outer Retinal Atrophy)
• c-ORA (complete Outer Retinal Atrophy)
• i-ORA (incomplete Outer Retinal Atrophy).
This OCT-based classification has been validated by histological examination of donor samples in a
comparative post mortem study of OCT and histology; the OCT data and histology data were found to
match exactly.
“iRORA” does not meet any of the three “cRORA” characteristics, instead displaying inhomogeneous
hypertransmission and an irregular break in the RPE and photoreceptor lines (Figure 4).
“cORA” is defined by a complete absence of the photoreceptor line and the presence of an intact RPE
line.
“iORA” is defined by a break in the photoreceptor layer, along with an intact RPE line, and no
hypertransmission (Figure 5).
Other characteristic signs of GA are visible in OCT but are not used in this classification. The most well-
known include ghost drusen8, outer retinal corrugation9, wedge-shaped subretinal hyporeflectivity10 and
quiescent neovascularisation11.
a) Autofluorescence
Autofluorescence imaging of the fundus is the gold standard for measuring GA progression in multi-
centre studies, using the semi-automatic surface quantification software "Region Finder”, developed by
Frank Holtz. It is also a very useful examination in everyday practice to confirm a diagnosis of GA and to
quantify it.
Two types of autofluorescence device are currently available on the market: blue light and green light.
Both work in similar ways.
When the RPE is stimulated by light, it responds by releasing its own physiological fluorescence. If
the RPE is damaged, increased fluorescence is generally observed around the edges of the atrophy
(hyperautofluorescence). In the case of severe damage, there is no fluorescence (hypoautofluorescence)
(Figure 1).
Fig. 6: Confluent progression of GA in autofluorescence images over six years: centrifugal progression is quicker that centripetal.
The main disadvantages of autofluorescence are significant glare for the patient and a lack of information
on the condition of the other retinal layers. In addition, blue autofluorescence will not work effectively
around the foveal zone due to blocking by xanthophyll pigment; near-infrared autofluorescence must be
used instead.
d) Electrophysiology
Electrophysiology is not a routine complementary examination for diagnosing GA, but it is crucial in
distinguishing between GA and cone dystrophy or cone-rod dystrophy in which the cones are primarily
affected. It will primarily be of use in cases where the atrophy is not associated with hyperautofluorescent
deposits.
Conclusion
GA is diagnosed based on a careful assessment of several different imaging findings,
plus information gathered from the patient. A new, OCT-based classification has
recently been put forward that includes histology findings for the first time. With
no treatment available at this time, better description and classification of this form
of AMD remains difficult but necessarily multimodal.
2. Hageman GS, et al. A common haplotype in the complement regulatory gene factor H (HF1/CFH) predisposes individuals to age-related macular
degeneration. Proc Natl Acad Sci U S A. 2005;102:7227– 7232.
3. Rivera A, et al. Hypothetical LOC387715 is a second major susceptibility gene for age-related macular degeneration, contributing independently
of complement factor H to disease risk. Hum Mol Genet. 2005;14:3227– 3236.
4. Bird AC, Bressler NM, Bressler SB, et al. An International classification and grading system for age-related maculopathy and age-related macular
degeneration. The International ARM Epidemiology Study Group. Surv Ophthalmol 1995;39:367–374.
5. Danis RP, Domalpally A, Chew EY, et al. Methods and reproducibility of grading optimized digital color fundus photographs in the Age-Related
Eye Disease Study 2 (AREDS2 Report Number 2). Invest Ophthalmol Vis Sci 2013;54(7):4548–4554.
6. Sadda SR, Guymen R, Holz FG et al. Consensus Definition for Atrophy Associated with Age-Related Macular Degeneration on OCT: Classification
of Atrophy Report 3. Ophthalmology 2018.
7. Garrity ST, Sarraf D, Freund KB et al. Multimodal imaging of nonneovascular age-related macular degeneration. Invest Ophthalmol Vis Sci. 2018;
59:AMD48–AMD64.
8. Bonnet C, Querques G, Zerbib J. et al. Hyperreflective pyramidal structures on optical coherence tomography in geographic atrophy areas. Retina.
2014.
9. Ooto S, Vongulkariti S, Sato T. et al. Outer retinal corrugations in age-related macular degeneration. JAMA Ophthalmol. 2014;132:806–813.
10. Querques G. Capuano V, Bandello F. et al. WEDGE-SHAPED SUBRETINAL HYPOREFLECTIVITY IN GEOGRAPHIC ATROPHY. Retina, 2015.
11. Capuano V. Miere A, Querques L, et al. Treatment-Naïve Quiescent Choroidal Neovascularization in Geographic Atrophy Secondary to
Nonexudative Age-Related Macular Degeneration. Am J Ophthalmol. 2017 Oct;182:45-55.
12. Schmitz-Valckenberg S, Sahle JA, Danis R. et al. Natural history of geographic atrophy progression secondary to age-related macular degeneration
(greographic atrophy progression study). Ophthalmology 2016.
13. Sacconi R, Corbelli E, Carnevali A. et al. OPTICAL COHERENCE TOMOGRAPHY ANGIOGRAPHY IN GEOGRAPHIC ATROPHY. Retina. 2017.
14. Corbelli E, Sacconi R, Rabiolo A. et al. Optical Coherence Tomography Angiography in the Evaluation of Geographic Atrophy Area Extension.
Invest Ophthalmol Vis Sci. 2017.
Differential diagnoses
for AMD
Polina Astroz, Olivia Zambrowski
AMD is the most common cause of visual impairment in the developed world. However, it is important
to be aware of several other more rare diagnoses that have their own specific characteristics but
can be complicated by atrophy or macular neovascularisation. There are many causes of macular
atrophy. Some of these diseases can also be complicated by neovascularisation. These are often
found in younger patients than those affected by AMD, but late-onset forms are not uncommon, and
patients may present at consultations in the scarring stage.
Patient interviews, family history, clinical examination and conventional multimodal imaging (MMI)
combined with OCTA are essential for both diagnosis and treatment.
1. Pachychoroid
This new entity was proposed in 2013 and covers a spectrum of clinical presentations with shared
characteristics: reduced visibility of the choroidal vessels in the fundus, focal or diffuse thickening
of the choroid, dilation of the "pachyvessels” in Haller’s layer with thinning of Sattler’s layer and the
choriocapillaris, and choroidal hyperpermeability visible in the late phase of indocyanine green (ICG)
angiography.
This primarily affects young men (72–88% of cases). Several risk factors have been described, including
endogenous or exogenous corticosteroid therapy and type A personality, among others.
