Geographic atrophy is the atrophic, late-stage manifestation of dry age-related macular degeneration in approximately 20 percent of AMD patients.1 GA is defined as a shƒarply delineated round or oval area of depigmentation or absence of RPE, in which choroidal vessels are more visible than the surrounding area and must be at least 125 µm in diameter. These affected areas have no visual function, since loss of the retinal pigment epithelium is associated with depletion of photoreceptors.1,2
Considering that the RPE, choriocapillaries and outer neurosensory retina are the principal target of AMD and that fundus autofluorescence well reflects the metabolic function of these structures, this tool should be considered a valid method of monitoring the disease. FAF is a fast and non-invasive tool, which provides functional images of the fundus and particularly of the photoreceptor/RPE complex by absorbing the stimulated emission of light from naturally occurring fluorophores.3
There are several commercially available devices for obtaining FAF images, two of which are mainly used: commercial scanning laser ophthalmoscopy (cSLO) (Spectralis HRA, Heidelberg Engineering), and a fundus camera-based system (Topcon). As these two tools are based on different principles, the resulting FAF images will not necessarily show the same results. In order to evaluate the clinical significance of differences between these two tools, this next section will compare FAF pictures of two patients with GA in AMD.
FAF in Practice
An 86-year-old woman with a history of atrophic age-related macular degeneration was referred for decreased vision in the right eye over the last six months. The patient complained of difficulties in reading small print. At presentation her visual acuity was 20/60 in the right and 20/25 in the left. By Amsler grid testing, she had a modest amount of distortion bilaterally. On fundus examination of the right eye, the patient presented a relatively large area of multilobular geographic atrophy, sparing the center of the foveal avascular zone (See Figure 1A). Areas of geographic atrophy, largely located superior to the macula were visualized in the left eye (See Figure 1B). To help further evaluate her ocular status, she underwent c-SLO and fundus camera FAF imaging. FAF of the right eye revealed hypoautofluorescent multilobular areas, whichalso appeared to involve the fovea in the c-SLO picture (See Figure 1C), but not with the fundus camera (See Figure 1E). The FAF imaging of the left eye also showed diffuse patches of geographic areas, that appeared to extend into the fovea with the c-SLO (See Figure 1D), but not with the fundus camera (See Figure 1F). Small puntiform zones of suffering RPE or pre-atrophic RPE were evident as hypoautofluorescent spots in the supero-temporal quadrant of the left eye with the fundus camera (See Figure 1F) and not with the c-SLO (See Figure 1D).
The second patient is an 86-year-old woman presenting for a routine eye exam. Her visual acuity was 20/40 in the right eye and 20/30 in the left eye. The fundus examination of the right eye showed extensive areas of GA in the central macula, but sparing of the fovea (See Figure 2A). A diffuse pattern of soft drusen and crystalline drusen was detected supero-temporally. The left eye manifested several patches of GA in the central macula with no involvement of the fovea, and a diffuse pattern of soft drusen in the supero-temporal quadrant (See Figure 2B). FAF imaging revealed the presence of circular hypoautofluorescent areas in the central macula bilaterally, which seemed to involve the fovea with the c-SLO FAF imaging (See Figure 2C). Supero-temporally, hyperautofluorescent zones and hypoautofluorescent spots, corresponding to the crystalline drusens and early RPE dysfunctions respectively, were well visible with the fundus camera bilaterally (See Figures 2E & F), while those lesions were not as well appreciable with the c-SLO (See Figures 2C & D).
Fundus Camera & c-SLO FAF Imaging
The c-SLO FAF seems to over-estimate the presence of atrophy at the fovea, revealing its presence when in fact it is absent. The fundus camera better reveals soft drusen and early dry-age related macular degeneration alterations at the level of the RPE in areas which appear otherwise normal on fundus exam.
The clinical significance of differences in FAF findings detected by these two devices is underlined by the functional and prognostic information associated with FAF alterations. An important role can be ascribed to the macular pigments, which include lutein, zeaxanthin and mesozeaxanthin, which are concentrated in the macula, in particular in the axons of the cone receptors.5 Therefore, the excitation light at 488 nm, used with cSLO, is blocked by macular pigments,6 which have an absorption spectra ranging from 400 nm to 540 nm, while with the fundus camera, the excitation light at 580 nm is not likely to be influenced. As a consequence, the macula pictured by cSLO appears darker in the center of the FAF images, as compared with the fundus camera, masking drusen and simulating the presence of GA.
