Longitudinal Serum Proteomic Profiles – A Step Closer to Personalized Monitoring in Dermatomyositis

Abstract

Dermatomyositis (DM) is a multisystemic immune mediated disease presenting with heterogeneous clinical features. Disease activity relies on biomarkers such as creatine kinase (CK) which can be unreliable, notably in cases with extra-muscular manifestations. Novel proteomic platforms have the potential to identify inflammatory biomarkers for personalized monitoring. The objective of this study was to identify proteins associated with disease activity in DM by using the Olink Target 96 Inflammation panel. Six DM patients from a multicenter inflammatory myopathy registry with available longitudinal biobanked sera were identified. All patients met the 2017 EULAR/ACR criteria for adult idiopathic inflammatory myopathy. Disease activity was categorized based on the International Myositis Assessment and Clinical Study (IMACS) physician global assessment (PhGA) as low (PhGA 0-3) or moderate/high (PhGA 4-10). The IMACS core set measures used for the 2016 ACR/EULAR Total Improvement Score were extracted. Serum samples were analyzed using proximity extension technology (Olink Proteomics Inc., Watertown, MA), simultaneously targeting 92 proteins involved in inflammatory processes. Results were reported as normalized protein expression (NPX) values (log2 scale) with higher NPX values representing higher protein concentrations. Mean NPX difference (ΔNPX) for each protein comparing moderate/high and low disease activity were calculated using paired t-tests with Benjamini-Hochberg correction for multiple testing. Six female DM patients with ages ranging from 34 to 53 years were included (Table 1). Clinical features at timepoint 1 included rash (n=6), muscle weakness (n=5), interstitial lung disease (n=5), Raynaud’s (n=3), arthritis (n=1) and dysphagia (n=1). CK values ranged from 37-6039 U/L. Autoantibodies included anti-MDA5, -TIF1y, -Mi2, -Ro52, and -Ku. At timepoint 1, 5 patients had moderate/high disease activity and at the timepoint 2, 5 patients had low disease activity. Eleven proteins were significantly upregulated when comparing moderate/high vs low disease activity. The strongest ΔNPX was observed for monocyte chemoattractant proteins, MCP-2 (3.0), MCP-1 (2.4), MCP-4 (2.2) (all adj. p=0.02) and C-X-C motif chemokine 11 (CXCL11, 3.0, adj. p=0.05). Other upregulated proteins included CX3CL1 (1.57), CCL11 (1.42), PD-L1 (1.27), IL-4 (1.1), CD40 (0.92), IL-15RA (0.9) and CSF-1 (0.61) (all adj. p=0.05). Table 1. Demographic and clinical characteristics Using serum inflammatory profiles, we identified proteins upregulated in DM patients with moderate/high disease activity compared to low disease activity in a small exploratory cohort. Those included chemokines involved in monocyte and T-cell recruitment that could represent potential biomarkers for disease activity monitoring. Further studies in larger DM cohorts are warranted to evaluate the role of longitudinal proteomic profiling in personalized disease activity monitoring.