MMI is essential to both diagnosis and monitoring. In autofluorescence imaging, RPE defects are typically
visible, associated with gravitational tracks. SD-OCT shows SRD, small PEDs corresponding to leakage
points, extension of the photoreceptor outer segments and—in chronic forms—a flat, irregular PED, fibrin,
degenerative macular oedema and outer retinal tubulation. Flat, irregular pigment epithelial detachment
(FIPED) in chronic CSCR is associated in 19–29% of cases with type 1 CNV in conventional MMI.
However, due to the RPE changes that occur in chronic forms, it is often difficult to identify this type 1
CNV. OCT-angiography (OCTA) is extremely useful in CSCR. In the literature, OCTA detects type 1 CNV
in 35–95% of cases versus19–29% of cases for conventional multimodal imaging (Figure 1). CSCR with
or without neovascularisation can be mistaken for exudative AMD.
2) Pachychoroid neovasculopathy
Type 1 CNV associated with a pachychoroid and with no other cause of CNV (AMD, myopia, inflammatory
diseases etc.) has been termed “pachychoroid neovasculopathy”.
D E F
G H
Fig. 1: Multimodal imaging and OCTA of the left eye of a male patient with central serous chorioretinopathy complicated by type 1
choroidal neovascularisation.
Colour fundus photographs (A). Autofluorescence image showing retinal pigment epithelium changes and gravitational tracks (B).
Late-phase fluorescein angiography showing macular pinpoints (C). Pachyvessels are visible in the early phase of ICG angiography
(D), along with late hyperpermeability (E). OCTA with manual segmentation, showing an area between the retinal pigment
epithelium (RPE) and 30 microns below it with neovascular flow (F). The flow is also visible in the corresponding B-scan (H). EDI-
OCT shows a FIPED associated with a SRD and a thick choroid measuring 460 microns (G).
Pachychoroid neovasculopathy and CSCR with type 1 CNV can be complicated by polyps in 36% of
cases. Type 1 CNV is the equivalent of the branching vascular network. Recently, a new term has been
put forward for PCV: “aneurysmal type 1 CNV”, but this remains controversial.
A B C
D E
Fig. 2: Multimodal imaging of a 65-year-old female patient with progressive loss of visual acuity and a central scotoma, associated
with disabling night vision impairment.
(A) In the autofluorescence photo, the area of atrophy is not particularly dense at this stage, and contrasts with the extent of the
impairment. The vertical appearance of the lesions also suggests a possible diagnosis of EMAP. (B) In the infrared photograph and
SD-OCT (C), the reticular pseudodrusen and macular atrophy with complete restructuring of the ellipsoid zone are also supportive of
this diagnosis. (D) The atrophy in the fundus has no specific appearance, but the peripheral areas should be checked for squamous
lesions. (E) Finally, full-field ERG reveals highly altered responses, with significant dysfunction of both the rod and cone systems.
The diagnosis of EMAP is confirmed.
Type 1: Men aged over 55 years, unilateral lesion, sometimes associated with peripheral vascular
abnormalities. Cystoid macular oedema (CMO) and exudates are common
Type 2: Men or women aged under 60 years, slight bilateral juxtafoveal capillary dilatation, generally in
the temporal area within the median raphe. Risk of neovascularisation
Type 3: Men aged 50 years, bilateral occlusive vascular capillary lesion with no peripheral vascular
abnormalities.
Type 2 or “mac tel” lesions can be quite easily mistaken for neovascular AMD. These lesions are
characterised by acute bilateral macular disease, visible in SD-OCT as intraretinal spaces (with no
systematic thickening of the macula) in the juxtafoveal zone, associated in fluorescein angiography with
moderate leakage in the late phase. These cavitation-type spaces initially appear in the inner layers of
the retina, then progress towards the outer layers, with the development of secondary atrophy or, in
rare cases, subretinal neovascularisation. OCTA is a new tool that can be used to diagnose the condition
by displaying vascular abnormalities in several different retinal layers. In the early stages, rarefaction of
the deep and superficial capillary networks is observed, with focal capillary dilatations associated with
a right-angle drainage vein. Next, the retinal capillaries in the outer retina are affected as a result of
the changes in the superficial and deep capillary plexus. If subretinal neovascularisation has occurred,
OCTA can be used to diagnose the neovascularisation without fluorescein diffusion getting in the way
(Figure 3).
A C D
Fig. 3: A 55-year-old man with decreased visual acuity and recent metamorphopsia in the left eye.
(A) The Multicolor® photograph reveals a non-haemorrhagic superior parafoveal lesion, (B) also visible in the infrared image.
(C and D) Leakage is visible around the lesion in fluorescein angiography, increasing over the course of the sequence, but the SD-
OCT image (E) shows cavitation in the inner layers of the macular with no overall increase in macular thickness, associated with
changes to the ellipsoid line, which is not typical of exudative AMD. (F) The OCTA image confirms the diagnosis of type 2 macular
telangiectasia (“mac tel”). The lesion is present in all the retinal slices, with rarefaction of the surrounding capillaries and, crucially,
a right-angle drainage vein.
a) Pattern dystrophy
Pattern dystrophy (or reticular, butterfly-shaped or Deutman’s dystrophy) covers heterogeneous bilateral
autosomal-dominant dystrophies characterised by yellowish macular deposits and pigment migration. In
general, the visual prognosis is good, with a slight decrease in central vision in the elderly.
d) Stargardt disease
This disease is genetically and phenotypically heterogeneous. It particularly affects young people, but can
sometimes be discovered later, making differential diagnosis with AMD more difficult. The most commonly
mutated gene is ABCA4 and transmission is autosomal recessive. It is characterised by flecking in the
posterior pole, no peripapillary involvement and a dark choroid in fluorescein angiography, progressing to
complete macular atrophy (Figure 4). Fundus flavimaculatus is characterised by a later onset and slower
progression. Rare cases of neovascularisation have been described in the literature.
e) Best disease
Best disease is autosomal dominant and of variable penetrance. It is characterised by the appearance of a
vitelliform macular deposit in young adults or children, and progresses through five stages: previtelliform,
vitelliform, pseudohypopyon, vitelliruptive and atrophy (Figure 5). Rare cases of neovascularisation have
been described. Differential diagnosis with AMD becomes difficult at the atrophic stage. The diagnosis
can be confirmed by examining the patient’s family history and identifying any changes in EOG response
or vitelliform material in the contralateral eye.
