Purpose or Case Report: To assess the equivalence of MRI without Gd-based contrast and MRI with contrast in the evaluation of synovitis in pediatric patients with juvenile idiopathic arthritis (JIA).

Contrast-enhanced imaging has been the gold-standard for MR evaluation of synovitis in patients with JIA (JAMRIS system). Conventional 2D MRI sequences did not allow reliable differentiation between synovium vs joint fluid. With higher field-strength magnets (1.5T or 3T), and high-resolution 3D sequences, the synovium may be differentiated from joint fluid on noncontrast MRI.
Methods & Materials: This is an IRB-approved, retrospective study performed in a tertiary children’s hospital. We included patients aged < 20 years who underwent knee MRI without and with contrast using a 1.5T or 3T magnet with confirmed diagnosis of JIA between Jan 2012 and Jan 2022. Each study was sorted into two anonymized worklists: one with only noncontrast images, and the other with only contrast-enhanced images. Studies were presented in random order.

Synovial thickness was independently measured by two pediatric radiologists (reader A, B). Each radiologist measured synovium at site of maximal thickness at 6 sites of the knee joint (patellofemoral, suprapatellar, infrapatellar, cruciate ligaments, and medial and lateral posterior condyle) in accordance to JAMRIS system. In this study, we analyzed numerical measurements (and their average over sites) in addition to ordinal scores (and their total over sites). Numerical measurements provided a plausible basis on which to define equivalence: noncontrast was considered equivalent to contrast if the true difference of means is within +/- 0.5 mm. Equivalence at the 5% level was established if the 90% CI for the difference fell in the equivalence interval (-0.5, 0.5).
Results: There were 165 studies who met the search criteria. By chart review, those studies without the diagnosis of JIA (127 studies) were excluded, and 38 studies from 35 patients (25 females, 10 males, mean age 13.3 yrs, range 2–20 yrs) were included. For reader A, 4/6 sites and the average showed equivalence. For reader B, 6/6 sites and the average showed equivalence. Inter-observer agreement was moderate (CCC = 0.55).
Conclusions: Unenhanced images are equivalent to contrast-enhanced images in assessment of synovitis with the use of higher field-strength magnets, and advanced sequences. Contrast could be eliminated for the evaluation of patients with JIA in these settings. Larger studies can confirm our findings.