SYMPOSIUM RECAP

Turning Our Attention to Dry AMD

Learn how spectral domain technology may factor into new drug development and improved outcomes.

By Philip J. Rosenfeld, M.D., Ph.D.

OCT has been an invaluable tool for monitoring patients undergoing anti-VEGF therapies, but once the drugs convert these patients from wet age-related macular degeneration (AMD) to dry AMD, time domain OCT has some limitations. Spectral domain technology provides additional qualitative and quantitative information for monitoring disease progression in the dry AMD patient.

As new drugs for the treatment of dry AMD enter clinical trials, the question arises: How will we apply imaging modalities to monitor changes in drusen, pigmentation, hyperpigmentation, geographic atrophy and nonvascularized pigment epithelium detachments — the manifestations of dry AMD? The answer is that we'll need to image macular disease in a way similar to how we manage glaucoma. We'll need to look for slow, degenerative changes in the macular layers and then choose therapies based on the slowing of disease progression. Several imaging techniques will be used, and spectral domain technology will be at the forefront.

SYMPOSIUM RECAP

By Joel S. Schuman, M.D. Diagnosing Glaucoma Earlier With Spectral OCT Optical coherence tomography (OCT) has become a gold standard in clinical decision-making, and this imaging technology continues to evolve. My colleagues explained how spectral domain OCT is advancing the study and treatment of retinal disease, and I'm going to tell you how this technology can improve the way we diagnose and treat glaucoma. Glaucoma Management Benefits With the increased speed of spectral domain OCT, we can acquire more images more quickly. The volume of data we acquire enables us to perform 3-D imaging. From there, we can section any dataset for review. Further, spectral domain OCT allows us to perform cube scans so that we can examine the macula, the nerve fiber layer (NFL) and the ganglion cell layer — all sites where we expect damage to occur in glaucoma. With time domain OCT, the C-scan provides multiple layers of the retina. Spectral domain OCT allows us to isolate and visualize a single layer of the retina. Furthermore, the image is adjustable so we can select the thickness or placement of the layer according to our needs. The tissue layer image offers a virtual dissection of the retina by extracting our specific layer of interest. For example, the inner retinal complex of the macular NFL, ganglion cell layer and inner plexiform layer could be viewed as a whole or individually. This could provide information with regard to glaucoma pathology, detection and assessment of progression that was previously inaccessible to physicians. Practical Applications for OCT The clinical utility of spectral domain OCT is demonstrated in the case of a 65-year-old woman diagnosed with glaucoma. Her IOP was 19 mm Hg and her vision was good. Using the GDx and the Stratus OCT (Carl Zeiss Meditec, Dublin, Calif.), we could see thinning of the neuroretinal rim superiorly with a corresponding inferior field defect, and we could see abnormalities. This patient was skirting the borderline area, even though the contour of the curve describing her retinal NFL was abnormal. The contour of the curve should have two peaks, but if that curve is flattened, the patient has lost the retinal NFL in the area where you should be able to measure it. The Heidelberg Retina Tomograph (Heidelberg Engineering, Vista, Calif.) also failed to find the area of abnormality. With a spectral domain 3-D scan, specific thinning of the NFL could be seen inferiorly and superiorly. So, we were able to detect details that we couldn't see with the other technology, again demonstrating that the damage was localized to the specific tissue layers affected by glaucoma. Spectral Domain OCT Improves Analysis With spectral domain OCT, we can identify specific tissue layers for examination. With 3-D capabilities and cube scans, we have the ability to easily acquire maps, images and comparative analyses that may help us diagnose glaucoma earlier in the disease process. Joel S. Schuman, M.D., F.A.C.S., is eye and ear professor and chairman of ophthalmology at the Eye and Ear Institute of the University of Pittsburgh School of Medicine and director of UPMC Eye Center.

In this article, I'll review how we may be able to use spectral domain technology to image dry AMD and monitor experimental therapies to improve outcomes.

