Pos0586 Anti-Type Ii Collagen Antibodies Associate With High Levels of Inflammation at Diagnosis but With Good Clinical Outcome – A Study Based on the Norwegian Early Rheumatoid Arthritis Arctic Cohort

Background: Collagen type II (CII) is the major protein in hyaline cartilage. We have previously detected anti-CII autoantibodies in a small subgroup (7-9%) of rheumatoid arthritis (RA) patients at the time of diagnosis (1,2). After forming immune complexes with surface-bound CII in vitro, anti-CII induce cytokines and chemokines via FcgRIIa- and TLR4-dependent mechanisms (3, 4). In our previous studies, the levels of anti-CII antibodies dropped during the first months after RA diagnosis, in parallel to declines in cytokine induction in vitro and declines in levels of CRP and ESR in vivo, arguing that anti-CII may be functionally active in vivo (1). We therefore hypothesize that anti-CII antibodies drive and identify an acute onset rheumatoid arthritis phenotype, and prognosticate a favourable clinical outcome. This has previously been shown to be the case in two Swedish RA cohorts (1,2) but not yet investigated in RA cohorts from other countries. Objectives: To extend our two Swedish clinical investigations on anti-CII antibodies (1, 2) to other early RA cohorts. Methods: 221 DMARD-naïve RA patients fulfilling the 2010 ACR/EULAR classification criteria with less than 2 years of symptom duration from the Norwegian ARCTIC (NCT01205854) strategy trial (5) were included between 2010-2013. All patients were treated with the same DMARD escalation strategy, with 13 visits during 2 years (5). Patients were followed with ultrasound (grey scale and power Doppler), DAS28 and its components, and markers of systemic inflammation including CRP, ESR and plasma calprotectin, as well as health-related quality of life measured with EQ-5D. IgG antibodies against native human CII were investigated with ELISA performed at Uppsala University in samples obtained at baseline and after 1, 2, 3, 6, 12, 16 and 24 months. The cut-off for positive reactions was set at the 95th percentile among population controls, in agreement with our previous publications (1, 2). The study was performed as part of the ScandRA consortium aiming at biomarkers for precision medicine in RA. Results: Mean age was 54 years, and 61% were females. Mean DAS28 at inclusion was 4.4. Anti-CII antibodies were found in 35/221 (16%) of the patients; without significant clinical differences compared to anti-CII negative patients. Most of the anti-CII positive patients remained positive during follow-up, but for five of the 35 initially anti-CII positive patients the anti-CII levels decreased > 50% during the first three months (denoted "anti-CII pos" in figures). These five patients differed from all other patients at baseline, showing higher CRP (median 54 vs. 7 mg/L; p=0.0026), calprotectin (196 vs. 14 AU/mL; p=0.01), swollen joint counts (17 vs. 6; p=0.0009), sum scores for ultra sound power Dopper (31 vs. 6; p=0.0005) and grey scale (54 vs. 17; p=0.0008). They also showed higher SDAI (51.1 vs. 22.1; p=0.001) and DAS28 (5.81 vs. 4.41; p=0.02) at inclusion. During the two-year follow-up, the differences between the groups disappeared, and inversed to a trend towards significantly lower CRP, calprotectin, SDAI and DAS28 among the five initially anti-CII positive patients (Figure 1). The initial differences in swollen joint counts and ultrasound measures disappeared (Figure 2), and the initially anti-CII positive patients ended up with significantly higher scores for EQ-5D (Figure 2). Conclusion: Anti-CII antibodies can be detected in early RA. RA patients with initially elevated anti-CII levels that decline during the first three months have an inflammatory presentation followed by a benign course during the first two years. Although the fraction of anti-CII positive RA patients is low, the corresponding clinical phenotype is strong and penetrant. Measurement of anti-CII may be a clinically useful biomarker to distinguish early RA patients with initially high disease activity but with a good long-term prognosis following optimized medical treatment. REFERENCES: 1 Mullazehi M et al. Ann Rheum Dis 2007;66:537-41 2 Manivel VA et al. Ann Rheum Dis 76(9):1529-1536 3 Mullazehi M et al. Arthritis Rheum 2006; 54(6):1759-71 4 Manivel VA et al Eur J Immunol 2016; 46(12):2822-2834 5 Haavardsholm EA et al. BMJ 2016;354:i4205 Acknowledgements: The study was performed as part of the ScandRA consortium aiming at biomarkers for precision medicine in RA. Disclosure of Interests: Johan Rönnelid Renumeration for lectures given for Thermo Fisher Scientific, Scientific Advisory Board member for Thermo Fisher Scientific and Inova/Werfen., Christine Möller Westerberg: None declared, Anna Svanqvist: None declared, Hilde Berner Hammer AbbVie, Lilly, Novartis, UCB, Joe Sexton: None declared, Linda Mathsson Thermo Fisher Scientific, Isabel Gehring Thermo Fisher Scientific, Johan Askling Agreements between Karolinska Institutet (with JA as PI) and Abbvie, BMS, Eli Lilly, Galapagos, MSD, Pfizer, Roche, Samsung Bioepis, Sanofi, mainly for the national safety monitoring of rheumatology immunomodulators in Sweden (ARTIS), Helga Westerlind: None declared, Siri Lillegraven: None declared, Espen A Haavardsholm: None declared.