Article Date: 10/1/2012

Imaging Dry AMD

Imaging Dry AMD

Making a case for a well-equipped toolbox.

By Jason S. Slakter, MD

When we discuss age-related macular degeneration (AMD), we arbitrarily divide it into wet and dry. We know wet AMD represents a conversion somewhere along the scope of progression to a more acute process that involves choroidal neovascularization (CNV), exudation, hemorrhage and potentially fibrovascular scarring. When we talk about dry AMD, we tend to think of geographic atrophy or atrophic degeneration, but in the clinic, we need to take a step back and determine where the patient’s disease is in the scope of progression.

Technically, dry AMD begins the first day that any abnormality associated with the disease occurs in the architecture of the macula, such as the formation of drusen, the beginning of retinal pigment epithelium (RPE) degeneration, the accumulation of pigment or the eventual development of true atrophy. In other words, dry AMD runs the gamut of the disease spectrum short of the development of CNV. For this reason, various imaging modalities play a crucial role at various stages of dry AMD.


In the earliest stages of dry AMD, we must confirm and document the presence of the pathology to establish a baseline for the patient. The most important imaging modalities at this stage are color fundus photography and autofluorescence imaging. Both are valuable in documenting the extent of the drusen, pigment or other changes, and to look for evidence of more widespread RPE disease. In particular, autofluorescence will show if there is evidence of more extensive RPE damage or dysfunction than was appreciated on clinical examination. If autofluorescence imaging shows some focal areas of hyperautofluorescence associated with soft drusen but no other abnormal patterns, we can be reassured the patient still has focal disease. Knowing this, we can have a more positive discussion with the patient with respect to prognosis and follow up. Conversely, if autofluorescence imaging shows extensive RPE dysfunction or atrophic changes that could not be appreciated on the color fundus photograph or examination, then our approach would change, particularly for follow-up. In the future, autofluorescence imaging may guide our selection of preventive therapies, once they are developed and become available.

Figure 1. Color photo (Top) of a 92-year-old woman with 20/400 vision shows extensive drusen (soft and calcified) and some generalized RPE color changes in the central macula. It is difficult to determine the location and extent of atrophic RPE changes. Autofluorescence image (Bottom) of the same patient shows clearly delineated and extensive loss of RPE throughout the central macula.

Imaging Studies: Critical Components in Research

Imaging is critical in clinical studies of dry AMD. Investigators must ensure they are enrolling a homogeneous population of patients who do not have exudative disease and who have atrophic lesions with similar rates of progression. In well-controlled, randomized clinical studies, the investigators must be able to determine the likelihood and extent to which the disease is truly being modified by the therapy.

In addition to identifying atrophic changes, autofluorescence imaging also identifies RPE dysfunction and hyperautofluorescent areas that are believed to represent accumulations of lipofuscin, A2E and other byproducts of the visual cycle. Normally, these are processed and eliminated by the RPE cells into the choroidal vasculature and carried away to the body, but they can build up within the RPE cells and cause damage or cell death. As a result, these toxic byproducts can impair vision and lead to atrophic degeneration. Therefore, identifying areas of hyperautofluorescence, particularly the location, extent and pattern of these brighter areas, can help us develop risk profiles for patients.

Alan Bird in England and Frank Holz in Germany have looked extensively at these autofluorescence patterns, which they describe with terms, such as banded, speckled and trickling among others. They found that certain patterns are associated with more rapid progression of atrophic degeneration than others. Although identifying these patterns in clinical practice may not have utility, they can help guide researchers select appropriate patients to enroll in therapeutic studies and even stratify patients within these trials.


When patients are visually symptomatic or manifesting disease progression or abnormalities that may indicate more extensive problems, we rely more heavily on autofluorescence imaging along with angiography and OCT imaging. Photography is still useful for documentation, but as the disease progresses, the photos become less critical, because it is often difficult to determine the extent of atrophy on a fundus photograph. Areas that appear lightly colored or hypopigmented on a photo may, in fact, still be functional and not representative of RPE atrophy. In addition, we may have difficulty determining if atrophy is present simply because of the associated drusen, media opacities or poor color balance. With autofluorescence imaging, everything is, quite literally, black and white. Black areas of hypoautofluorescence truly represent regions of dead or nonfunctional RPE. Autofluorescence imaging clearly localizes and delineates atrophic degeneration and often helps to explain a patient’s symptoms (Figure 1).

For example, we sometimes see patients whose visual acuity is 20/20, but they complain bitterly about their visual function. A careful clinical examination and photograph might be enough for a diagnosis, but when autofluorescence imaging shows a large patch of atrophy next to fixation or, in the worst cases, a tiny area of preserved RPE cells surrounded by a large area of atrophy (Figure 2), we know that, yes, the patient can see an individual letter in a sentence, but everything surrounding it is gone. From a functional point of view, the patient cannot read or track anything on a page. Autofluorescence imaging clearly shows why this is happening.


