POS0440 THE ANALYSIS OF THE INFLAMMATORY PROTEOME IN RHEUMATOID ARTHRITIS IDENTIFIES COMMON SIGNATURES ASSOCIATED WITH THE CLINICAL RESPONSE TO DMARDs AND TNFi THERAPIES

Background The clinical outcome of the most common therapeutic options of rheumatoid arthritis (RA) patients, such as conventional disease-modifying antirheumatic drugs (DMARDs) and TNF inhibitors (TNFi) is still unpredictable, since a high percentage of patients do not response to the therapy. Innovative analyses combining high-throughput technologies and thorough clinical assessments are needed to gain insight about the management of this prevalent autoimmune disorder. Objectives To evaluate the systemic inflammatory proteome of RA patients, to identify useful biomarkers associated with distinctive clinical outcomes. Methods Serum samples from 140 subjects, including 40 healthy donors (HC) and 100 RA patients with high activity disease (mean DAS28=4.7), were profiled with the innovative proteomic methodology “proximity extension assay” (Olink) which analyses one panel of 92 pro-inflammatory proteins. Samples from RA active patients included 40 from newly-diagnosed RA patients before taking conventional DMARDs and 60 from biologics-naïve patients (mean disease duration=10 years) before receiving TNFi drugs. Clinical outcomes were evaluated following EULAR criteria after 6 months of treatment and patients were classified as responders or non-responders to the different therapeutic interventions. Unsupervised hierarchical clustering methodologies were applied to identify subgroup pf patients based on the proteomic profiles. Gene ontology enrichment were used to interrogate the biological meaning of the distinctive molecular signatures identified. Results The inflammatory proteome analysis identified 33 proteins differentially expressed and upregulated in RA patients compared with HC including several chemokines (CCL-11, -19, -20, -23, -28; CXCL-10, -11, -9; MCP-1, -3), interleukins (IL-6, -8, -18, -10, -17c), and other relevant proinflammatory mediators (VEGFA, CD40, MMP-1, CSF-1, OPG, FGF23) among others (FDR<0.05). Most of these molecules were associated with disease activity (DAS28) and the autoimmunity profile (Rheumatoid factor and ACPAs antibodies) of RA patients. The unsupervised clustering analysis using the proteomic profile of RA patients before TNFi identified two subgroups of patients. Cluster 1 (C1) was characterised by patients with higher levels of several pro-inflammatory mediators compared with Cluster 2 (C2), where a signature of 16 chemokines was significantly enriched (CCL-3, -4, -10, -11, -20, -23; CX3CL1; CXCL-1, -10, -11, -5, -6, -9; MCP-1, -3, -4). Clinically, 25% of the non-responders’ patients was included in C2, while 75% was located in C1, suggesting that a prominent circulating chemotaxis profile prior therapy is associated with a poor clinical outcome. These data were similarly observed in patients before receiving DMARDs, where a signature of upregulated chemokines and pro-inflammatory mediators characterised a cluster with a high % of non-responder patients. Conclusion A pro-inflammatory signature, where chemokines are predominantly up-regulated in the serum of RA patients before therapy, is associated with a poor clinical outcome. This newly identified signature, which deserves a more in-depth analysis, might be clinically useful guiding precision medicine and novel therapeutic approaches. Acknowledgements Supported by ISCIII (PI21/005991 y RICOR-RD21/0002/0033) co-financed by FEDER, Fundacion Andaluza de Reumatología (FAR) and Consejería de Conocimiento, Investigación y Universidad de la Junta de Andalucía (P20_01367). Disclosure of Interests None declared.