BACKGROUND: Magnetic resonance imaging (MRI) and ultrasonography (US) are more sensitive than clinical evaluation in assessing inflammation in rheumatoid arthritis (RA). Data is scarce regarding potential link between patient-reported flares and inflammation on imaging. The aim of the study was to explore the pattern and longitudinal associations of inflammatory lesions detected by serial MRI and US in relation to patient-reported flares in patients with RA.

METHODS: Eighty RA patients with baseline DAS28CRP < 3.2 and no swollen joints were examined at baseline and followed for 1 year. Patients were requested to contact the hospital in case of patient-reported hand flare accompanied by ≥ 1 tender and swollen joint. The 29 patients who reported hand flare had four extra visits within 4 months from flare onset comprising clinical examination, patient-reported outcomes, MRI, and US of wrists and hands. MRI synovitis/tenosynovitis/bone marrow edema (BME) and US synovitis/tenosynovitis were scored. MRI and US scores at and after the flare were compared to baseline before the flare, and associations were explored by linear mixed models for repeated measurements.

RESULTS: Synovitis and tenosynovitis by MRI/US increased significantly at flare onset. Synovitis waned quickly, as did US tenosynovitis. BME showed delayed increase yet persisted, once the patient-reported flare had resolved, as did MRI tenosynovitis. In univariate models, patient-reported flares were associated with all MRI and US inflammatory markers, except for BME, which was only associated with SJC28 and long-lasting flares > 14 days. Independent associations were observed between patient-reported flares and tenosynovitis by MRI and US (p < 0.05).

CONCLUSIONS: Patient-reported flares were linked to inflammation detected by serial MRI and US. Differential patterns of inflammatory lesion evolution were observed by serial imaging with early synovial and tenosynovial inflammation, followed by delayed-onset BME.