Autofluorescence images, which are obtained through the use of confocal laser scanning ophthalmoscopy (cSLO), are playing an increasing role in the understanding of drusen and age-related macular degeneration (AMD). An example of a typical, unfiltered autofluorescence image in a normal retina can be seen in Figure 1. Autofluorescence (AF) images are derived largely from lipofuscin accumulated in the retinal pigment epithelium (RPE).1,2 It is believed that metabolic activity of the RPE correlates with the content of lipofuscin in RPE cells.3 In humans, lipofuscin accumulates with age within the lysosomal compartment.4 Excessive lipofuscin accumulation in the RPE, as imaged by focally increased autofluorescence (FIAF), has been proposed to be a marker of RPE disease and photoreceptor cell degeneration.5,6 Focally decreased autofluorescence (FDAF), especially over larger areas, is equivalent to geographic atrophy (GA).7 These correlations between macular pathological states and AF imaging indicate that a better understanding and ability to analyze these images can help in the staging and treatment of AMD.
Figure 1. Typical Unfiltered AF Image of a Normal Retina.
AF photo of a 50-year-old woman taken by SLO at a resolution of 512 pixels square. The circle encloses the central foveal region of interest (1500 microns squared or 100 pixels squared). This region is hypofluorescent (darker) in the center and becomes progressively hyperfluorescent (lighter) away from the center.
Development of Automated Techniques for AF Image Analysis
Similar to color fundus photography analysis, the analysis of AF images via manual grading can be cumbersome and variable between graders. As such, improved, automated techniques could allow for more precise and confident measurements of macular drusen and atrophy in the setting of AMD. However, automated analysis in the setting of AF images is complicated by the absorption of the 488-nm laser light used to obtain the images by macular pigment and melanin granules as well as the lipofuscin granules in the RPE. In order to resolve this difficulty, we began by applying the mathematical model that we have used previously in the analysis of color fundus photographs8-10 to normal AF images in order to confirm their geometry.11
Since areas of decreased and increased AF have been associated with atrophy and drusen, respectively,5-7 it is important that such assessments be made against a uniform background, which we were able to do using a 12-zone mathematical model. An example of this model can be seen in Figure 2, in which some scattered drusen peripherally in the 6000-micron zone of interest cannot be analyzed adequately in the first photograph (A), given the variability in the background. However, after leveling of the background (B), a single definition of increased AF can be applied with a reasonable segmentation (C).11
Figure 2. Mathematical Model and Segmentation of a Normal AF Scan
(A) Right eye of a 54-year-old woman showing significant background variability and foveal decreased fluorescence due largely to luteal pigment. (B) Twelve-zone mathematical model of the AF background in (A), presented as a contour graph. Note how the model captures the background variability of the original scan. (C) The image in (A) leveled by subtracting the model in (B). The background of the leveled image is now homogenous. Increased autofluorescence was defined as 2.0 standard deviations above the mean, and can be shown in pink. Comparison of the increased FAF with the original image (A) demonstrates a very reasonable selection. By contrast, the use of any single threshold in the unleveled image (A) to define increased FAF would cause major errors, due to the image variability.
For the additional concern of images in which there are areas of GA, we developed a semi-automated method of GA segmentation in which the health care provider can draw the area of GA rather than to label the area manually, as can be seen in Figure 3. We incorporated this technique in the analysis of serial images from patients by registering the images on computer software (Matlab 7.0; The Mathworks, Inc., Natick, MA).12 In Figure 4, the application of these techniques to AF images from a patient obtained at an initial visit and another visit 3 years afterwards are shown. There is limited overlap of the areas of FIAF over the areas of newly-developed GA.
Figure 3. User-interactive GA Segmentation Tool
On the far left, one can see an AF image with an area of GA. In the center image, using the automated GA segmentation tool, the user demarcates the background and the areas of GA with a few simple lines. On the right, the software finds areas of GA (compare to original AF image on the far left).