Moreover, the c-SLO scales the image in grayscale histogram, truncating the black portion of it,4 modifing regions of reduced autofluorescence, such as in GA. This may explain the more clear visibility of early FAF alterations, representing dysfunctioning RPE cells, with the fundus camera imaging, compared with the c-SLO. Indeed, the recognition of abnormal FAF as clinical biomarkers for disease progression is underscored by c-SLO FAF imaging.
There are several reasons why the fundus camera shows soft drusen deposits more clearly than cSLO. First, the two systems use different optics: the cSLO is a confocal system, where images only conjugate on the selected plane in the fundus, and points not lying on this conjugated plane are omitted.7,4 In contrast, the fundus camera is not a confocal system, and it does not omit light coming from adjacent layers. Secondly, the brightest areas of an image might be cut off during image processing with cSLO if these are within zones of markedly increased or decreased FAF.8 Third, drusen consist mainly of lipofuscin and precursors of A2-E, better imaged with the fundus camera.1 The cSLO uses excitation wavelengths, which optimize the detection of lipofuscin, and not of the precursors of A2E. Since the barrier filter of the fundus camera has a bandpass with a longer wavelength than that for cSLO, these signals from the precursors of A2E are more likely to be captured by the fundus camera.9
To minimize interference from fluorophores in the lens, which mainly emits between 510 and 670 nm, Richard Spaide, MD, modified excitation (bandwidth 535 to 580 nm) and emission wavelengths (bandwidth 615 to 715 nm). This drastically has improved the signal-to-noise ratio and image quality, so despite the nuclear opacities of the patients the quality of images is good with both tools.5
Therefore, we can consider FAF a useful examination in routine clinical settings for the evaluation of the functional and prognostic status of GA, allowing clinicians not only to monitor the evolution of the disease, but also to execute clinical interventional trials with "fast progressors." Therefore, retinal specialists should be aware of the characteristics of each of these systems, particularly so as to avoid misinterpretation of FAF findings.
Dr. Bruè is an international fellow at Vitreous Retinal Macula Consultants of New York. Contact her at email@example.com.
1. Schmitz-Valckenberg S, Bindewald-Wittich, Fleckenstein M, et al. Age-Related Macular Degeneration II-Geographic atrophy. Medical Retina. Springer Berlin Heidelberg New York 2007:147-151.
2. Brar M, Kozak I, Cheng L, et al. Correlation between spectral-domain optical coherence tomography and fundus autofluorescence at the margins of geographic atrophy. Am J Ophthalmol 2009;148:439-444.
3. Schmitz-Valckenberg S, Fleckenstein M, Hendrik PN. Fundus autofluorescence and progression of age-related macular degeneration. Surv Ophthalmol 2009;54:96-117.
4. Spaide RF. Autofluorescence imaging with the fundus camera. Marion P. Atlas of Fundus Autofluorescence Imaging. Berlin, Springer 2007; 49-54. 1st ed.
5. Trieschmann M, Spital G, Lommatzsch A, et al. Macular pigment: Quantitative analysis on autofluorescence images. Graefe's Arch Clin Exp Ophthalmol 2003;241:1006-1012.
6. Ermakov IV, McClane RW, Gellerman W, Bernstein PS. Resonant Raman detection of macular pigment levels in the living human retina. Opt Lett 2001; 26:202-204.
7. Schmitz-Valckenberg S, Fitzke FW, Holz FG. Fundus autofluorescence imaging with the Confocal Scanning Laser Ophthalmoscope. Marion P. Atlas of Fundus Autofluorescence Imaging. Berlin, Springer 2007;31-36. 1st ed
8. Schmitz-Valckenberg S, Luong VY, Fitzke FW. How to obtain the optimal fundus autofluorescence images with the confocal scanning laser ophthalmoscope. Marion P. Atlas of Fundus Autofluorescence Imaging. Berlin, Springer 2007; 37-47.1st ed.
9. Bessho K, Gomi F, Harino S, et al. Macular autofluorescence in eyes with cystoid macula edema, detected with 488 nm excitation but not with 580 nm excitation. Graefe's Arch Clin Exp Ophthalmol 2009;247:729-734.