1822: The Circulating Immune Profile of Systemic JIA Patients with HLA-DRB1*15 alleles, Eosinophilia, and Lung Disease

Background/Purpose: We aimed to explore the association between HLA-DRB1*15 alleles, eosinophilia, and lung disease (LD) in patients with systemic juvenile idiopathic arthritis (sJIA) through proteomic profiling of circulating immune proteins.

Methods: Peripheral blood (PB) was obtained from children with sJIA (PRINTO criteria) and healthy controls evaluated at our center. Disease status (active/remission) was defined by the treating clinician and confirmed through chart review based on the presence/absence of fever, rash, arthritis, serositis, hepatosplenomegaly, and/or elevated inflammatory markers. Eosinophilia was defined as an absolute eosinophil count (AEC) > 500/μl. A pulmonologist adjudicated the diagnosis of sJIA-LD. HLA typing was done on a clinical basis or through research testing of samples in our institutional biorepository. A proximity extension assay (PEA, Olink Proteomics) was used to profile 45 immune proteins in plasma. Mann-Whitney U tests corrected for an FDR of 0.1 for our discovery analysis were used to compare groups. Volcano plots were generated with GraphPad Prism while PCA plots were created with Clustvis Web Tool.

Results: Table 1 summarizes study participants. Most sJIA patients treated with IL-1 inhibitors at the time of sampling. As expected, broad differences in protein expression were found between active sJIA patients and controls (Fig1A). Previously reported sJIA biomarkers, CCL7, CXCL10, CXCL11, HGF, IL-6, MMP1, and OSM, were confirmed to be significantly elevated in patients with active disease compared to those in remission (Fig1B). We also identified novel proteins differentially expressed in this comparison, including CCL3, CSF3, and TNFSF12/TWEAK (Fig1B). Next, we evaluated sJIA patients with or without eosinophilia at the time of sampling, stratified by disease activity. By univariate testing, IL-33 was elevated in sJIA patients with vs. without eosinophilia, regardless of disease status (Fig2A). In patients with active sJIA, IL-13 and LTA were increased while CXCL12 and OSM were decreased in those with vs. without eosinophilia (univariate testing) (Fig 2A). Levels of circulating immune proteins were not significantly different in patients with and without HLA-DRB1*15:01 alleles when stratified by disease activity and PCA confirmed this finding (Fig2B). Finally, PCA showed that samples obtained from sJIA patients during an episode of eosinophilia closely associated with samples from patients with sJIA-LD who did not have eosinophilia at the time of testing (Fig 2C).

Conclusion: We identified a network of immune proteins in sJIA linked with eosinophilia and type 2-related cytokines. This includes OSM, an IL-6 family member that induces eotaxin-1 release, in sJIA patients with active disease. The IL-1 cytokine family member IL-33, which induces innate lymphoid 2 cells (ILC2s) to produce IL-4, -5, and -13, was associated with eosinophilia in sJIA patients. HLA status did not significantly impact the profile of circulating immune proteins in this cohort. Our studies suggest that the sJIA cytokine milieu may promote type-2 immunity and these responses are likely important in patients who develop eosinophilia and lung disease.