RPE Loss is Correlated with Baseline Circumference in Atrophic AMD, Stargardt Disease and Choroideremia

Abstract

Purpose : Clinical trials are limited by relatively short duration and loss to follow-up, making it virtually impossible for a single trial to determine the natural history of atrophy enlargement over several decades. Herein we apply a novel mathematical approach to infer the decades-long natural history of atrophy expansion secondary to age-related macular degeneration (AMD), Stargardt disease (STGD), and choroideremia (CHM).

Methods : We extracted study-level data in AMD (2942 eyes), and individual-level data in STGD (228 eyes) and CHM (68 eyes) regarding retinal pigment epithelium (RPE) atrophy enlargement from publications in 6 databases. We analyzed the atrophy progression by testing 3 models: the area linear model, radius linear model, and area exponential model, in which the lesion area, radius, or Log area changes linearly over time. To correct for differences in patients’ entry times into each study, we added a horizontal translation factor to shift each dataset one month at a time until r² was maximized for the cumulative trendline.

Results :
After the correction for entry times, the log residual RPE area in eyes with CHM declined linearly over 60 years with a rate of 0.049 log(area)/year (r² = 0.997; compare Fig 1B with 1C). The predicted age at 20° residual RPE diameter was 27.0±11.8 years. Using the same method, we found that the effective radius of atrophy in AMD and STGD enlarged linearly over at least 10 and 40 years, respectively. The growth rate is 56.7% higher in AMD (0.163 mm/year; r²=0.986) than STGD (0.104 mm/year; r²=0.927). Interestingly, based on the derived models, the loss rates of RPE cells were strongly correlated with the baseline circumference in all 3 diseases (Fig 2), suggesting that a process at the border of atrophy might drive RPE atrophy progression.

Conclusions : The RPE atrophy enlargement in AMD, STGD, and CHM follows a 1-stage model over decades. Log area of residual RPE can serve as a reliable outcome measure in trials for CHM, while the effective radius of RPE atrophy can serve as a reliable endpoint in trials for AMD or STGD.

This is a 2020 ARVO Annual Meeting abstract.