THU0296 PROTEIN PROFILING IN INDIVIDUALS BEFORE ONSET OF ANCA-ASSOCIATED VASCULITIS. (13th June 2020)

• Record Type: Journal Article
• Authors: Berglin, E.; Esberg, A.; Dahlqvist, J.; Sjöwall, J.; Lundquist, A.; Lejon, K.; Johansson, I.; Mohammad, A. J.; Rantapää Dahlqvist, S.
• Is Part Of: Annals of the rheumatic diseases. Volume 79(2020)Supplement 1
• Page Start: 376
• Page End: 376
• Publication Date: 2020-06-13
• Subjects: Rheumatism -- Periodicals; 616.723005
• DOI: 10.1136/annrheumdis-2020-eular.3033
• Languages: English
• ISSNs: 0003-4967

Abstract

Abstract : Background: Etiology and pathogenesis of ANCA-associated vasculitis (AAV) is multifactorial and understanding of the processes leading from a healthy immune system to autoimmunity and on to debut of symptoms in AAV is rudimentary. Objectives: To identify inflammatory proteins related to the early processes preceding AAV development, and potential novel biomarkers, using large-scale protein analyses Methods: The Swedish National Patient Register of in-patient carevand the Swedish Cause of Death Register with discharge diagnosis from ICD-9 and-10 for AAV were co-analysed with the registers of 4 different blood biobanks to identify AAV individuals with available samples predating onset of symptom. Of the pre-AAV cases 86 (36 male, 50 female; mean age (SD); 51.9 (16.9) years) were identified with at least one plasma or serum sample (28 plasma, and 100 serum) pre-dating symptom onset (mean (SD); -4.3 (3.1) years), and 14 had 2-3 samples. Serum and plasma control samples matched for sex, age and sampling date were identified (n=198; 82 male, 116 female; mean age (SD); 51.9±15.9 years). The samples were analysed for levels of 92 proteins using proximity extension assay (OLINK inflammation panel, SciLifeLab, Uppsala, Sweden). Data were analysed using routine statistical methods, random forest and Partial Least square-discriminant analysis (PLS-DA). Results: As previously described for the assay significant difference between plasma and serum samples were observed both in pre-AAV individuals and controls. In pre-AAV plasma samples significantly increased concentrations of interleukin (IL)-2, chemokine ligand (CCL)-4, fibroblast growth factor (FGF)21, IL-4 and CCL20 were found closer to symptom onset, (<5 years) than later (> 5 years) and compared with controls. In serum tumor necrosis factor receptor superfamily member (TNFRSF)9, CXCL9, osteoprotegerin and vascular endothelial growth factor-A were significantly increased <5 years before onset vs. later (>5 years) and compared with controls. PLS-DA score scattered plot separated the pre-AAV individuals from healthy controls (R 2 =0.26), with significantly increased levels of CCL23, CXCL5, and matrix metalloproteinases-1 (MMP-1), transforming growth factor -ß, orosomucoid, en-rage (S100A12) and IL-7 and decreased FGF-19 level in serum. Binary logistic regression analyses comparing tertiles for these proteins confirmed significantly increased odds ratios for disease development of CCL23, CXCL5 and MMP-1. The findings were confirmed in random forest analysis where these factors were among the 20 most discriminatory factors between pre-symptomatic AAV and controls. Conclusion: In serum samples collected years before symptom onset of AAV, proteins involved in immune system activation were increased, suggesting that the inflammatory process is initiated long before clinical manifestations of the disease appear. These findings propose the elevated proteins as novel biomarkers for disease progression. References: [1]Watts et al. Ann Rheum Dis 2007;66:222-22 Acknowledgments: Vasculitis Foundation, USA Disclosure of Interests: Ewa Berglin: None declared, Anders Esberg: None declared, Johanna Dahlqvist: None declared, Johanna Sjöwall: None declared, Anders Lundquist: None declared, Kristina Lejon: None declared, Ingegerd Johansson: None declared, Aladdin J Mohammad Speakers bureau: lecture fees from Roche and Elli Lilly Sweden, PI (GiACTA study), Solbritt Rantapää Dahlqvist: None declared