OP0028 ADVANCED PROTEOMIC ANALYSIS IN SJÖGREN’S DISEASE: INSIGHTS INTO PATHOPHYSIOLOGY AND TREATMENT RESPONSE FROM THE RepurpSS-I TRIAL

Background: The combination of leflunomide and hydroxychloroquine (LEF-HCQ) has emerged as a potential therapy for Sjögren's disease. Given the variability in patient responses, comprehending the action mechanisms of this treatment and the complex pathophysiology of the disease is vital for developing personalized treatment strategies. Objectives: To examine large-scale serum proteomic profiles in healthy individuals versus patients with primary Sjögren's disease (pSD) undergoing LEF-HCQ treatment, and to delineate the dissimilarities between healthy controls (HCs), non-responders (NR), and responders (R), identifying biomolecular pathways and predictive biomarkers. Methods: Serum samples from 28 pSD patients (10 NR, 11 R, 7 placebo) were collected at baseline and after 24 weeks of treatment (clinical endpoint). For comparison, 22 HCs were included. The Olink platform was used to quantify 2926 proteins, with a focus on areas like inflammation, oncology, neurology, and cardiometabolism. Data processing involved the use of UMAP for correlation matrix visualization and HDBSCAN for clustering. Spearman's correlation was employed to evaluate associations with various clinical indicators (including ESSDAI, ESSPRI, UWS, Schirmer, sIgG, RF, STAR, CRESS). Results: Our analysis identified six distinct clusters of proteins, highlighting differences between HC and pSD patients but also between non-responders (NR) and responders (R). Remarkably, in five of these six clusters, NRs showed characteristics that were intermediate to those of HCs and Rs. All patient groups, in comparison to HCs, demonstrated a pronounced pro-inflammatory profile, as evident in clusters 1 and 2. This was characterized by significantly higher levels of chemokines (e.g., CXCL9/10/11/13), cytokines (e.g., IL10, IL12B, TNFSF8), and soluble cell surface markers (e.g., B2M, CD80, FCRL5), with these elevations being more notable in Rs than in NRs. Cluster 3, which was primarily composed of soluble transcription factors (e.g., BACH1, STAT2, STAT5B, FOXO1), showed predominant expression in NRs. Compared to HCs, pSD patients showed lower expression in clusters associated with proteins involved in neutrophil degranulation (cluster 4), complement signaling (cluster 6), and various cellular processes including vesicular transport, apoptosis, and the MAPK signaling pathway (cluster 5). Intriguingly, certain proteins within cluster 1 were correlated with clinical parameters such as the ESSDAI biological score, sIgG, ESR, RF, and IFN score. Furthermore, cluster 4 was linked with TRIM21/SSA expression, suggesting an association between neutrophil activity and autoantibody production. LEF-HCQ upregulated 40 and downregulated 192 proteins across all patient groups, with changes varying between NR and R groups. Particularly, proteins that were more abundantly expressed at baseline in the R group underwent more significant downregulation following 24 weeks of treatment. Notably, baseline levels of certain proteins, such as IFNAR1, strongly correlated with treatment response scores STAR and CRESS (r = -0.74 and r = -0.61, respectively). Furthermore, IFNAR1 levels over time correlated with STAR and CRESS scores (r = 0.66 and r = 0.59, respectively), suggesting that a greater decrease in IFNAR1 levels over time is indicative of a more favorable treatment response. Conclusion: This study highlights the complex proteomic profile of Sjögren's disease and the varying effects of LEF-HCQ treatment. It underscores the molecular diversity and the role of proteomic profiling in developing personalized treatments. The prominence of specific proteins, particularly STAT in non-responders, suggests new targeted treatments. Moreover, the link between proteins like IFNAR1 and treatment response highlights their potential as predictive biomarkers, aiding in understanding Sjögren's pathophysiology and improving treatment strategies. REFERENCES: NIL. Acknowledgements: NIL. Disclosure of Interests: Valentin Baloche: None declared, Wing-Yi Wong: None declared, Arno Concepcion: None declared, Marie-Claude Gaudreau Immunologist and Scientific Leader (AbbVie), Thierry Sornasse Senior Clinical Scientist (AbbVie), Michelle D Catalina Immunologist and Scientific Leader (AbbVie), Helen Leavis: None declared, Joel van Roon: None declared.