abstract colorful image of cells

Mosaic of retinal pigment epithelial cells, visualized with indocyanine green (ICG) and adaptive optics. [Image: Johnny Tam / U.S. National Eye Institute]

Researchers at the U.S. National Eye Institute (NEI), Bethesda, Md., are using in vivo fluorescence ophthalmoscopy and adaptive optics (AO) to capture mosaic patterns created by the retinal pigment epithelium (RPE) in human subjects (JCI Insight, doi: 10.1172/jci.insight.124904). The RPE plays an important role in maintaining the health of photoreceptors in the eye; however, it is typically difficult to image. The new NEI technique employs a clinically approved fluorescent dye called indocyanine green (ICG) to stain RPE cells, and AO to achieve cellular-level resolution for the fluorescent RPE mosaics captured by the ophthalmoscope.

In a proof-of-concept study, the team observed that RPE patterns remained constant in individuals with healthy eyes, but changed over time in people with vision diseases linked to damaged RPEs. The researchers concluded that changes in the normally static fluorescent RPE mosaics could someday be used to track retinal disease onset and progression.

Why look at the RPE?

The RPE is a thin layer of cells below the photoreceptors (that is, rods and cones) on the retina. It delivers nutrients to and removes cellular wastes from the photoreceptors; therefore, the health of the retina is very much linked to the health of the RPE. Unfortunately, the RPE is hard to image because the cells contain varying amounts of light-absorbing pigment and because optical aberrations created by the structure of the eye prevent resolution of the layer on a cellular level.

To solve these problems, Johnny Tam and his NEI colleagues used a fluorescent dye (ICG) to make the highly pigmented RPE cells more visible to their imaging system, and employed AO technology to compensate for optical aberrations created by the eye. After validating their approach using mouse models, the team says that its AO-ICG system is capable of resolving individual RPE cells and capturing the fluorescent mosaic patterns they create.  

Testing the AO-ICG system on human eyes

The AO-ICG system consists of a custom-made multimodal AO retinal imager with a computer-controlled fixation system. Tam and his team say the system has a lateral resolution of 2–3 µm and a 0.23–0.60 mm field of view on the retina.

The researchers first looked at the RPE mosaic patterns of five human subjects with healthy eyes. Six minutes after ICG was injected intravenously, the fluorescence signal from the subjects’ RPE cells became stable and remained visible for hours. The total image-acquisition time for each eye ranged between 1 and 3 hours with built-in breaks for the patient. 

The researchers recorded the RPE mosaic patterns for each eye over a period of days after the initial ICG injection and over a period of 3 to 12 months after subsequent ICG injections. They created an automated algorithm to quantify changes in the RPE patterns over time. Tam’s team found that for these five human subjects, the RPE mosaic patterns remained the same.  

To see if its AO-ICG method could be used to detect damage to the RPE, Tam’s team imaged the eyes of people with late-onset retinal degeneration. For these individuals, the team found that the RPE pattern changed only slightly over time. However, for individuals with Bietti crystalline dystrophy—a disease that causes drastic damage to RPE cells—the AO-ICG method showed extreme changes in the RPE mosaic patterns.

Based on the results of these proof-of-concept studies, Tam and his colleagues believe their AO-ICG imaging technique could be used to help diagnose and monitor the progression of retinal diseases, and to pre-clinically detect cellular-level damage to the RPE by nondestructively charting changes in the fluorescent RPE mosaic over short and long periods of time.