A B
A B D
C E
Genetic cases of retinal dystrophy can be discovered very late. If a patient presents with an atrophic or
exudative macular lesion, it can be easily confused with AMD, particularly in the presence of macular
oedema and, in rarer cases, neovascular complications.
If the retinal imaging (OCT, autofluorescence) or patient history are atypical, it is important to look for
photophobia, nyctalopia or a family history of retinal disease, as well as associated signs in the fundus,
particularly in the retinal periphery (e.g. osteoblasts or white spots). Electrophysiological examination
should also be considered (full-field ERG, electro-oculogram, multifocal ERG). If there are any functional
abnormalities in the cone and/or rod system as a whole, it cannot simply be a case of AMD, since the
functional impairment in this case spreads far beyond the limits of the posterior pole.
In order to distinguish between macular dystrophy, cone dystrophy and cone-rod (or rod-cone) dystrophy,
it is useful to examine the initial symptoms (nyctalopia or photophobia), family history, associated signs
in the fundus and the percentage reduction in the cone and rod system response ranges in full-field ERG
(Figure 6).
A B
C D
Fig. 6: 82-year-old woman monitored for several years for atrophic AMD.
The atrophic lesions in the fundus (C) are extensive and the patient reports a considerable loss of independence, although the fovea
remains unaffected (A and B). Further discussion with the patient reveals consanguinity, photophobia dating back several years
and, recently, nyctalopia: a full-field ERG is therefore performed. (D) The ERG responses confirm cone-rod dysfunction, compatible
with a diagnosis of cone-rod dystrophy, probably of genetic origin and autosomal recessive.
5. Drug toxicity
1) Synthetic antimalarials
Synthetic antimalarials (SAMs) are widely used to treat chronic inflammatory diseases.
Their toxic effects on the retina have been known for a long time, but the mechanism by which this occurs
remains poorly understood. Prevalence varies depending on the dose and treatment duration: the main
risk factor is the daily dose, correlated with actual weight.
In the early stages, the effects of SAM toxicity cannot be confused with AMD: it is characterised by
thinning of the ellipsoid zone and the outer nuclear layer, which has a “flying saucer” appearance. In the
late stages, the RPE atrophy begins to look like a bull’s-eye, but can sometimes extend across a large area
of the posterior pole. It is in these later forms that differential diagnosis can become difficult, particularly
if the treatment was discontinued a long time ago, if the disease developed slowly and if the patient
presents at the age of around 60 years complaining of unexplained reduced visual acuity with significant
macular atrophy.
The diagnosis can be confirmed through targeted questioning of the patient and if there is no or only
partial foveal involvement, no neovascularisation, highly symmetrical atrophy, and, in general, changes in
the cone and rod system responses in full-field ERG (Figure 7).
2) Tamoxifen
Tamoxifen is an oestrogen blocker primarily used as an adjuvant therapy for hormone-dependent breast
cancer. Maculopathy was first identified as a side effect in 1978, and is characterised by yellowish
perifoveal deposits associated with pigment migration and angiographic leakage.
Maculopathy is also found with lower doses of tamoxifen, involving macular thinning, cavitations that
appear atrophic and, in some cases, pseudoholes. Some patients may display changes in the ellipsoid
zone and foveal photoreceptors. In these cases, the patient should be asked whether they have used this
medicine—particularly over the long term—before a diagnosis of atrophic AMD or macular telangiectasia
is made. Atrophic lesions, pigment migration and pseudocystic cavities do not appear to be reversible
upon treatment discontinuation, unlike other potential signs, such as corneal deposits and macular
oedema (Figure 8).
3) Deferoxamine
Deferoxamine is an iron and aluminium chelator used in the treatment of haemochromatosis and in
patients who receive multiple blood transfusions. Maculopathy is just one of its many side effects, and
was first identified as such in the 1980s. It primarily affects the retinal pigment epithelium, with a direct
toxic effect on the cells of the RPE and/or Bruch’s membrane, with pigment epithelium changes in the
fundus, visible more clearly in autofluorescence images. In the foveal zone, SD-OCT reveals thinning of
the RPE, with deposits, irregularities or atrophy; breaks in the ellipsoid line; photoreceptor attenuation;
and, in some cases, tubulation around the areas of atrophy. In the early stages of intoxication, OCT may
reveal a serous retinal detachment with extension of the outer segments, which resolves upon treatment
discontinuation.
A B C
D
Fig. 7: A 57-year-old female patient with progressive,
bilateral loss of visual acuity. The fundus (A) is pale,
but the atrophy is very clear, although it does not
affect the fovea. The autofluorescence photo (B) shows
subfoveal atrophy extending to the arcades, with
areas of heterogeneous hypoautofluorescence and
hyperautofluorescence. In SD-OCT (C), the unaffected
fovea probably explains why the patient has retained
her independence; she has not seen a doctor despite
receiving synthetic antimalarials for 15 years. Finally, the
full-field ERG responses (D) are of normal morphology,
but the amplitudes are lower, confirming a disorder of
both the cone and rod systems.
A B C D
Fig. 8: A female patient presenting with decreased visual acuity during selective oestrogen receptor modulator therapy for breast
cancer.
The presence of a crystalline deposit in the fundus (A) in this context suggests this is the most likely diagnosis. The crystals
are easily visible in the infrared (C) and autofluorescence (B) images. In SD-OCT (D), the pseudo-telangiectasia, with intraretinal
cavitations associated with ellipsoid line changes also confirms the diagnosis.
A B C
D E
Fig. 9: An 85-year-old female patient seeking a second opinion for atrophic AMD, since she has experienced a significant loss of
independence in recent months.
In the fundus (A), the atrophic lesion is aspecific, with several small pigmented spots superior to the fovea. The atrophy around the
fovea has a “bunch of grapes” appearance, which calls the diagnosis into question (B). In the ultra-wide-field photos (C), there are
highly heterogeneous areas of hypoautofluorescence and hyperautofluorescence, with no obvious distribution pattern. Fluorescein
and indocyanine green angiography rule out an inflammatory disease (white dot syndrome). SD-OCT (D) does not provide any
further information on the origin of these lesions. Full-field ERG (E) confirms moderate overall retinal dysfunction which—in this
patient with recent, rapid loss of visual acuity—is suggestive of a paraneoplastic syndrome. Further aetiological investigation
confirmed the presence of colon cancer, which no doubt caused the paraneoplastic syndrome.
