MRI-Detected Erosions of the Small Joints Not Predictive of Later Rheumatoid Arthritis

Pensive doctor looking the xray exams
Magnetic resonance imaging-detected erosions of the hand and feet were not predictive of rheumatoid arthritis in patients with arthralgia.

Magnetic resonance imaging (MRI)-detected erosions of the hand and feet were not predictive of rheumatoid arthritis (RA) in patients with arthralgia, according to study data published in the Scandinavian Journal of Rheumatology. Instead, MRI-detected subclinical inflammation was the strongest correlate of later disease progression in patients with clinically suspect arthralgia.

The study cohort comprised patients with recent-onset (<1 year) arthralgia in the small joints who received care at a rheumatology clinic in Leiden, the Netherlands. Patients were enrolled between April 2012 and October 2018. Individuals were excluded if clinical arthritis was detected at baseline or if another etiology for joint arthralgia was present. At baseline, patients underwent a physical examination, blood draw, and high-contrast MRI of the hands and feet. Erosions were scored using the Rheumatoid Arthritis Magnetic Resonance Imaging Scoring system. Follow-up visits were conducted at 4, 12, and 24 months. Patients who initiated treatment with disease-modifying antirheumatic drugs during follow-up were excluded from analyses.

The primary outcome was diagnosis of inflammatory arthritis, determined by the treating rheumatologist. Multivariable logistic regression was performed to assess the relationship between MRI-detected lesions and later diagnosis of arthritis. Analyses were stratified by anticitrullinated protein/peptide antibody (ACPA) positivity and adjusted for patient age and subclinical inflammation levels.

The study cohort comprised 490 patients, of whom 83 developed inflammatory arthritis during follow-up. Median follow-up time to inflammatory disease diagnosis was 14 weeks (interquartile range, 3-23 weeks). Median follow-up duration in patients who did not progress to arthritis was 103 weeks (interquartile range, 51-113 weeks). Median total erosion score was greater in patients who progressed to arthritis compared with patients who did not.

The presence of MRI-detected erosions was not significantly associated with arthritis development in either the univariate model (hazard ratio, 1.40; 95% CI, 0.86-2.28) or the model adjusted for age and subclinical inflammation (hazard ratio, 0.97; 95% CI, 0.59-1.59). Analyses were then restricted to RA-specific erosions: grade ≥2 erosions, erosions of the fifth metatarsophalangeal joints (MTP5), and erosions in MTP1 in patients aged <40 years. None of these erosion types were predictive of later progression to inflammatory disease.

In analyses stratified by ACPA status, these trends persisted: MRI-detected erosions did not predict disease progression in either ACPA subset. The researchers instead found subclinical inflammation level to be the strongest predictor of later disease progression. Area under the receiver operating characteristic curve (AUC) was 0.73 for subclinical inflammation vs 0.54 for MRI erosion. When MRI-detected erosion was added to the model, the AUC remained at 0.73, suggesting that prognostic accuracy was not improved by erosion assessment.

Results from this longitudinal study indicated that MRI-detected erosions of the hands and feet are not predictive of later arthritis development in patients with clinically suspect arthralgia. “This implies that evaluating MRI erosions of [patients with clinically suspect arthralgia] is superfluous if MRI-detected subclinical inflammation is assessed,” the investigators wrote.


Wouters F, Matthijssen XME, Boeters DM, et al. Do magnetic resonance imaging-detected erosions predict progression to rheumatoid arthritis in patients presenting with clinically suspect arthralgia? A longitudinal study [published online June 2, 2020]. Scand J Rheumatol. doi:10.1080/03009742.2020.1737221