The definition of dry AMD is the absence of CNV, but how do we prove that? A patient may have visual changes, drusen and pigment changes on clinical examination, and atrophy on both clinical examination and autofluorescence imaging, but there may be areas of irregularity or thickening, as well. We must confirm that the macula is indeed “dry” before deciding what course we will take. We do this primarily with fluorescein angiography.

Because we do not have therapeutics for dry AMD at this time, our imaging studies cannot guide therapy. We use them to confirm the diagnosis, monitor for changes in the fundus to explain visual symptoms, and educate and reassure patients. Most importantly, these studies help us identify patients who do need therapy. Fluorescein angiography’s crucial role is to rule out CNV in the setting of dry AMD, or more importantly, to identify unsuspected CNV that would be amenable to intervention.

Although not in the scope of this article, I must acknowledge the important role of OCT. OCT complements fluorescein angiography as we look for subtle areas of exudation that might not be appreciated on clinical examination. Thus, OCT is essential for ruling out exudative disease and confirming the non-exudative nature of the process. (See also “OCT of the Macula” on page 10).

Figure 2. Color photo (Top) of a patient with 20/25 central vision OD (OS is 20/200) but complaints of significant difficulty with reading and driving. Note focal RPE changes and few yellow drusen-like flecks. Autofluorescence image (Bottom) clearly delineating extent of atrophy that surround the fovea (*), thus explaining the visual symptoms in spite of a measured acuity of 20/25.


One unique area worth mentioning during any discussion of fluorescein angiography is myopic degeneration. Although not considered to be the same disease process as AMD, it still qualifies as an atrophic degeneration.

Myopic degeneration can occur at any age, but as patients age, the degeneration worsens. Patients with myopia are difficult to study with OCT. One reason is because of the shape of the myopic eye. If there is an extensive staphyloma, particularly in the posterior pole, the quality of the OCT image may be poor, or at least difficult to interpret. Therefore, conversion from a dry atrophic condition to exudative disease in these patients may be difficult to detect on OCT. Often, the neovascular tissue is not associated with extensive exudation or hemorrhage. Instead, there may be minimal thickening and little in the way of intraretinal or subretinal fluid, which can be missed on an OCT scan. In these cases, we rely on fluorescein angiography to identify the active neovascular tissue, which is often easily seen even against the backdrop of the RPE atrophy commonly seen in these patients (Figure 3, next page).

When I have a patient with myopic degeneration who has had a change in vision, new distortion, or in whom I see something suspicious on the clinical examination, I rely very heavily on fluorescein angiography to determine if the patient has developed active CNV. The angiogram is absolutely critical in identifying exudative disease in this population.


The most recent advance in fluorescein angiography is the advent of widefield imaging. Although its utility is limited for dry AMD, having a widefield image of the fundus is helpful for documenting more extensive, widespread disease. When therapies become available for dry AMD, then this widefield imaging may be more important, because we will want to look globally at the back of the eye. Certainly, for retinal vascular disease, where there’s a significant component of the pathology in the periphery, wide-field imaging is important.

Autofluorescence imaging is a relatively new and evolving technology. A useful feature of some of the newer systems is the combination of autofluorescence, fundus photography, angiography and SD-OCT all in one unit. This enables us to do point-to-point comparisons of what we are seeing with each imaging modality.

Hyperspectral imaging is an emerging technology that localizes structures and biomolecules in the retina using their specific spectral signatures. With this technology, researchers can determine the concentration of certain chemicals within the macula and use that data to determine the macular pigment level. This information could potentially be used in clinical trials to study treatment effects on molecular chemistry in the macula and perhaps in the future to help guide some forms of therapy. Hyperspectral imaging is currently not widely adapted for clinical use, but it does potentially have relevance for dry AMD if it helps identify modifiable risk factors or features in the back of the eye that are associated with progression.


We are at a time now in dry AMD similar to where we were in the 1980s with wet AMD. We are starting to understand the disease, skirting around the edges of trying to find therapies that work but not at a point where we can truly modify the disease.

If there is a caveat for clinicians today, it is that we must carefully analyze all imaging studies, including fundus photos, fluorescein angiograms and OCT, to help determine if what we see represents exudative or nonexudative disease. What looks like some hyperfluorescent drusen on an angiogram, for example, may actually harbor occult CNV, which OCT will pick up. By the same token, if an area is suspicious for exudative disease on the angiogram, you should use OCT for confirmation.

As with everything in medicine, particularly in retina, there is never one test or one approach for everything. If you have a tool chest and you’re building something, you certainly don’t

always use a hammer. You probably want a screwdriver sometimes and maybe you want both. It’s the same thing in retina. Each of our imaging modalities offers a different view of the retina, a different piece of information. Depending on the situation, you may want to utilize many or all of those modalities to fully evaluate the status of an individual patient and determine how you want to proceed in his management.

Dr. Slakter is a partner at Vitreous-Retina-Macula Consultants of New York and a clinical professor at New York University School of Medicine. He is the founding editor of Retinal Physician.

Retinal Physician, Issue: October 2012, page(s): 5 - 8