At present, some authors believe these two clinical presentations are distinct, while others believe
they lie on a continuum of the same disease. The visual prognosis for these patients depends on the
level of macular complication: active and/or atrophic inflammatory macular lesions; CNV (6.7–76.9%),
usually type 2; macular oedema (0–37.5%); and epiretinal membrane (0–11%). Fundus examination
reveals white-to-yellowish macular and/or peripheral patches progressing towards atrophy, potentially
complicated by CNV. It can be difficult to differentiate the active patches from CNV using conventional
MMI. Only 66.7% of CNV presents exudative signs in SD-OCT. OCTA has proven highly useful in this
disease due to its ability to diagnose CNV at an early stage (Figure 10). In the literature, 6–14% of lesions
thought to be active inflammatory patches in conventional MMI were found to have neovascular flow in
OCTA.
APMPPE has been described as a disease of the retinal pigment epithelium. It is characterised by the
appearance of whitish polycyclic plaques in the posterior pole, which heal spontaneously in the majority
of cases. However, relapse is possible, sometimes with atrophic macular complications or, more rarely,
CNV. The observed abnormalities may be caused by ischaemia of the choriocapillaris.
A B
D E
F H
Fig. 10: Multimodal imaging of a 57-year-old female patient with multifocal choroiditis complicated by type 2 choroidal
neovascularisation along the edges of an area of macular atrophy in the right eye.
The colour fundus photo (A) shows several atrophic patches that are hypoautofluorescent (B). The SD-OCT slice (C) through the
superior section of the macular atrophy (dotted line in image E) shows subretinal hyperreflectivity associated with a SRD compatible
with active type 2 CNV. Fluorescein angiography shows early hyperfluorescence (D) with late leakage from the lesion (E) and a late-
phase plaque in ICG (F). Finally, OCTA shows the neovascular flow in the manually segmented slab, between the retinal pigment
epithelium and 30 microns below it (G), with flow in the corresponding B-scan slice (H), confirming the neovascular complication.
6) Birdshot chorioretinopathy
This is a rare, bilateral chorioretinopathy with an understated inflammatory reaction in the anterior
chamber, vitritis, vasculitis and whitish spots. It is strongly associated with HLA-A29. Its complications
are macular oedema (100%), epiretinal membrane, CNV (11%), atrophy and vasculitis.
Conclusion
The most common cause of atrophy and neovascularisation in patients aged over
55 years is AMD. However, several differential diagnoses must be considered,
particularly in the absence of drusen. A carefully conducted patient interview
(disease history, family history, medication), clinical examination with MMI plus
OCTA and, in some cases, additional examinations (ERG, EOG) are essential to
making the correct diagnosis and adapting treatment and monitoring accordingly.
no Fundus, ERG
yes
OCTA EMAP
Quick
ERG +/- normal progression
White dot syndromes
OCT, Nyctalopia
Mactel ICG Vertical lesion
Stargardt
Pattern MFC
dystrophy Pachyvessels: Serpiginous
Best CSCR AZOOR
Adult-onset
foveomacular
Neovasculopathy MEWDS
Polyps AMPPE
vitelliform
dystrophy PPM
DVA: decreased visual acuity; F: fundus; FAF: fundus autofluorescence; MMI: multimodal imaging;
FA: fluorescein angiography.
- EMAP:
Hamel, CP, Meunier I, Arndt I, et al. Extensive Macular Atrophy with Pseudo-drusen-like Appearance: A New Clinical Entity. American Journal of
Ophthalmology ; 2009: 609 20.
Douillard A Picot MC, Delcourt C, et al. Dietary, Environmental, and Genetic Risk Factors of Extensive Macular Atrophy with Pseudo-drusen, a Severe
Bilateral Macular Atrophy of Middle-Aged Patients. Scientific Reports ; 2018: 6840
- Macular telangiectasia:
Gass JD, Blodi BA. Idiopathic juxtafoveolar retinal telangiectasis. Update of classification and follow up study. Ophthalmology ; 1993 :1536-1546
Yannuzzi LA., Bardal AM, Freund KB et al. Idiopathic Macular Telangiectasia. Retina 2006 450 60.
Toto, L, Di Antonio L, Mastropasqua R, et al. Multimodal Imaging of Macular Telangiectasia Type 2: Focus on Vascular Changes Using Optical
Coherence Tomography Angiography. Investigative Ophthalmology & Visual Science ; 2016 : 268-276.
Capuano, V, Miere A, Amoroso F, et al. Uncommon retinal vascular diseases. 2016 ; 453-73.
Spaide RF, Yannuzzi LA, Maloca PM. Retinal-choroidal anastomosis in macular telangiectasia Type 2 . Retina ; 2018 : 1920 29.
- Drug toxicity:
Marmor MF, Kellner U, Lai TYY, et al. American Academy of Ophthalmology. Recommendations on screening for chloroquine and hydroxychloroquine
retinopathy (2016 Revision). Ophthalmology 2016:1386-94.
Nair AA, Marmor MF. ERG and other discriminators between advanced hydroxychloroquine retinopathy and retinitis pigmentosa. Documenta
Ophthalmologica. 2017: 175-183
Koinzer S, Klettner A, Treumer F, et al. Correlation of fundus autofluorescence, spectral-domain optical coherence tomography, and microperimetry in
late deferoxamine maculopathy. Retinal Cases Brief Report 2012:50-5.
Kaiser-Kupfer MI, Lippman ME. Tamoxifen retinopathy. Cancer Treat Rep 1978:315-20.
Kaiser-Kupfer MI, Kupfer C, Rodrigues MM. Tamoxifen Retinopathy: A Clinicopathologic Report. Ophthalmology 1981:89-93.
Faure C, Paques M, Audo I. Electrophysiological features and multimodal imaging in ritonavir-related maculopathy. Documenta Ophthalmologica.
2017:241-8.
- CAR syndrome:
Sadowski, B., Kriegbaum C, Apfelstedt-Sylla E. Tamoxifen Side Effects, Age-Related Macular Degeneration (AMD) or Cancer Associated Retinopathy
(CAR)? European Journal of Ophthalmology. 2001: 309 12.
Micronutrition
and age-related
macular degeneration
Sergio Piscitello, Pierre Sustronck
AMD is a complex, multifactorial disease. Genetics has been shown to play a major role in
predisposition to AMD, but environmental factors—both protective and aggravating—are also
fundamental. Several genetic polymorphisms have been identified to date, notably genes involved
in the complement system (CFH, C3, C2, CFI, CFB), cholesterol metabolism (CETP, LIPC, ABCA1,
APOE), extracellular matrix remodelling (TIMP3 and COL8A1) and oxidative stress (ARMS2). While
genetics, the presence of these polymorphisms and patient age are non-modifiable risk factors,
there are other modifiable factors. The main ones are smoking, excess weight and a diet lacking in
certain micronutrients. There are three major types of micronutrients that appear to play a protective
role in AMD: antioxidant vitamins and minerals, macular pigments and omega-3 polyunsaturated
fatty acids.
