Update Of Disease Activity Assessment In Rheumatoid Arthritis: Comparison Between Clinical, Ultrasound And Mri Scores And Introduction Of Volumetric Inflammation Measure Concept

Background: Therapeutic decisions in rheumatoid arthritis (RA) are driven by the estimated grade of disease activity. It is clinically evaluated by composite scores such as DAS28, SDAI and CDAI (1). Ultrasound (US) and MRI imaging of the joints provide further characterization of disease activity. US assessments search for the presence of synovitis, joint effusion and Power Doppler (PD) scored semi-quantitatively from 0-3 on joints of DAS28 or proposed reduced score systems (2). MRI joint imaging is evaluated through RAMRIS (RA MRI Scoring) grading separately with a semi-quantitative score (0-3) synovitis after contrast enhancement (CE-MRI), extension of bone marrow edema and erosions on 7 joints of each hand. In 2010 SAMIS (Simplified RA MRI Score) was proposed for reducing examination time by analyzing the most tender or dominant hand only (3). Despite all efforts for objectively quantifying disease activity, all evaluations depend on the subjective feeling of patient or examiner and on the interpretation of imaging data. Objectives: Comparison of clinical, US and MRI scores and volumetric measure of synovial inflammation by dedicated MRI software. Methods: Methods: Clinical data (CRP, ESR, DAS28, CDAI, SDAI and HAQ) were collected from 32 RA patients. Rheumatologists assessed further the number of joints positive for presence of synovitis, effusion and PD (DAS28 and 12-joint US score). Three radiologists independently performed RAMRIS and SAMIS validation on CE-MRI. Moreover, an software was developed in house to estimate total volume of contrast perfusion in joint spaces of both hands. Correlations were analyzed by Spearmen test. Results: SAMIS and RAMRIS correlated significantly for synovitis, edema and erosions (p<0,0001). Interoperator agreement was nearly perfect (Kendall coefficient=0,99). SAMIS and RAMRIS synovitis correlated significantly with DAS28, SDAI, CDAI, HAQ, PCR and ESR (p<0,05), whereas US scores did neither correlate with MRI nor clinical scores. The measure of “inflammation volume” estimated with the software was assumed to be a proxy of synovitis, but did not correlate with clinical, US and MRI scores. Conclusions: MRI remains the gold standard for estimating disease activity in RA, since it correlates with clinical scores more than US-derived scores. Software evaluation of total contrast perfusion or “synovitis volume” is an objective measurement that might provide an independent variable, but its role has yet to be fully tested.