PROJECT SUMMARY / ABSTRACT Loss of eyesight is regarded among the worst possible diseases by most Americans. Despite being rare causes of vision loss, inherited retinal diseases (IRDs) are molecularly simple single-gene defects, and advances in genetics and genomics have raised hopes for the development of gene-based treatments. For most IRDs, the natural history of disease involves progressive vision loss and successful interventions must demonstrate a clinically meaningful slowing of the natural progression. Many good metrics have been developed to measure vision loss but the continued lack of approved IRD treatments strongly suggest better outcome measures are still needed. The long-term objective of the research is to predict retinal locations maximally vulnerable to progression over the next 2 years individualized for each patient across distinct IRDs. Successful predictions of loci of vulnerability will drive reliable measurements of functional and/or anatomical changes over the duration of typical clinical trials. The focus of the current project is on two human IRDs that remain without approved treatments – autosomal dominant retinitis pigmentosa caused by RHO mutations (RHO-ADRP) and Stargardt disease caused by ABCA4 mutations (ABCA4-STGD). Average tendencies for spatio-temporal progression in both diseases are well investigated. However, the reliable and individualized prediction of the photoreceptor locations maximally vulnerable to fast progression remains a major challenge. Current literature and preliminary studies support the hypothesis that retinal cross-sectional structure at each location when considered together with the structure of its immediate neighborhood retains enough information to predict vulnerability to disease progression. The current project will involve a combination of retrospective longitudinal and prospective longitudinal studies that operate on different spatial scales and structure/function dimensions, to test the hypothesis and provide a more complete understanding of the range of photoreceptor vulnerability to disease progression in IRDs. Aim 1 will first use a unique existing data set obtained serially in RHO-ADRP patients and train an artificial intelligence (AI) model to learn input OCT features that correspond to disease progression. Trained AI will be applied to another unique existing data set obtained in ABCA4- STGD patients. With special attention to heterogenous transition zones, retinal locations maximally vulnerable will be mapped and validated against serial data. Aims 2 and 3 will use prospective serial studies in ABCA4- STGD to test predictions directly with en face imaging, ultra-wide angle OCT recordings, microperimetric evaluation of rod- and cone-specific light sensitivities, and novel adaptive-optics OCT imaging of the outer retina. The project should provide novel insight into the interaction of human photoreceptors with their diseased neighbors and allow optimum localization of visual f...