In terms of five-year visual acuity, the probability of vision loss of at least 15 letters in category 3 and
4 patients was 29% for the placebo group and 23% for those treated with zinc + antioxidants. One of
the limitations of the study was that it was not designed to identify whether dietary supplements reduce
disease progression in patients who already have severe AMD in both eyes.
Table 1
• Vitamin C 500 mg/day
Antioxidants • Vitamin E 400 IU/day
• Beta-carotene 15 mg/day (only if patient a non-smoker)
Zinc and • Zinc oxide 80 mg/day
copper • Copper oxide 2 mg/day (to prevent anaemia)
Table 2
40% Placebo
Antioxidants
Zinc 28%
30%
Estimated Probability
Antioxidants + Zinc
20%
20%
10%
P versus A+Z - p<0.01
0%
A 25% Risk
Reduction 0 1 2 3 4 5 6 7
Years
AREDS Report No. 8. Arch Ophtholmol. 2001;119:1417-36.
2. Macular pigment
1) Definition
There are over 600 carotenoids in the natural world, 140 of which are edible. Three of these are the main
components of the yellow macular pigment. The three isomers in macular pigment are lutein (structurally
related to alpha-carotene), zeaxanthin (related to beta-carotene) and, to a lesser extent, meso-zeaxanthin
(an isomer of zeaxanthin). They belong to a subcategory of carotenoids known as xanthophyll pigment.
There are twenty or so carotenoids found in human serum, the five main ones being lycopene, alpha-
carotene, beta-carotene, lutein and zeaxanthin. Only lutein, zeaxanthin and meso-zeaxanthin are found
in the macula, at concentrations around 10,000 times those found in serum.
The maximum concentrations are found in the centre of the fovea, where zeaxanthin predominates.
However, lutein is found in greater concentrations than zeaxanthin in the periphery area. All three are
primarily concentrated in the Henle fibre layer (zeaxanthin mainly in the cones and lutein in the rods), but
are also found in the retinal pigment epithelium and photoreceptor outer segments.
Macular pigment improves the visual performance of the macula by reducing chromatic aberrations and
glare. It also has antioxidant properties, both direct (as an antioxidant agent) and indirect (by filtering out
short wavelengths, which provides photochemical protection via lipofuscin), as well as anti-inflammatory
properties. Zeaxanthin and lutein are not synthesised by the body and must be sourced entirely from
food. Numerous studies have highlighted the protective properties of a diet rich in carotenoids vis-à-vis
the risk of AMD.
In the 1990s, Seddon et al. were the first to demonstrate that increasing one’s consumption of foods
rich in certain carotenoids—in particularly certain dark green leafy vegetables—can reduce the risk of
developing advanced or exudative AMD3. The prospective POLA2 study, conducted on 2584 patients,
proved that the risk of AMD (early or severe) reduces by 79% in subjects with high plasma concentrations
of zeaxanthin and lutein. In the Eye Disease Case Control Study Group5, subjects with higher carotenoid
serum concentrations had a markedly reduced risk of neovascular AMD. In 2003, a study by Gale4 et al.
found that this risk was significantly higher in people with lower plasma concentrations of zeaxanthin.
In 2006, the CAREDS6 study found that diets rich in zeaxanthin and lutein can protect healthy women
under the age of 75 from developing intermediate AMD. Other epidemiological studies have had
similar findings22. The AREDS2 study also looked at omega-3 supplementation, comparing this to the
supplementation considered in the first AREDS study23.
The randomised, double-blind, interventional, prospective LAST7 study, of sound methodology, assessed
90 patients with atrophic AMD over 12 months.
The authors evaluated the effects of supplementation with lutein alone or in combination with other
carotenoids and antioxidants on macular pigment optical density (MPOD), measured using heterochromatic
flicker photometry (MacularMetrics® device) and in objective vision tests. MPOD increased by 36% in the
lutein group (L) and by 43% in the lutein plus antioxidants, vitamins and minerals group (L/A) (Figure 2). In
addition, visual acuity increased by 5.4 letters in the L group and 3.5 in the L/A group. Contrast sensitivity
improved in both groups, and quality of life (measured by a VF-14 questionnaire relating to subjective
recovery after glare) improved in the L/A group.
These results suggest that xanthophyll pigments play a protective role in the prevention of AMD.
0.5
MPOD Thickness
0.4
0.3
0.2
0.1
0
Lutein Lutein / A Placebo
Fatty acids are referred to as "unsaturated” when they contain at least one double covalent bond
between the carbon atoms (C=C). Some of these fatty acids (such as omega-3 and omega-6) cannot be
synthesized by the body and must be provided through the diet.
2) DHA
Eicosapentaenoic acid (EPA) is partially converted into docosahexaenoic acid (DHA) and omega-3
after being absorbed through the intestines. Omega-3 has numerous biological effects on the blood
vessels and tissues, acting through signal transduction, gene expression regulation and cell membrane
remodelling. DHA is a key lipid component of the photoreceptor membranes, where it plays a crucial role
in maintaining their structural and functional integrity. In the retina, DHA increases mitochondrial activity
and has antioxidant, anti-inflammatory, antiapoptotic and antiangiogenic effects. The continual renewal
of the retinal membranes requires a constant source of omega-3. This is why diets rich in DHA and EPA
can improve retinal function and delay the development of AMD.
In the 2000s, numerous scientific studies found a relationship between diets rich in DHA and low in
saturated fats and a reduction in the risk of neovascular AMD.
Seddon et al. showed that increased omega-3 intake is associated with a lower risk of AMD in people
with diets low in linoleic acid (an omega-6)8. Studies by Smith W9, Augood10 and Merle11, the US Twin
Study12 and the POLANUT13 study found that more frequent consumption of omega-3-rich fish is
associated with a reduced risk of neovascular AMD. The findings of these studies show that the role of
different types of dietary lipids is complex, in particular the relationship between saturated fats, omega-3,
omega-6 and cholesterol.
In the Blue Mountains Eye Study, a clear association was observed between more frequent consumption
of fish and a reduced risk of severe AMD14. Another study, conducted by Augood et al., specifically
examined the protective effect of EPA and DHA in elderly subjects and found that consuming fish oil at
least once a week protected against neovascular AMD, as did a high dietary intake of DHA and EPA15.
