FRI0581 Ultrasonography in the detection of joint destruction in ra patients: a comparison with conventional radiography

Background RA evolves erosive polyarthritis resulting in destructive changes in the joints. Ultrasonography (US) is used in current practice as early diagnostic modality for identification of structural damage to the articular surfaces. Objectives To compare US and radiographic assessment of hands and feet joint destruction in RA patients, and to evaluate the contribution of early US detection of damage into long-term RA outcomes. Methods 75 patients with RA, mean age 54.0,44,0; 62,0 disease duration 74; 20 months, were treated with MTX and biologics according to Treat-To-Target concept. Hands and feet US were analysed before initiation of treatment and in 3, 6, 9 and 12 months after. Deepening of the bony contour >2 mm in width and >1 mm in depth, visualised in 2 orthogonal planes, was considered as the key US sign of destructive changes (erosions) according to OMERACT criteria. A binary scoring system (presence/absence of erosions) of the joints examined was used. Radiographs were obtained at baseline, at 12 month and 4 years, radiographic changes were assessed using Sharp/van der Heijde modified scoring method. Radiographic progression was documented based on Sharp/Van der Heijde modified score changes during the follow up. Results There was a significant correlation between the counts of joints with erosions obtained with two diagnostic methods – US and radiography. This correlation was moderate before initiation of therapy (r=0,37 ; p=0,0008), and weak – after 12 month follow up (r=0,28 p=0,016). During one year US showed increase in the count of joints with erosions (from 1 [0; 2] to 21; 3) while radiography did not show any significant change (from 0 [0; 1] to 0 [0; 1]). Bland-Altman analysis showed statistical agreement between the results obtained by two methods. Mean difference between the two modalities before initiation of treatment was −0,42 (95% CI −0,68; −0,16), and at 12 month follow up −1,16 (95% CI −1,52; −0,80), which is comparable with actual values. We identified the relationship between the difference in variables and the count of affected joints before and 12 month after initiation of treatment (r= −0,35, r= −0,53 respectively). 8% of variables were outside 2 standard deviations at baseline, and 4% – at 12 month. Logistic regression analysis showed no relationship between annual radiological progression and US diagnosed increasing count of joints with erosions at 3, 6 and 9 months follow up. However, dynamic radiographs assessment at 4 years revealed a correlation with US diagnosed count of joints with erosions at 6 month and 9 month follow up (r=0.24, p=0.03; r=0.24, p=0.04, respectively). Quality indicator of US diagnosed count of erosions at 6 month follow up: OR=2,8 95% CI 1.05–7,5, p=0.037, with 71% sensitivity and 54% specificity; Quality indicator of US diagnosed count of erosions at 9 month follow up: OR=2,73 95% CI 1.02–7,27, p=0.041, with 64% sensitivity and 61% specificity. Quantitative assessment of the dynamics at these time periods did not show any relationship. Conclusions Therefore, our study confirms the relevance of US in assessment of bone erosions in spite of week agreement with radiography data. We found a prognostic value of US-diagnosed erosions during the first year of follow upfor long-term (4 years) clinical outcomes, and the relevance of control assessment at 6 month after initiation of treatment. Disclosure of Interest None declared