NIH Study Uncovers New Insights into Stargardt Disease Progression Using AI

NIH Study Uncovers New Insights into Stargardt Disease Progression Using AI

For individuals and families affected by inherited retinal diseases (IRDs) like Stargardt disease, understanding the progression of vision loss is crucial. A recent study from the National Eye Institute (NEI), part of the National Institutes of Health (NIH), has brought significant advancements in classifying vision loss and retinal changes in Stargardt disease. This research, published in JCI Insight on January 25, 2022, offers new hope for better monitoring, understanding, and ultimately, treating this challenging condition.

Stargardt disease, the most common form of which is ABCA4-associated retinopathy, is an autosomal-recessive genetic disorder affecting approximately 1 in 9,000 people. It is caused by variants in the ABCA4 gene, which provides instructions for a protein vital to light-sensing photoreceptor cells in the retina. Individuals develop Stargardt disease when they inherit two mutated copies of ABCA4, one from each parent. While carriers with only one mutated copy do not develop the disease, the variability in age of onset and progression among those with ABCA4 gene variants can be wide-ranging.

AI-Powered Method for Tracking Disease Progression

NEI researchers developed and validated an artificial-intelligence-based method to evaluate Stargardt patients. This innovative approach quantifies the loss of light-sensing retinal cells, providing valuable information for monitoring patients, understanding the genetic causes, and developing therapies. The study followed 66 Stargardt patients (132 eyes) over five years, utilizing spectral-domain optical coherence tomography (SD-OCT), a retinal imaging technology that uses light to image layers of the retina.

Many of the SD-OCT scans were analyzed using deep learning, a type of artificial intelligence where imaging data is fed into an algorithm that learns to detect patterns. Six retinal layers were segmented and analyzed for changes in thickness. The researchers found that the loss of the ellipsoid zone (indicating severe photoreceptor degeneration) and thinning of the outer nuclear layer (indicating subtle photoreceptor degeneration) followed a predictable pattern.

Implications for Patients and Future Therapies

According to Michael F. Chiang, M.D., director of the NEI, these results provide a framework to evaluate Stargardt disease progression, which will help manage the significant patient-to-patient variability and facilitate therapeutic trials. This predictability allowed researchers to classify the severity of 31 different ABCA4 variants. Importantly, the study also revealed that photoreceptor degeneration was not confined to the area of ellipsoid zone loss.

This new AI-driven method holds promise for standardizing how Stargardt disease is assessed, which is a critical step for clinical trials aiming to test new treatments. By providing a more precise way to measure disease progression, this research could accelerate the development of much-needed therapies for the IRD community.