In 2009, the pilot study NAT1 found that omega-3 supplementation (DHA 480 mg/day; EPA 720 mg/
day) was well tolerated and omega-3 concentration increased considerably in serum and red blood cell
membranes16.
NAT2 was a randomised, double-blind, prospective, comparative study conducted in patients with
unilateral neovascular AMD (CNV) receiving DHA (840 mg/day) and EPA (270 mg/day) or a placebo orally
for a period of three years. The time to bilateral involvement was not significantly different between the
treated group (19.5 ±10.9 months) and the placebo group (18.7 ±10.6 months). Similarly, there was no
significant difference in the incidence of CNV between the two groups (28.4% vs 25.6%, respectively).
However, an analysis of fatty acid levels in the red blood cell membranes identified a subgroup (the
upper tertile) of patients receiving the supplement who regularly had high levels of DHA+EPA in these
membranes. This subgroup had an almost 70% lower risk of developing CNV in the second eye than
those in the lower tertile (p = 0.047).
The results of the AREDS2 study (multi-centre, randomised, double-blind over five years) were
unexpected. In this study, 507 patients supplemented with EPA+DHA (650 mg/day + 350 mg/day)
developed severe AMD compared to 493 in the control group treated with the AREDS formula (HR
0.97; 95% CI 0.82–1.16, p = 0.7), which suggests that omega-3 does not provide significant additional
protection. However, it may be that the study design prevented any potential prophylactic effect of
omega-3 from being demonstrated, given the low DHA intake during the study, the use of DHA ethyl
ester and the failure to check actual capsule consumption by measuring membrane fatty acids.
It should be noted that there are several factors that could affect the reliability of these studies, including
patient consumption of relevant functional foods, additional unreported self-supplementation in the
control group, and treatment noncompliance in the treated group. The only way to truly identify the
benefits of omega-3 on a slowly progressing disease such as AMD is to measure membrane fatty acids.
Complex interactions between genotype and diet should not be underestimated. A study by Reynolds
et al.19 found that subjects who are homozygous for the genotype ARMS2/HTRA1, which confers a high
risk of AMD, can be protected from geographic atrophy by increased DHA intake. Those who do not have
the genotype for this risk are not protected. Similarly, Merle BM et al. suggest that a genetic predisposition
to AMD resulting from the CFH Y402H variant limits the benefits of DHA supplementation20. Regardless,
it appears that people with a genetic predisposition to AMD can potentially benefit from taking omega-3
supplements. Pragmatically speaking, since the risk of taking omega-3 is low16, the potential benefits can
be considered greater than the risks, and the consumption of omega-3-rich supplements can therefore
be recommended.
4. Dietary advice
A healthy, varied and nutritionally balanced diet can complement standard medical treatments. Patients
can be advised to follow a diet rich in certain vitamins, minerals and omega-3.
Lutein and zeaxanthin are found in cabbage, broccoli, spinach, turnips, lettuce, peas, sweetcorn, green
beans, carrots and celery. These can be recommended for consumption four times a week.
Vitamin E can be found in vegetable oils, nuts and seeds (walnuts, hazelnuts, almonds, pecans, pine nuts
and sunflower seeds) and fresh fruit and vegetables (fennel, peas, salsify, avocado, spinach, parsley,
cabbage, kiwi, blueberries, mango and chestnuts). Two tablespoons of a variety of vegetable oils should
be consumed daily, along with plenty of the aforementioned fresh fruit and vegetables.
Foods rich in vitamin C include fresh vegetables such as peppers, broccoli, cauliflower, cabbage, fennel
and spinach—particularly when eaten raw—and fruit such as kiwi, citrus fruit, blackcurrants, strawberries,
mango, blackberries, redcurrants, passion fruit, raspberries and melon. One raw vegetable and one fruit
is recommended at every meal. High concentrations of omega-3 are found in oily fish (salmon, herring,
trout, mackerel, sardines and tuna), nuts, and rapeseed, walnut and soya oil. Oily fish should be consumed
at least twice a week. The main zinc-rich foods are meat, vegetables, eggs, wholegrain cereals, fish and
dark chocolate.
After discussing diet and lifestyle, vitamin supplements can be offered to patients
with advanced, unilateral AMD, since these seem to have a protective effect,
reducing the risk of bilateral involvement at five years by 25%, according to the
AREDS study1,21.
Treatment protocols
for exudative AMD
Hassiba Oubraham
Anti-VEGF intravitreal injections are the first-line treatment for sight-threatening exudative AMD.
Three anti-VEGF drugs have been proven effective for this indication: ranibizumab, aflibercept and
bevacizumab. The first two have an MA in France, while the third has a “temporary recommendation
for [off-label] use” (“RTU”) in hospitals.
The treatment regimens can be subdivided into two broad categories: a “reactive” regimen, which
delivers the treatment after a relapse has occurred, and a “proactive” regimen, which treats the
condition “a priori” before a relapse happens.
1. Reactive regimens
1) Simple PRN (“pro re nata” or “as needed”)
After an induction phase of three injections each administered one month apart, the patient is monitored
on a strict monthly basis. One or more further anti-VEGF injections are given if the disease is reactivated of
if there are still signs of disease activity (Figure 1). The standard activity criteria for neovascular AMD are:
- A reduction in visual acuity of more than five letters (on the ETDRS scale) compared to the previous
examination, attributable to choroidal neovascularisation
- The recent appearance of subretinal haemorrhage
- The presence of a macular serous retinal detachment
- The presence of intraretinal oedema spaces
- Significant expansion of a vascularised pigment epithelial detachment or signs of disease activity
revealed by fluorescein angiography1 or indocyanine green angiography
- Hyperreflective subretinal lesions in OCT2, which are also considered to be signs of neovascular
activity.
Practitioners must be aware of these signs and take them into account when making treatment decisions1,2.
L M M J V S D L M M J V S D L M M J V S D L M M J V S D L M M J V S D L M M J V S D L M M J V S D L M M J V S D L M M J V S D L M M J V S D L M M J V S D L M M J V S D L M M J V S D
Month 0 1 2 3 4 5 6 7 8 9 10 11 12
IVI
In total, after 12 months of monitoring, the patient had an induction phase with 3 intravitreal injections (IVI)
Strict monthly check-ups revealed 2 relapses, which were retreated
Total of 5 IVIs in the first year
D E
F G
Fig. 2: Example of type 2 neovascularisation treated with anti-VEGF and monitored based on the PRN regimen, with two relapses and visual
loss due to subretinal fibrous scarring.