Figure 4. Predictive Value of FIAF for Progression of GA
(A) Initial AF image. (B) Segmentation into GA (purple) with border zone outline (white), and increased FAF (pink). The increased FAF was defined as 1.5 standard deviations above the leveled image mean for this illustration. (C) Final AF image, 3 years later. The image is of only fair quality, but the GA is defined adequately. Note how the image has been stretched and rotated by the registration software to align with the initial image in (A). (D) The total GA is composed of new GA (light purple) and original GA (dark purple). The original increased FAF (pink) in (B) is superimposed.
AF Images Correlated with Color Fundus Photographs in AMD Staging
Using our automated methods, we found that the relationships of AF lesions to photographic fundus abnormalities in AMD fall into patterns depending on the disease state: early (large, soft drusen with pigmentary abnormalities), atrophic, or neovascular. We noted that the presence of large, soft drusen without atrophy, regardless of the presence or lack of pigment changes, was associated with areas of increased autofluorescence (FIAF). FIAF was also associated with pigment abnormalities in this subset of patients. However, in patients with GA, FIAF is found mostly adjacent to drusen and areas of GA, rather than co-localizing with those areas. An example of this can be seen in Figure 5, which shows the eye of a patient with just soft drusen (above) as compared to the eye of a patient with some atrophic changes and GA in the fellow eye (below). One can see less co-localization of the areas of increased autofluorescence with the drusen segmentation in the eye with more atrophy. Finally, images with evidence of choroidal neovascularization (CNV) were remarkable for an increased finding of reticular autofluorescence patterns.13
Figure 5. Comparison of Soft Drusen Only to Soft Drusen and GA.
(A-E) An eye with soft drusen only. (F-J) An eye with soft drusen whose fellow eye has GA. (A, F) Fundus photographs were cropped to the 3000-micron region. (B, G) Drusen was segmented (green) using the automated leveling and thresholding software. (C, H) The corresponding AF images, which were registered with their respective fundus photographs. (D, I) Segmentation (pink) of areas of FIAF. (E, J) Areas of co-localization can be seen in black, with more overlap occurring in the eye of the patient with only soft drusen (E) rather than the eye of the patient with GA in the fellow eye and some atrophic changes in the pictured eye (J).
Reticular Pseudodrusen and AMD
Reticular pseudodrusen (RPD), which were first identified by Soubrane in patients with age-related maculopathy (ARM),14 are defined as "drusen that form ill-defined networks of broad, interlacing ribbons" in the Wisconsin Grading System for maculopathy.15 Their presence was linked to CNV by Arnold et al., who found that 66% of the patients they studied either had or later developed CNV in one or both eyes. Based on histopathology, Arnold et al. concluded that RPD were associated with deficiencies in the choroidal vasculature, most pronounced in the larger choroidal veins.16
A recent, comprehensive study by Klein et al. demonstrated that the incidence of RPD increases with age, most commonly occurring in study patients who were 75 years of age or older. They also found an almost 2.5-fold increase in the incidence of RPD in women compared to men.17 Interestingly, there was an increased prevalence of RPD in patients who were heterozygous or homozygous for the CFH gene variant Y402H,17 the same one that has been implicated in the pathogenesis of AMD.18 The study also found increased incidence of visual impairment as well as both GA and CNV in patients with RPD as compared to eyes with soft indistinct drusen, though they did not note a preferential outcome from the two.17 Finally, and most concerning, the study found that RPD were associated with poorer survival in patients, even after controlling for factors such as smoking status, cardiovascular history, and cancer.17 These findings lend greater credence to the idea that these lesions are the result of an inflammatory condition affecting the choroidal circulation in addition to other possible systemic effects, as has been suggested.16,19
Autofluorescence Imaging and RPD
Since reticular findings were first noted on blue channel fundus photographs,14 it has been postulated that there may be significant under-reporting of RPD given the more common use of color fundus photography without the added green or blue channel photos.17 As such, it has become increasingly important to use more than one imaging modality when evaluating patients for these lesions in the setting of AMD.
