THU0187 Circulating autoantibodies as indicators to predict the clinical response to infliximab in rheumatoid arthritis

Background Approximately a third of Rheumatoid Arthritis (RA) patients treated with tumour necrosis factor (TNF)-α inhibitors such as Infliximab (IFX) fail to respond. This has prompted a widespread interest in the finding of measures for predictors of response to TNFα inhibitors. Objectives To search for serum autoantibodies that aid to identify RA patients most likely to benefit from IFX. Methods We analysed serum of 170 biologic-naive RA patients at baseline assigned to receive IFX plus methotrexate. The serum samples were distributed in 3 independent samples sets that were provided by 3 different sources: 1 discovery sample set (n=24) collected from Hospital Clinico Universitario of Santiago de Compostela (Spain) and 2 validation sample sets collected from Hospital Universitario de A Coruna (Spain), (n=61); and the Swedish Farmacotherapy (SWEFOT) trial (Sweden), (n=85). The European League Against Rheumatism (EULAR) criteria were used to assess the clinical response at six months of follow-up: good response (GR, n=60), moderate (MR, n=60) and non-response (NR, n=50). A suspension bead array platform built on protein fragments within Human Protein Atlas and selected from an initial screening using an array containing 42 000 antigens was employed to identify the IgG and IgA autoantibodies in the discovery sample set and validate the results within the 2 validation sets. Thresholds for autoantibodies were calculated by Receiver Operating Characteristics (ROC) curve analysis performed with SPSS 24. Results Our data revealed a more prevalent IgG reactivity and higher IgG autoantibody levels against the antigen Centromere Protein F (CENPF) in GR when compared with NR, showing an overall reactivity of 31% vs 0%, 45% vs. 26% and 17% vs 4% in the three sample sets analysed respectively. The area under the ROC curve was 0,649 [p-value=0.049; IC 95% (0.510–0.789)]. CENP-F is a proliferation-associated and cell cycle-dependent centromere autoantigen that might be involved in the increased or abnormal cell proliferation that occurs during RA process. Interestingly, our results also showed that IgA autoantibodies levels toward the antigen Solute Carrier family 39 member 2 (SLC39A2), a zinc transporter protein, were decreased in GR when compared with MR in the discovery sample set and this trend was significantly validated (p=0.018) in the SWEFOT cohort. The area under the ROC curve was 0,681 [p-value=0,019; IC95% (0,543–0,818)]. Conclusions We have identified two autoantibodies in RA that associate with IFX response. Our findings could potentially be useful to guide the treatment decisions for IFX and lead to further studies focusing on the role of the autoantibodies against CENPF and SLC39A2 within RA. Disclosure of Interest None declared