A: Initial presentation: exudative lesion in the central macula, with pre-epithelial hyperreflective material, intraretinal fluid and intraretinal
cysts in the central fovea. B: En-face OCTA shows the presence of flow, with a characteristic jellyfish appearance. C: One month after the
induction phase with three anti-VEGF intravitreal injections (IVI): the intraretinal and subretinal fluid has disappeared, as have the intraretinal
cysts. Anatomical response is good. D: First relapse two months after the last IVI: reappearance of intraretinal fluid with thickening of the pre-
epithelial hyperreflective structure (CNV) and diffuse, very mild retinal thickening (arrow). An injection is given. E: Anatomical improvement
after the fourth IVI. F, G: Major relapse despite regular monthly monitoring, causing a major, irreversible loss of visual acuity due to fibrous
scarring of the central fovea (star). H: Visual acuity curve (based on ETDRS letter score) during the monitoring period. The irreversible drop in
visual acuity after the second major neovascular relapse can be seen, caused by subretinal fibrous scarring (the vertical dotted lines represent
the anti-VEGF IVIs).
The PRN regimen was found to be highly effective in the PrONTO1 uncontrolled pilot study in terms of
improving and maintaining visual acuity over 24 months. Randomised, controlled trials have proven its
non-inferiority compared to strict monthly treatment (the treatment method used in the pivotal MARINA3
and ANCHOR4 studies): see the HARBOR5 study and CATT6 study for ranibizumab.
However, its transposition into clinical practice (real-life studies) has been a resounding failure, both in
France7-10 and worldwide11,12. There are many reasons for this failure, but the two main ones are doctors’
inability to impose strict monthly monitoring, and a lack of patient compliance with this schedule due to
comorbidities and transport problems13.
Not being able to treat patients in a timely fashion and waiting for relapse can result in fibrotic lesions and
irreversible loss of visual acuity14 (Figure 2).
In 2009, Lala et al.15 introduced the concept of a “relapse treatment” involving three injections at one-
month intervals (instead of the single injection recommended in the PrONTO study). Their retrospective,
single-centre study of 316 patients used a treatment protocol comprising a 3-injection induction phase
followed by monthly check-ups with repeat treatment as needed. Relapses were treated with three
monthly injections, while stable patients were treated every three months as standard. The average
visual improvement after 3 years of follow-up was 8 letters for an average of 17 injections over a 3-year
period.
This “reinforced” protocol was also used in the IVAN16 study (a prospective, multi-centre, randomised
study of 610 patients), which found that “reinforced” PRN was as effective as monthly treatment, but
with an average of 7 injections per year (versus 12 for the monthly protocol).
This regimen involves administering treatment at variable intervals depending on whether or not the
retina appears dry during the patient’s check-up. At each check-up, the patient is automatically retreated
with an anti-VEGF injection, and the “dry” or "not dry” status of the retina determines not whether
retreatment is required, but how long it will be until the patient’s next check-up and treatment. Hence the
name “treat and extend”.
L M M J V S D L M M J V S D L M M J V S D L M M J V S D L M M J V S D L M M J V S D L M M J V S D
IVI
Follow-up
visit
(Visual Acuity, Oct) NR NR P NR
Disease
progression NR: No relapse - P: Disease progression
Fig. 3: Diagram of the “treat and extend” (or “inject and extend”) protocol
The first check-up after the three induction phase IVIs takes place at six weeks (it could also take place at four weeks). If the patient
is dry, another IVI is given and the time until the next check-up and IVI is extended by two weeks, i.e. to eight weeks. However, if
the lesion is active, the IVI is given and the time to the next check-up is shortened to four weeks (or maintained at four weeks) and
will remain at this length until the retina dries out.
This is the treatment regimen required by the 2014 European MA for aflibercept, based on the results of
the VIEW21 studies, which compared aflibercept 2 mg administered every two months after an induction
phase of three monthly injections to ranibizumab 0.5 mg/0.05 ml administered every month. The study
demonstrated the non-inferiority of aflibercept. The administration guidelines for aflibercept 40 mg/ml
were amended in 2018, and fixed bimonthly treatment can now be replaced with a “treat and extend”
regimen after the induction phase of three monthly injections.
A1 A2
B C
For example: a patient is treated with three induction injections then is seen monthly. The first relapse
is observed between four and eight weeks after the third injection. The patient is treated and their next
injections scheduled for six weeks’ away. Treatments with two or three injections—carried out with or
sometimes without a prior OCT examination—are therefore possible. As a result, the treatment becomes
less of a burden for both patients and medical teams, with less of an impact on patient quality of life. This
concept of a relatively fixed relapse pattern, individual to each patient, is demonstrated in Mantel et al.23.
After each round of treatment (two or three injections) based on the identified relapse pattern, a full
functional and anatomical assessment needs to be carried out to check whether the relapse pattern
has changed. If the retina has dried out, the period between treatments can be extended by two weeks.
Conversely, if fluid is found, the reinjection period must be shortened by two weeks. It is important to
check the condition of the contralateral eye at least once every four to six months, and as a matter of
urgency if the patient reports problems.
Check-up 1 2 3 4
visit
IVI 1 2 3
After the first three induction IVIs, an observation phase begins, where the patient is checked monthly, with treatment as needed if
relapse occurs (PRN). This phase lasts six to nine months, during which the patient’s relapse profile is established. At the end of this
observation period, the treatment is individualised based on the patient relapse profile, in anticipation of further relapses.
4. Combined anti-VEGF
and photodynamic therapy
Combined anti-VEGF (ranibizumab) and verteporfin photodynamic therapy (vPDT) has proven effective
at improving visual acuity and reducing the size of polypoidal lesions24. However, the EVEREST study—a
randomised study of 61 patients comparing ranibizumab as monotherapy with combined ranibizumab-
vPDT therapy—was not able to prove the superiority of the combined therapy in terms of improving
visual acuity beyond six months25.
The EVEREST II study—a large, randomised, controlled, multi-centre study conducted in 322 Asian
patients and lasting 24 months—compared the long-term effects of the combined therapy to ranibizumab
as monotherapy in polypoidal choroidal vasculopathy26.
All the patients received three initial monthly injections of ranibizumab and were then monitored as per
the PRN regimen. The patients were randomised to also receive either vPDT or no vPDT (sham-PDT
group). At 12 months, the combined therapy group had an improvement in VA of 8.3 letters versus
5.1 letters for the monotherapy group (p = 0.01), with 4 IVIs versus 7 for the monotherapy group. It would
therefore appear that for polypoidal choroidal vasculopathy which remains active after induction with
three anti-VEGF IVIs, combined therapy is superior to monotherapy26.