The use of AF in the imaging of RPD began with the identification of a reticular subtype of AF images characterized as a grouping of ill-defined hypofluorescent lesions against a background of mildly elevated AF.20 These reticular findings on AF were subsequently linked to RPD through the use of our exact image registration techniques, an example of which can be seen in Figure 6, which shows a side-by-side comparison of color fundus and AF images for one of our study patients.13 Through the methods for automated image analysis described earlier, we were able to find a strong correlation between RPD as noted on color fundus photographs of study patients and findings of reticular AF in such patients, with only two patients who had reticular findings on color fundus photos without reticular AF patterns.
Figure 6. Reticular Autofluorescence and Reticular Pseudodrusen
(A, D) Original color photographs of right and left eyes of a patient with unilateral CNV. The reticular pseudodrusen are difficult to appreciate. Optic disc drusen were also present incidentally. (B, E) The photographs were contrast-enhanced for better visualization of the reticular pseudodrusen. They were present in all quadrants (E) and above the optic nerve (B). (C, F) Corresponding AF images. The optic disc drusen showed characteristic hyperautofluorescence. Reticular AF was present above the optic nerve (C) and was also present in all quadrants (F). Hence, in both eyes, the damage demonstrated by reticular AF corresponded very closely to that suggested by the reticular pseudodrusen in the photographs.
1. Delori, F.C. et al. In vivo fluorescence of the ocular fundus exhibits retinal pigment epithelium lipofuscin characteristics. Invest Ophthalmol Vis Sci 1995; 36:718-729.0
2. von Ruckmann, A. et al. In vivo fundus autofluorescence in macular dystrophies. Arch Ophthalmol 1997; 36:718-729.
3. Kitigawa, K. et al. In vivo quantification of autofluorescence in human retinal pigment epithelium. Ophthalmologica 1989; 199:116-121.
4. Weiter, J.J. et al. Retinal pigment epithelial lipofuscin and melanin and choroidal melanin in human eye. Invest Ophthalmol Vis Sci 1986; 27:145-152.
5. Delori, F.C. et al. Autofluorescence distribution associated with drusen in age-related macular degeneration. Invest Ophthalmol Vis Sci 2000; 41:496-504.
6. Dorey, C. et al. Cell loss in the ageing retina: relationship to lipofuscin accumulation and macular degeneration. Invest Ophthalmol Vis Sci 1989; 30:1691-1699.
7. Holz, F.G. et al. Fundus autofluorescence and development of geographic atrophy in age-related macular degeneration. Invest Ophthalmol Vis Sci 2001; 42:1051-1056.
8. Smith, R.T. et al. A method of drusen measurement based on reconstruction of fundus reflectance. Br J Ophthalmol 2005; 89:87-91.
9. Smith, R.T. et al. Automated detection of macular drusen using geometric background leveling and threshold selection. Arch Ophthalmol 2005; 123:200-207.
10. Smith, R.T. et al. A method of drusen measurement based on the geometry of fundus reflectance. BioMed Eng Online 2003; 2:10.
11. Smith, R.T. et al. Autofluorescence characteristics of normal foveas and reconstruction of foveal reflectance. Br J Ophthalmol 2005; 89:87-91.
12. Hwang, J.C. et al. Predictive value of fundus autofluorescence for development of geographic atrophy in age-related macular degeneration. Invest Ophthal Vis Sci 2006; 47:2655-2661.
13. Smith, R.T. et al. Autofluorescence characteristics of early, atrophic, and high-risk fellow eyes in age-related macular degeneration. Invest Ophthalmol Vis Sci 2006; 47(12):5495-5504.
14. Mimoun, G. et al. Macular drusen. J Fr Ophthalmol 1990; 13:511-530.
15. Klein, R. et al. The Wisconsin Age-Related Maculopathy Grading System. Ophthalmology 1991; 98:1128-1134.
16. Arnold, J. et al. Reticular pseudodrusen: a risk factor in age-related maculopathy. Retina 1995; 15:183-191.
17. Klein, R. et al. The epidemiology of retinal reticular drusen. Am J Ophthalmol 2008; 145:317-326.
About our author(s):
R. Theodore Smith, MD, PhD
Associate Professor of Ophthalmology and Biomedical Engineering
New York, NY