The natural history of facial features observed in Sanfilippo syndrome (MPS IIIB) using a next generation phenotyping tool

Abstract

Sanfilippo syndrome (MPS III) is a rare lysosomal storage disease resulting from a deficiency in one of four enzymes needed to breakdown the glycosaminoglycan heparan sulfate. Excessive storage of heparan sulfate leads to debilitating, progressive neurodegeneration while systemic disease is relatively less severe. Despite MPS III being the most prevalent form of mucopolysaccharidosis, significant delays occur in diagnosis due to the less pronounced facial and physical features. Several distinct facial features are present in nearly all patients progressive coarsening facies, prominent eyebrows, and frontal bossing. Face2Gene is a freely available tool for medical providers, that uses facial analysis, deep learning, and artificial intelligence to analyze 2D photographs and suggest possible syndrome matches. To establish an accurate composite facial image for which to compare new cases, the system must contain a critical mass of disease specific images. Earlier collaborations between Cure Sanfilippo Foundation, Jonah’s Just Begun Foundation and Face2Gene collected patient entered images and established a composite image for MPS IIIB. Here, we assessed the technology’s capability to characterize the unique facial phenotype of MPS III across different age groups. Through a secure patient portal, families submitted patient photographs across the patients’ lifespan. Images were stratified by age group and comparisons conducted between MPS IIIB age groups, between “normal” subjects and non-MPS IIIB syndromic facies. Statistical analysis of the receiver operating characteristic (ROC) curve was performed to calculate the area under curve (AUC) assessing differences between age groups and build the facial natural history. Face2Gene was able to accurately identify MPS IIIB patient photos in the 1-3 year-old age a critical window for intervention. Development of a facial natural history for MPS III could aid clinicians in identifying affected patients. This study further illustrates the value of patient engagement in the advancement of rare disease diagnosis and characterization.