Objective: Stargardt disease type 1 (STGD1) is a rare, inherited, progressive retinal disorder caused by variants in the ABCA4 gene. It leads to the accumulation of toxic vitamin A metabolites and lipofuscin resulting in progressive loss of central vision due to photoreceptor and retinal pigmented epithelium (RPE) death. The fundus autofluorescence (FAF) imaging is an effective noninvasive way in determining the extent of Stargardt disease and can provide an accurate prognosis for affected patients. ProgStar studies assess the natural history of the progression of atrophy secondary to STGD1. The objective of this analysis was to develop a quantitative model of disease progression using atrophic lesion area data derived from FAF imaging from the prospective ProgStar study.
Methods: The definitely decreased autofluorescence (DDAF) area data collected over 24 months from a total of 313 study eyes of 174 STDG1 participants in the prospective observational ProgStar study dataset was available. Participants were 10 to 69 years of age with a median disease onset age of 20 years. A disease progression model was built using nonlinear mixed-effects modeling software, NONMEM 7.4. DDAF area data was square root transformed for modeling. Categorical and continuous measures for demographics, pathophysiologic and other relevant factors were evaluated for their effect on disease progression. Standard model evaluation techniques were used to assess adequacy and robustness of the model.
Results: A linear model adequately described the DDAF area progression over 24 months observed in ProgStar study. The model estimated DDAF progression slope of 0.04mm2 per month. The slope was comparable between left and right eyes. Baseline DDAF area and multifocal lesions were found to be significant covariates impacting DDAF progression slope such that a 2-fold higher baseline resulted in 30-40% greater progression. The presence of multifocal lesions lead to 80-90% greater progression. The FAF heterogenous background signal (localized low AF signal at the macula surrounded by a heterogeneous background with widespread foci or multiple areas of low AF signal at the posterior pole with a heterogeneous background, with/without foci) was found to be a significant covariate on baseline DDAF area, associated with approximately 2-fold higher baseline DDAF compared to homogeneous background signal (localized low AF signal at the fovea surrounded by homogenous background with/without perifoveal foci). The other evaluated covariates such as sex, age, disease onset age, vitamin A supplements, smoking status, flecks outside of arcade and baseline electroretinography (ERG) did not affect baseline DDAF or progression rate.
Conclusions: The disease progression model provided robust characterization of observed longitudinal DDAF area data from ProgStar study. The modeling framework can be used to inform clinical trial design for evaluation of potential treatments.
Citations: Disclaimer: The source of the data is the Foundation Fighting Blindness ProgStarStudy, but the analyses, content and conclusions presented herein are solely the responsibility of the authors and may not reflect the views of the Foundation Fighting Blindness. 1.Strauss RW, Ho A, Jha A, et al, Progression of Stargardt Disease as Determined by Fundus Autofluorescence Over a 24-Month Period (ProgStar Report No. 17). Am J Ophthalmol. 2023 Jun: 250:157-170.