Quantifying Drusen and Geographic Atrophy

Using spectral domain technology, we can look at cross-sections of the OCT fundus map and get a nice point-to-point correlation, and that helps us better appreciate drusen. Going one step further, we can remove the OCT map from the image and examine each of the drusen individually. Doing so helps us differentiate between the large and small drusen. We even see things that look like drusen on fundus photographs but, on further evaluation using OCT, don't have any substance or volume at all.

To assess the morphology and the change over time of these drusen, we can use segmentation with the spectral domain OCT. We can look at the retinal pigment epithelium (RPE) map and perform point-to-point correlations, seeing what has substance and what does not. With the RPE layer segmentation map, you can view the landscape of drusen and appreciate the 3-D contour of drusen (Figure 1).

Using a two-dimensional retinal map, we might see deformations in retinal thickness overlying the drusen (Figure 2). With the 3-D high-definition thickness map, we can see craters in the landscape view that correlate with retinal thinning wherever there are drusen. Even though I've looked at B-scans with time domain OCT for many years, I never truly appreciated how thin the retina really is overlying the drusen. The maps generated by the Cirrus HD-OCT are remarkably reproducible and detect subtle changes, which is important because dry AMD is a dynamic disease.

When enrolling patients in a clinical trial, we must ask ourselves what is most important: the area of drusen or the volume of drusen. We've compared two fundus images of dry AMD patients showing drusen. If we grade these photographs based only on fundus photography, as we've done in the past, you may see a 9% difference in the area of drusen, but when we use spectral domain OCT to image the drusen, we observe a 170% difference in the volume of drusen. Therefore, from this point forward in monitoring dry AMD, I would prefer to know the volume of drusen rather than the area of drusen.

Key Formula Still in Development

Until now, we've referred to drusen as hard, soft, calcific and cuticular, but we couldn't describe them morphologically. Giovanni Gregori, Ph.D., of Bascom Palmer Eye Institute, is conducting additional research to help us realize the full potential of spectral domain technology and how it can help us to better quantify drusen and geographic atrophy.

Figure 1. This image shows segmentation with spectral domain OCT.

Dr. Gregori normalizes the RPE as if there weren't drusen, which is known as "RPE fit" on the Cirrus HD-OCT. He interpolates the RPE from an area that appears normal, providing a flat bed of RPE. He then subtracts the interpolated RPE from the actual RPE (Figure 3). A future Cirrus release will incorporate Dr. Gregori's process so we'll be able to perform correlation analyses that measure the number, distribution, curvature and shape of the drusen and make point-to-point correlations.

Figure 2. Here, segmentation and retinal thickness maps are viewed simultaneously.

Figure 3. Dr. Gregori's formula uses the RPE difference map correlated with the fundus photo.

In a case with large, soft confluent drusen, for example, we'll be able to superimpose the difference map that Dr. Gregori is developing on top of the OCT fundus map, so we can measure the area and the volume. This will allow us to differentiate between drusen with substance and those without substance.

New Dimension in Monitoring

Spectral domain technology will take monitoring of dry AMD to a whole new level. We'll have quantitative techniques to evaluate cases, and we'll be able to measure and track patients over time.

The potential future capabilities of spectral domain OCT also are extremely important. They'll allow us to refine the entry criteria for clinical trials so that we can select a more uniform population of patients with the same kind of dry macular degeneration.

For me, the Cirrus HD-OCT is the complete package. It gives us everything fundus autofluorescence gives us, plus a 3-D analysis that we've never had before. We'll be able to measure the volume of drusen. We'll have a more complete understanding of disease progression. With these improved capabilities, we can better assess patients in clinical trials, which will aid in determining the effectiveness of new drugs and treatment regimens.

Philip J. Rosenfeld, M.D., Ph.D., is professor of ophthalmology, Bascom Palmer Eye Institute, Miller School of Medicine, University of Miami.