In silico functional meta-analysis of 5,962 ABCA4 variants in 3,928 Stargardt disease and cone-rod dystrophy cases

Abstract

Purpose : Variants in ABCA4 are associated with autosomal recessive Stargardt disease (STGD1) and cone-rod dystrophy (arCRD). The clinical outcome depends on the severity of the two variants. To provide an accurate clinical prognosis and to select patients for novel treatments, it is important to assess the functional significance of non-truncating ABCA4 variants. Employing a meta-analysis of the published ABCA4 variants and associated phenotypes, we aimed to functionally classify the variants.

Methods : We collected all ABCA4 variants from 3,928 retinal dystrophy cases published up to 2015 in a Leiden Open Variation Database (www.LOVD.nl/ABCA4). The frequencies of ABCA4 variants were compared between 3,270 Caucasian STGD1/arCRD cases and 33,370 non-Finnish European control individuals. We performed in silico predictions on the functional effects of missense and non-canonical splice site variants. We also analyzed the homozygous occurrence of variants and the age of onset of affected persons. Based on this information and the American College of Medical Genetics and Genomics (ACMG) guidelines, all variants were classified according to their pathogenicity.

Results : 316 of 913 ABCA4 variants, were found to be significantly enriched in Caucasian STGD1/arCRD cases. We predicted 81 missense variants to have a (moderately) severe pathogenic nature and 30 variants to be benign. Assessing the homozygous occurrence of frequent variants in STGD1/arCRD cases based on the allele frequencies in control individuals, confirmed the mild nature of the p.[Gly863Ala, Gly863del] variant and identified 3 additional mild variants (p.Ala1038Val, c.5714+5G>A, p.Arg2030Gln). Based on these data, in silico analyses and the ACMG guidelines, we provide pathogenicity classifications for all variants.

Conclusions : Our meta-analysis and application of the ACMG guidelines allowed us to classify all ABCA4 missense variants on the 5-tier scale from benign to pathogenic and to predict their functional effect. We share these results through the LOVD database thereby facilitating direct variant annotation, e.g. for diagnostic application. The next challenge is to promote ABCA4 variant sharing by diagnostic centers to further enhance the value of the database and to allow classification of variants for which there is currently not sufficient data for reliable classification.

This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.