LAUSR

Background The 2010 rheumatoid arthritis (RA) classification criteria has been verified to classify patients early as having RA more efficiently than the 1987 criteria. However, sensitivity of this criteria decreased remarkably in patients whose rheumatoid factor (RF) and anticitrullinated antibodies (ACPA) were both negative. Modern imaging technique including ultrasound (US) and magnetic resonance imaging (MRI) are more sensitive than physical examination for detecting joint inflammation objectively, however, US may offer only slight additional value when assessing patients with positive ACPA and RF. Reliability and value of inflammatory findings detected by US and MRI in seronegative, undifferentiated arthritis (UA) patients are still unclear. Objectives To clarify benefits of US and MRI for predicting futuristic diagnosis of RA in UA patients with neither ACPA nor RF. Methods Consecutive, untreated, early arthritis patients who underwent both US and contrast enhanced MRI of 22 sites including bilateral wrists, MCP and PIP joints were enrolled. Synovitis and tenosynovitis were defined as inflammatory findings of US and MRI. Concordance between swollen joint counts (SJC) by experienced physician, inflammatory findings of US and MRI were assessed. We defined UA as non-fulfilment of the 2010 RA classification criteria and the clinical diagnosis pf RA as the initiation of disease modifying anti-rheumatic drugs. Results Seventy one patients were included in the analysis. Fifty eight (82%) were female, the median age was 63 years old, and the mean symptom duration 3 months. Forty eight (67.6%) did not fulfil the 2010 criteria being defined as UA, among which thirty six were seronegative. In all patients, the concordance of detecting inflammation was quite high between MRI and US (κ: 0.67–0.70). In 36 seronegative UA patients, SJC detected by physical examinations were fewer than US or MRI (2.2 vs 3.9, p<0.05) and discrepancy between clinical SJC and inflammatory findings by US/MRI were more frequent in seronegative UA patients (κ: 0.18) compared to the patients who fulfilled the 2010 RA criteria (κ: 0.32). Although only 4 of the seronegative UA patients would have fulfilled the 2010 criteria if inflammation findings by US/MRI were used instead of physical examinations, 21 seronegative UA were diagnosed with clinical RA (sensitivity 19% and specificity 100%). One or more inflammatory arthritis in wrists and MCP joints detected by US or MRI in seronegative patients significantly predicted the development of RA with a good sensitivity of 62% and specificity of 87% (Odds ratio, 10.5). Conclusions Our study suggested that both US and MRI inflammation findings are reliable with a good concordance and can be useful as predictor for futuristic development of RA in UA patients without ACPA and RF. References [1] Kaneko Y, et al. Rheumatology (Oxford). 2011Jul;50(7):1268–74. [2] Zufferey P, et al. Joint Bone Spine. 84(3): 299–303, 2017 [3] Kawashiri S, et al. Mod Rheumatol (2013) 23:36–43 [4] Hirata A, et al. Arthritis Care Res (Hoboken). 2017Jun;69(6):801–806. Disclosure of Interest None declared