Abstract
Impaired glycosaminoglycans (GAGs) catabolism may lead to a cluster of rare metabolic and genetic disorders called mucopolysaccharidoses (MPSs). Each subtype is caused by the deficiency of one of the lysosomal hydrolases normally degrading GAGs. Affected tissues accumulate undegraded GAGs in cell lysosomes and in the extracellular matrix, thus leading to the MPS complex clinical phenotype. Although each MPS may present with recognizable signs and symptoms, these may often overlap between subtypes, rendering the diagnosis difficult and delayed. Here, we performed an exploratory analysis to develop a model that predicts MPS subtypes based on UHPLC-MS/MS measurement of a urine free GAG profile (or GAGome). We analyzed the GAGome of 78 subjects (38 MPS, 37 healthy and 3 with other MPS symptom-overlapping disorders) using a standardized kit in a central-blinded laboratory. We observed several MPS subtype-specific GAGome changes. We developed a multivariable penalized Lasso logistic regression model that attained 91.2% balanced accuracy to distinguish MPS type II vs. III vs. any other subtype vs. not MPS, with sensitivity and specificity ranging from 73.3% to 91.7% and from 98.4% to 100%, depending on the predicted subtype. In conclusion, the urine GAGome was revealed to be useful in accurately discriminating the different MPS subtypes with a single UHPLC-MS/MS run and could serve as a reliable diagnostic test for a more rapid MPS biochemical diagnosis.