Reviewed by Amitha Domalpally, MD, PhD
Amitha Domalpally, MD, PhD, discussed multimodal assessment of optical coherence tomography (OCT) of geographic atrophy (GA) and its precursors and reported both that multimodal imaging offers an added perspective on progression of age-related macular degeneration (AMD) and that multimodal technology bridges the gap between advanced AMD biomarkers and clinical features.
Dr. Domalpally, from the Wisconsin Reading Center, Madison, discussed this topic at the 21st Annual Angiogenesis, Exudation, and Degeneration 2024 Conference sponsored by Bascom Palmer Eye Institute, Miami.
She and her colleagues focused on a multimodal comparison of the OCT-based biomarkers, ie, nascent GA (nGA), incomplete retinal pigment epithelial and outer retinal atrophy (iRORA), and complete RORA (cRORA), with color photographs and autofluorescence images. Their goal was to develop a translational map to gain an understanding of how advanced imaging biomarkers are manifested in typical diagnostic imaging, she explained.
What was known is that previous research uncovered how GA develops, especially from drusen. When retinal pigment epithelium (RPE) is released, areas of increased pigmentation develop followed by areas of depigmentation. This inevitably leads to GA over time and subsequent enlargement of the area of GA.
In light of this, the researchers wanted to determine where specifically the biomarkers nGA, cRORA, and iRORA enter the picture.
The study population included patients who participated in the Age-Related Eye Disease 2 Study (AREDS2) at the University of Wisconsin, Madison. The imaging used included Heidelberg OCT, autofluorescence, and color fundus photography. nGA, iRORA and cRORA were mapped to infrared images and overlaid on fundus photographs/fundus autofluorescence images.
To accomplish the comparison of the imaging technologies, Dr. Domalpally described that the lateral borders of the OCT biomarkers were identified first and, using custom software, they were mapped onto corresponding registered infrared images. She demonstrated the marked lateral edges of each B-scan for cRORA on an OCT image.
The infrared image with those markings then was registered with color photographs and autofluorescence images. This results in stacked registered images. “The regions of interest on the images are essentially windows into the location of the biomarkers that help us to see what’s happening in autofluorescence and color photographs,” she said.
Among the 100 eyes that were eligible for the study from 69 participants, 73 eyes had intermediate AMD on color fundus photographs. Among these eyes, 26 had nGA lesions, 42 iRORA lesions, and 89 cRORA lesions; 12 eyes had both nGA and iRORA lesions.
The results of the mapping process on autofluorescence images, according to Dr. Domalpally, showed that cRORA is the easiest pathology to interpret. “Almost 100% of the images had hypo-autofluorescence. With nGA and iRORA, about 50% to 60% had hypo-autofluorescence. The rest were split between either hyper- or normal fluorescence,” she reported.
On the color photographs, cRORA again was the easiest, with almost 50% that map to GA. Of the remaining that did not have GA on color photographs, drusen with pigment and pigment only were seen in a handful of images. With nGA and iRORA, no GA was seen on the color photographs. However, with nGA, more images had drusen only. With iRORA, both drusen and drusen with pigment were seen. Fading drusen were seen with all 3 lesion types.
The take-home points of the presentation are
Dr. Domalpally summarized, “Based on the color photographs and trying to map to what is happening with these advanced OCDT biomarkers, it appears that nGA and iRORA are in the earlier part of the spectrum with drusen and pigment changes. cRORA seems to encompass the entire spectrum. Multimodal imaging offers a different perspective of AMD progression and it is important that we develop these advanced AMD biomarkers to derive some meaning regarding the clinical features and for a translational impact.”