Validation of JIA Patient Stratification by Synovial Proteome Profiles in an Independent Cohort

Purpose:Juvenile idiopathic arthritis (JIA) comprises a poorly understood group of chronic, childhood onset, autoimmune diseases with variable clinical presentations, outcomes and therapeutic responses. In a prior study, the synovial fluid (SF) proteome of an oligoarticular patient subgroup was compared to patients who show a spread of disease in the first year to involve new joints, the extended oligoarticular subgroup. Current laboratory tests are unable to flag those patients at a higher risk of disease spread to multiple joints, who could benefit form earlier therapy to prevent joint damage. This study investigated whether profiling of the SF proteome based a difference gel electrophoresis (DIGE) approach could differentiate between subgroups of this chronic disease.Methods:To construct a predictive model, SF samples from 22 JIA patients were analysed: 10 oligo-, 5 extended oligo- and 7 polyarticular disease. Samples from each patient were labeled with Cy dyes and subjected to protein separation by 2-DE. The predictive value of using a panel of SF proteins to stratify patients was tested on an independent pool of 28 JIA patients: 13 oligo-, 5 extended oligo- and 10 polyarticular disease. Protein spots of interest which over expressed beyond a two fold threshold between patient subgroups were identified by MALDI-TOF. Discrimininant analysis was used to build a model to predict patient subtype of the latter cohort.Results:Samespots software analysis of SF gel scans was used to highlight joint-specific proteins which were differentially expressed across disease classifications. Hierarchical clustering based on the expression levels of a previously selected set of 40 proteins matched across the three clinical subgroups segregates the polyarticular from the oligoarticular patients (Fig 1). The constructed statistical model classified each of the independent ‘test’ patient cohort into the respective disease subgroups. Proteolytic fragments of apolipoproteins, complement component C3c, haptoglobin and transferrin were identified (p