FRI0585 Prevalence of anti-acetylated protein antibodies in inflammatory arthritis, osteoarthritis, connective tissue diseases and its discriminative capacity as diagnostic marker for early rheumatoid arthritis

Background Numerous post-translationally modified proteins have been described as auto-antigens in rheumatoid arthritis (RA) patients. Antibodies (abs) against acetlyated (ac) peptides (AAPA) have recently been reported in RA patients, but not yet been evaluated in other inflammatory and non-inflammatory rheumatologic conditions; therefore their specificity (spec) and sensitivity (sens) remains unclear. Objectives To determine the prevalence of AAPA in RA, healthy subjects and other rheumatic diseases in order to evaluate their diagnostic potential for discriminating RA, healthy and other rheumatic diseases. Methods We obtained serum samples of patients with early untreated RA, established RA (>3 years), osteoarthrits (OA), systemic lupus erythematosus, granulomatosis with polyangiitis (GPA), polymyositis, axial spondyloarthritis, primary Sjogren’s syndrome and healthy subjects. AAPA were measured by ELISA using peptides derived from mutated vimentin (acetylation of lysine or ornithine in position 7 or 2 (inverse peptide), as antigen. Receiver operating characteristics and logistic regression analyses were used to assess the discriminative capacity of AAPA. Results Areas under the curves (AUC) were significant in early RA (eRA; n=120; 75% female, mean disease duration: −0.07±0.51 years, mean symptom duration 1.49±2.01 years) versus healthy subjects for IgG-abs against ac lysine, inverse lysine and ornithine (AUC of 0.666, 0.687, 0.800, respectively). We chose a cutoff of 20 U/ml putting an emphasise on high spec, with balanced sens (ac-lysine: spec: 97.0%; sens: 32.5%;+likelihood ratio (LR) 10.7, CI: 3.4–33.7; ac-inverse-lysine: spec: 80.7%; sens: 42%;+LR 2.2, CI: 1.3–3.6; ac-ornithine: spec: 93.9%; sens: 39.2%;+LR 6.5; CI: 2.9–14.5). Analyses of positivity for multiple ab-reactivity revealed increasing +LR by number of abs, with 100% specifity when all 3 AAPAs are detected (table 1). Testing this cutoff against OA patients showed similar specificities, but with lower +LR (2 AAPA:+LR 3.48, CI: 1.9–6.6). Sens is increased when testing established RA versus healthy controls, with ac-ornithine performing best (ac-lysine: 49.2%, CI: 42.0–56.5; ac-inv-lysine: 35.2%, CI: 28.5–42.4; ac-ornithin: 53.9%, CI: 46.6–61.0) We found that practically only RA patients showed three different AAPA reactivities (in eRA: 39% positive for ac-ornithine abs, 33% for ac-lysine abs, 48% for inverse ac-lysine abs). Polymyositis and GPA patients showed the lowest prevalence of AAPA ( Graph 1A ). Among eRA patients 17% were found to be exclusively positive for AAPA, while 39% were also positive for rheumatoid factor (RF) and anti-citrullinated antibodies (ACPA) (distribution in Graph 1B ). Also in RF- and ACPA- patients the presence of one AAPA identified RA patients vs. healthy subjects with a spec of 77.7% and those with 2 AAPA reactivities with even 97% respectively. Conclusions AAPA are highly prevalent autoantibodies in early RA, closing a further gap of seronegativity, with only 24.6% of early RA patients remaining negative for RF, ACPA or AAPA. In particular, multiple reactivity to AAPA increased the specificity for eRA, also adding diagnostic value beyond RF and ACPA. Disclosure of Interest P. Studenic: None declared, H. Bang Employee of: Orgentc Diagnostika GmbH, A. Alunno: None declared, D. Sieghart: None declared, D. Aletaha: None declared, S. Bluml: None declared, H. Haslacher: None declared, J. Smolen: None declared, G. Steiner: None declared