Background/Purpose: Clinical evaluation of synovitis in rheumatoid arthritis (RA) is difficult in obese (O) and overweight patients, due to the fat pad located around the joint, that can over- or under estimate the number of joints regarded as swollen by clinical examination.
Objectives: To compare the level of agreement between the number of joints considered abnormal by Power Doppler ultrasound (PDUS) and clinical examination (swollen joint count (SJC) as component of SDAI) in O (i.e. Body Mass Index (BMI) >30) versus non-obese (NO) (BMI≤30) RA patients.

Methods: RA patients ≥18 years fulfilling 2010 ACR-EULAR criteria, treated with conventional synthetic (cs) or biologic (b) DMARDs were included in the cross-sectional multicenter (13 centers) French observational RABODI study (ClinicalTrials.gov Identifier: NCT03004651). Clinical synovitis was evaluated on 44 joints. ESR and CRP were collected and several composite scores were calculated: SDAI, DAS28, DAS. A standard US examination on 44 joints for the presence of synovial hypertrophy (SH) and PD signal according to the EULAR-OMERACT scoring system was performed in each center by an independent investigator blinded to clinical data. Presence of US synovitis was defined by a SH≥1 and PD signal≥1 on a semi-quantitative scale. Levels of agreement between number of synovitis defined by PDUS and clinical examination were compared in O versus NO patients using Chi2 test, and Kappas (k) and ORs were calculated. A patient was considered “discordant” if ≥1 joint was discordantly classified by PDUS and clinical examination. SDAI was calculated and compared, with SJC defined either by clinical examination or PDUS.

Results: 121 patients were included: mean (SD) age of 58.5 (12.7) years, mean disease duration of 11.1 (9.7) years. 81% were female, 84.3% anti-CCP positive, 63.6% had erosive disease. Mean SDAI was 12,6 (±10,2). 53 (43.8%) had a BMI >30 and 68 (56.2%) ≤30. 59 (48.7%) and 62 (51.2%) had a SDAI≤11 and >11, respectively. 78.5% received csDMARD, 58.7% bDMARD and 40.5% glucocorticoids. The 2 groups were comparable, except for weight (mean (SD) 65.4 (13.5) vs 96.7 (14.7) kg, p< 0.001), some comorbidities (diabetes, asthma and fibromyalgia more frequent in O patients, anemia less frequent), tender joint count (mean 4.04 (±5.23) in NO vs 7.38 (±8.64) in O, p=0.021). Mean number of clinical synovitis was 2.4 (3.3), and 6.7 (±6.3) by PDUS. Levels of agreement between clinical and PDUS findings were comparable in O vs. NO patients regarding SDAI and other scores (Table). Patients with ≥3 discordant joints was numerically higher in O patients compared to NO (26/53 (49.1%) vs 22/68 (32.4%), p=0.062). Results were similar when overweighted patients (BMI >25) were compared with normally weighted patients (BMI≤25) or when SH≥1 was considered in place of PDUS synovitis. At the joint level, discordance was higher in O patients in some specific joints: MCP4 (p=0.057), wrist (p=0.089).

Conclusion: In RA patients, despite a perceived higher difficulty to clinically detect synovitis (SJ) in O patients, the discrepancy between clinically- and PDUS defined synovitis was not significantly higher than in NO patients, and did not impact the extend of the definition of disease activity level.

Table Abstract ACR

« Back to 2019 ACR/ARP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/impact-of-body-mass-index-on-the-agreement-between-ultrasound-and-clinical-assessments-of-disease-activity-in-rheumatoid-arthritis-multicenter-and-cross-sectional-study/