Artificial intelligence
and AMD
Daniel Seknazi, Oudy Semoun
1. Definition
Artificial intelligence (AI) is currently one of the most promising and dynamic fields of research in the
medical world. It is a new area of medicine and it is important to understand the new vocabulary and
concepts that come with it.
1) Artificial intelligence
AI refers to all the theories and techniques used to develop programs capable of simulating certain
aspects of human intelligence.
An artificial neural network (Figure 1) is an information system based on a network of units known as
artificial neurons, which are organised in layers. Each neuron receives information through its dendrites
and then activates or inhibits one of the neurons in the next layer via its axon. The message is thus sent
down from the first neural layer into the last.
Artificial neuron
Input Output
Input 1 → Weight 1
Neuron
Input 2 → Weight 2
Activation
or inhibition
of the next
Input 3 → Weight 3 neuron
Machine learning (ML) is one of the techniques used in AI. It involves analysing a set of markers or
characteristics by means of data (e.g. retinal thickness in an OCT image or number of haemorrhages). The
AI algorithms learn to identify and categorise these characteristics based on a set of examples provided
in advance. The efficacy of these traditional models of automatic learning depends primarily on how easy
it is to distinguish between the selected characteristics. In traditional machine learning, the task of the
engineer (or doctor) is therefore to manually define the characteristics specific to the field in question.
Normal
retina
4) Deep learning
Deep learning (DL) is a type of machine learning. DL involves a series of different neural layers. Each
neural layer breaks down the input signal into increasingly abstract data and transfers the information via
its synapse to the next neural layer. The message thus descends through the layers until it reaches the
final neural layer. The network cannot only find and classify data as instructed: it can also decide for itself
what it is looking for and how it wishes to classify the data.
The value of DL lies in its capacity for feedback, which enables it to learn from its mistakes and improve
the sensitivity and specificity of a diagnosis. This feedback mechanism is known as “back propagation”.
The main advantage of DL is therefore that its performance continually improves (Figure 3).
Input Output
Cat
Normal
retina
AMD
Back propagation
2. Application to AMD
1) Diagnosis
- Screening
One of the main uses of AI in a disease such as AMD is to screen for at-risk and affected patients.
Burlina et al.1 compared deep learning with “human” diagnosis using fundus photographs to determine
the presence of AMD. They achieved a sensitivity of 88%, a specificity of 93% and a precision of 90% (in
artificial intelligence, the term “precision” refers to the percentage of correctly categorised images). The
results in the DL group were similar to those obtained in the human group. However, fundus photography
is clearly not the best examination for diagnosing AMD.
- Drusen
One of the first studies to use AI in AMD was conducted by Van Grinsven et al.2,3. The authors used
ML capabilities to identify drusen and reticular pseudodrusen. In their study, ML performed similarly
to humans when detecting drusen. The area under the ROC curve was 0.95 for drusen and 0.94 for
pseudodrusen.
2) Prognosis
3. Future perspectives
AI is a powerful tool that offers infinite possibilities. The role of doctors in the future has been firmly called
into question.
With regard to AMD specifically, the potential usefulness of patient self-monitoring is already being
considered. Devices already exist that allow OCT images to be taken at home12.
Fundus camera software is also available for smartphones13,14.
Combined with intelligence capable of making diagnoses and suggesting treatment, this recent progress
will undoubtedly improve patient care and lead to a new role for the healthcare provider—one that has
yet to be defined.
2. Grinsven MJJP van, Lechanteur YTE, Ven JPH van de, et al. Automatic Drusen Quantification and Risk Assessment of Age-Related Macular
Degeneration on Color Fundus Images. Invest Ophthalmol Vis Sci. 2013;54(4):3019-27.
3. Van Grinsven MJJP, Buitendijk GHS, Brussee C, et al. Automatic Identification of Reticular Pseudo-drusen Using Multimodal Retinal Image
Analysis. Invest Ophthalmol Vis Sci. 2015;56(1):633-9.
4. Srinivasan PP, Kim LA, Mettu PS, et al. Fully automated detection of diabetic macular edema and dry age-related macular degeneration from
optical coherence tomography images. Biomed Opt Express. 2014;5(10):3568.
5. Treder M, Lauermann JL, Eter N. Automated detection of exudative age-related macular degeneration in spectral domain optical coherence
tomography using deep learning. Graefes Arch Clin Exp Ophthalmol. 2018;256(2):259-65.
6. Treder M, Lauermann JL, Eter N. Deep learning-based detection and classification of geographic atrophy using a deep convolutional neural
network classifier. Graefes Arch Clin Exp Ophthalmol. 2018;256(11):2053-60.
7. Venhuizen FG, van Ginneken B, van Asten F, et al. Automated Staging of Age-Related Macular Degeneration Using Optical Coherence
Tomography. Invest Ophthalmol Vis Sci. 2017;58(4):2318-28.
8. Prahs P, Radeck V, Mayer C, et al. OCT-based deep learning algorithm for the evaluation of treatment indication with anti-vascular endothelial
growth factor medications. Graefes Arch Clin Exp Ophthalmol. 2018;256(1):91-8.
9. Schmidt-Erfurth U, Bogunovic H, Sadeghipour A, et al. Machine Learning to Analyze the Prognostic Value of Current Imaging Biomarkers in
Neovascular Age-Related Macular Degeneration. Ophthalmol Retina. 2018;2(1):24-30.
10. Bogunovic H, Waldstein SM, Schlegl T, et al. Prediction of Anti-VEGF Treatment Requirements in Neovascular AMD Using a Machine Learning
Approach. Investig Opthalmology Vis Sci. 2017;58(7):3240.
11. Schmidt-Erfurth U, Waldstein SM, Klimscha S, et al. Prediction of Individual Disease Conversion in Early AMD Using Artificial Intelligence. Investig
Opthalmology Vis Sci. 2018;59(8):3199.
12. Chew EY, Clemons TE, Bressler SB, et al. Randomized Trial of a Home Monitoring System for Early Detection of Choroidal Neovascularization
Home Monitoring of the Eye (HOME) Study. Ophthalmology. 2014;121(2):535-44.
13. Nazari Khanamiri H, Nakatsuka A, El-Annan J. Smartphone Fundus Photography. J Vis Exp JoVE. 2017;(125).
14. Sharma A, Subramaniam SD, Ramachandran KI. Smartphone-based fundus camera device (MII Ret Cam) and technique with ability to image
peripheral retina. Eur J Ophthalmol. 2016;26(2):142-4.
Théa Pharma
12 Rue Louis-Blériot
THE DMLA 0419