SPS9-05 Proteomic Profiling Identifies Autoantibody-Associated Proteomic Signatures and an Injury-Stress Endotype Linked to Poor Outcomes in Myositis-Associated Interstitial Lung Disease

Rationale

Myositis-specific autoantibodies (MSAs) define clinical subgroups in idiopathic inflammatory myopathy-associated interstitial lung disease (IIM-ILD), yet serology alone does not explain heterogeneity in disease severity, progression, or outcomes. Integrating quantitative autoantibody levels with circulating proteomic biomarkers may help delineate shared and antibody-specific biological mechanisms and identify risk-aligned disease states.

Methods

We quantified MSAs using a luminescent immunoprecipitation system and measured targeted serum proteomics (inflammation and organ damage panels) in 226 adults with IIM and 35 matched healthy controls. Patients were categorized by ILD-risk (Jo-1, MDA5, PL-7, PL-12, EJ) and non-ILD-risk (Mi-2, NXP2) autoantibodies. Correlation analyses, pathway enrichment, unsupervised proteomic clustering, and exploratory longitudinal analyses of lung function were performed.

Results

Distinct autoantibody-associated proteomic signatures were observed. Anti-Jo-1 showed the highest CXCL9 expression; anti-MDA5 demonstrated prominent upregulation of ATP6AP2 and PSMA1 with marked downregulation of TNNI3; anti-PL-12 showed the greatest increases in IL-6 and FKBP1B; and anti-PL-7 exhibited pronounced upregulation of CASP8, FES, TIGAR, and EN-RAGE. Across ILD-associated autoantibodies, a shared inflammatory program was evident, characterized by upregulation of LTA4H, CXCL9/10/11, CASP8, and MCP-3, with downregulation of CRH and 4E-BP1. Pathway analyses highlighted IL-6/JAK/STAT3 and interferon-γ signaling coupled with epithelial stress and apoptotic responses. Quantitative autoantibody levels correlated with interferon-related chemokines, pro-fibrotic signaling (LAP-TGF-β1), and metabolic-vascular markers (e.g., NCF3, EGFL7). Unsupervised clustering of the targeted proteome identified four biologically distinct endotypes (Figure 1A), which were characterized by differential protein expression patterns and associated survival curves (Figure 1B). One injury-stress endotype was enriched for epithelial injury, apoptotic signaling, neutrophil oxidative stress, and vascular remodeling proteins and was associated with higher autoantibody levels, lower lung function, and lower survival. In exploratory longitudinal analyses, SERPINA9 emerged as the most consistent circulating protein associated with subsequent lung function decline over time.

Conclusions

Quantitative autoantibody profiling integrated with targeted proteomics reveals both autoantibody-associated heterogeneity and a shared interferon-driven inflammatory program in IIM-ILD. Beyond serologic classification, proteomic clustering identifies a convergent injury-stress endotype linked to impaired lung physiology and poorer survival. These findings support the use of circulating protein signatures to enable biologically grounded risk stratification and autoantibody-aligned precision phenotyping in IIM-ILD.

This abstract is funded by: Mentored Research Fellowship (The Myositis Association); Bench-to-Bedside and Back Program (National Institutes of Health)