Revisiting genotype-phenotype correlations in ABCA4-related Stargardt's disease

Abstract

Purpose : The phenotypic information surrounding ABCA4-related Stargardt's disease most commonly available in the literature is the age of onset. Based on that information only, we note that the genotypes where neither ABCA4 allele produces a functional protein result in the early onset of the disease (Fig 1A and 1B). However, to begin to understand the disease progression, the longitudinal data are of particular value, Fig. 1C.

Methods : We have collected the disease progression data from 38 patients diagnosed with ABCA4-related Stargardt's disease. The medical literature search resulted in further 600 genotype-phenotype pairs. Therein the longitudinal data, we note, remains scarce.

Results : Based on our dataset, we propose the model in which the rapid decline in the visual acuity is a consequence of the feedback interplay between the photoreceptors (PRs) and the retinal pigment epithelium (RPE). In a nutshell, we argue that if the RPE declined over time in a simple proportion to the number of its cells under the toxic attack by A2E, the byproduct of dysfunctional ABCA4 in the PRs, the decline would be exponential. However, because the PRs themselves depend on the RPE for their viability, the decline must be faster, leading to super-exponential decline in the RPE and, consequently, in visual acuity (VA). This corresponds well with the clinical observation, Fig 1D.

Conclusions : Our hope is that the model will serve as the starting point for discussion of the Stargardt's disease phenotype as a function of ABCA4 genotype (rather than individual alleles). The hope is also that it will motivate systematic collection and publication of Stargardt’s disease progression data, and further experimental characterization of the disease-related variants needed for the development of detailed models applicable in the clinical setting.

This is a 2021 ARVO Annual Meeting abstract.

 

Figure 1. A) The distribution of the age of onset in our dataset . B) The same as A) for both alleles effectively null. C) VA decline in one of our patients, modelled as a super-exponentially decaying function. D) Elementary mathematical motivation for the model. N(t): number of RPE cells, t: time. α: proportionality constant. When α is itself a function of time (with the proportionality factor β), the N(t) decays faster than the exponential function.