Abstract
Purpose :
The retinal pigment epithelium (RPE) is thought to be functionally different in different regions of the eye, implicating its role in regional retinal diseases like age-related macular degeneration (AMD). Recently, we completed a morphometric map of human RPE and defined five concentric rings containing subpopulations with varying cell sizes. Going from the macula to the periphery, these subpopulations were named P1 to P5. Using this map, this study aims to determine if any subpopulation had a higher prevalence of lesions in non-AMD and AMD eyes and to use changes of cell morphometry to predict the initiation of AMD.
Methods :
Flatmounts of 17 non-AMD donor eyes and 5 AMD eyes were stained and analyzed with a machine learning based software created to quantify RPE morphometrics. After identifying micro lesions, 3 RPE-retina flatmounts were optically cleared with BABB and stained for RPE borders and photoreceptors.
Results :
RPE cells in the lesioned area exhibited weak phalloidin staining, elongated borders, and a larger area. Analysis of lesion prevalence in each subpopulation indicated that P5 (far-periphery) had significantly more lesioned area than P3 (mid-periphery) in non-AMD eyes (24% vs 7%). AMD eyes showed the same pattern, albeit without statistical significance. To understand RPE morphometric changes with age, we divided the eyes into age groups (<65, 65-79, 80-100, and AMD 80-100). We found that, compared to younger donors, the 80-100 age group of non-AMD eyes had more elongated and less hexagonal cells in P1 (macula). AMD 80-100 data presented with more variability but generally showed worse aspect ratio, hexagonality, and an irregular number of neighbors, indicating abnormalities in the RPE.
Conclusions :
The findings show that far peripheral RPE presents with more risk for developing sub-RPE deposits as compared to mid-peripheral RPE in non-AMD donors, highlighting the importance of imaging peripheral retina. Changes in RPE morphometry with age are more evident in the macular region and advances in the resolution of adaptive optics could allow analysis of RPE morphometrics in real time and could be a predictive tool for clinicians.
This abstract was presented at the 2022 ARVO Annual Meeting, held in Denver, CO, May 1-4, 2022, and virtually.