Background/Purpose: In metatarsalphalangeal (MTP) joints 2-5 of patients with early rheumatoid arthritis (RA), we aimed to assess the agreement between i) Magnetic Resonance Imaging (MRI) and ultrasound (US) findings of inflammation, and ii) clinical swelling and each of MRI and US findings of inflammation.

Methods:  Participants with early RA (ACR criteria) were recruited. Clinical exam was followed by an MRI scan (1.0 Tesla peripheral MRI) and US scan of MTPs 2-5 of the most symptomatic foot. MRI scans were scored by a blinded radiologist using OMERACT RAMRIS criteria for bone marrow edema (BME) (0-3) and synovitis (0-3). Using OMERACT criteria, US images were assessed for synovial thickening (0-3), synovial flow (power Doppler (PD)) (0-3), and erosions (present/absent) in each MTP joint by a rheumatologist, who also assessed swelling (presence/absence) in each MTP. To compare inflammation on MRI versus US, kappa statistics were determined for: MRI synovitis and US synovial thickening, MRI synovitis and US PD, MRI BME and US synovial thickening, and MRI BME and US PD. Agreements between clinical swelling and each MRI and US feature of inflammation (0 = absent, ≥1 on MRI or US = present) were assessed using kappa statistics. The proportions of imaging-detected inflammation also detected clinically were determined.

Results: The study included 33 women, 6 men; mean age 51.6 years (standard deviation=10.2). Table 1 presents weighted kappa statistics for agreement between MRI and US findings of inflammation. Like agreement between synovitis on MRI and US PD, agreement between MRI detected BME and US PD was fair to good. Table 2 presents agreement between clinical swelling and inflammation on MRI and US. Neither US synovial thickness nor MRI synovitis agreed with clinical swelling. There was fair agreement (kappas 0.2322-0.3664) between clinical swelling and each of US PD and MRI BME, but this was stronger in some joints than others. Table 1: Agreement in Inflammation detected by MRI and Ultrasound

*denotes kappa is significant (p<0.05) Table 2: Agreement in Swelling on Clinical Exam compared to Ultrasound and MRI findings of Inflammation

*denotes kappa is significant (p<0.05) There was fair agreement between ultrasound PD and MRI features of BME and synovitis; and ultrasound synovitis and MRI synovitis and BME, but only for MTP4 and 5. There was similar fair agreement between clinical swelling and imaging features of inflammation (BME and PD), but poor for clinical swelling and imaging measures of swelling.

Conclusion: These findings confirm agreements in the imaging parameters of inflammation and suggest that imaging by US and MRI are better than clinical assessment in determining synovial inflammation.

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/assessing-agreement-among-measures-of-inflammation-detected-on-magnetic-resonance-imaging-ultrasound-and-clinical-findings-in-the-feet-of-patients-with-early-rheumatoid-arthritis/