Aim: to study genotype-phenotype correlations in patients with inherited retinal diseases with mutations in ABCA4 gene in Russian Federation.Patients and methods. 21 patients from Russian population aged from 7 to 51 years old (mean age 20 ± 11 years with best-corrected visual acuity from 0,02 to 0,6 (0,14 ± 0,11) with ABCA4-associated retinopathy, verified by molecular genetics methods. All patients besides standard ophthalmic examination and photodocumentation were performed Spectral-Domain OCT and fundus autofluorescence on Spectralis ®HRA+OCT (Heidelberg Engineering, Germany). Full-field electroretinogram (ERG), 30-Hz flicker ERG and macular chromatic ERG (MERG) to red stimulus were recorded on electroretinographic system MBN (MBN, Russia). (Russia) Molecular genetic studies were performed using Next Generation Sequencing (NGS) and Sandger direct sequencing. Results: In ABCA4-associated Stargardt disease 1 type (STGD1) genotype [p.L541P, p.A1038V] of «frequent» mutations was revealed in 9 patients, in 2 cases in was associated another “frequent” mutation p.G1961E. In 4 patients with genotype [p.L541P, p.A1038V] “severe” phenotype of Stargardt disease was found: with large defect of the ellipsoid zone and large zone of central reduced autofluorescence, severely subnormal macular ERG (MERG) to red stimulus and subnormal 30 Hz flicker and full-field maximal ERG. In one patient with these mutations in homozygous state ABCA4-associated cone-rod dystrophy (CORD3, clinically looking alike secondary retinal dystrophy is diagnosed. In 2 patients with genotype [p.L541P, p.A1038V] and mutation p.G1961E was found mild phenotype. One patient with homozygous mutation p.R653C autosomal recessive ABCA4-associated retinitis pigmentosa (RP19) was diagnosed. Clinical picture and autofluorescence were polymorphic in all patients.Conclusions. Our study with ophthalmological, molecular genetics and instrumental methods widens the spectrum of clinical signs of inherited eye diseases associated with mutations in АВСА4 gene, widens the spectrum mutations in Russian Federation and reveals clinicо-genetic genotype-phenotype correlations.
Reported growth rates (GR) of atrophic lesions in Stargardt disease (STGD1) vary widely. In the present study, we report the longitudinal natural history of patients with confirmed biallelic ABCA4 mutations from five genotype groups: c.6079C>T, c.[2588G>C;5603A>T], c.3113C>T, c.5882G>A and c.5603A>T. Fundus autofluorescence (AF) 30° × 30° images were manually segmented for boundaries of definitely decreased autofluorescence (DDAF). The primary outcome was the effective radius GR across five genotype groups. The age of DDAF formation in each eye was calculated using the x-intercept of the DDAF effective radius against age. Discordance between age at DDAF formation and symptom onset was compared. A total of 75 eyes from 39 STGD1 patients (17 male [44%]; mean ± SD age 45 ± 19 years; range 21–86) were recruited. Patients with c.3113C>T or c.6079C>T had a significantly faster effective radius GR at 0.17 mm/year (95% CI 0.12 to 0.22; p < 0.001 and 0.14 to 0.21; p < 0.001) respectively, as compared to those patients harbouring c.5882G>A at 0.06 mm/year (95% CI 0.03–0.09), respectively. Future clinical trial design should consider the effect of genotype on the effective radius GR and the timing of DDAF formation relative to symptom onset.
(1) Background: Recessive Stargardt disease (STGD1) and multifocal pattern dystrophy simulating Stargardt disease (“pseudo-Stargardt pattern dystrophy”, PSPD) share phenotypic similitudes, leading to a difficult clinical diagnosis. Our aim was to assess whether a deep learning classifier pretrained on fundus autofluorescence (FAF) images can assist in distinguishing ABCA4-related STGD1 from the PRPH2/RDS-related PSPD and to compare the performance with that of retinal specialists. (2) Methods: We trained a convolutional neural network (CNN) using 729 FAF images from normal patients or patients with inherited retinal diseases (IRDs). Transfer learning was then used to update the weights of a ResNet50V2 used to classify the 370 FAF images into STGD1 and PSPD. Retina specialists evaluated the same dataset. The performance of the CNN and that of retina specialists were compared in terms of accuracy, sensitivity, and precision. (3) Results: The CNN accuracy on the test dataset of 111 images was 0.882. The AUROC was 0.890, the precision was 0.883 and the sensitivity was 0.883. The accuracy for retina experts averaged 0.816, whereas for retina fellows it averaged 0.724. (4) Conclusions: This proof-of-concept study demonstrates that, even with small databases, a pretrained CNN is able to distinguish between STGD1 and PSPD with good accuracy.
In the most prevalent retinal diseases, including Stargardt disease and age-related macular degeneration (AMD), byproducts of vitamin A form in the retina abnormally during the vitamin A cycle. Despite evidence of their toxicity, whether these vitamin A cycle byproducts contribute to retinal disease, are symptoms, beneficial, or benign has been debated. We delivered a representative vitamin A byproduct, A2E, to the rat's retina and monitored electrophysiological, histological, proteomic, and transcriptomic changes. We show that the vitamin A cycle byproduct is sufficient alone to damage the RPE, photoreceptor inner and outer segments, and the outer plexiform layer, cause the formation of sub-retinal debris, alter transcription and protein synthesis, and diminish retinal function. The presented data are consistent with the theory that the formation of vitamin A byproducts during the vitamin A cycle is neither benign nor beneficial but may be sufficient alone to cause the most prevalent forms of retinal disease. Retarding the formation of vitamin A byproducts could potentially address the root cause of several retinal diseases to eliminate the threat of irreversible blindness for millions of people.