Contrast-enhanced MRI of the knee in asymptomatic pediatric controls compared to juvenile idiopathic arthritis patients to validate synovitis scores

Methods An axial fat-saturated T1-weighted MRI sequence of the knee of 25 asymptomatic controls and 25 JIA patients was collected, blinded and randomized. The asymptomatic controls were children who underwent MR enterography with IV contrast for unrelated diseases, had no (history of) joint complaints or signs of joint inflammation and gave permission for an additional sequence of the knee. JIA patients were age/sex-matched and divided in three clinical subgroups: new-active, relapse and inactive. Two readers independently scored synovial hypertrophy (SH) on a scale from 0-2 (none, 2-4mm, ≥4mm) at six locations, based on the Juvenile Arthritis MRI Scoring system (JAMRIS). Afterwards agreement on incongruent cases was obtained. Differences in SH score and (subgroups) of JIA patients were assessed. Results Mean age of all subjects (42% female) was 13.5 years (SD 2.5). Contrast-enhanced thickened synovium was present upon imaging of the knee in 60% of the controls (total SH score range 1-3) and 76% of the patients. A significant difference (p=0.007) in the total SH score was found between controls and JIA children. SH score could differentiate controls from the clinically active JIA subgroup (p=0.002) but not from the clinically inactive JIA subgroup (p=0.303). Findings only observed in the asymptomatic group consisted of an diffuse ‘contrastoutfading’ pattern in 28% of the controls and ‘pseudoenhancement’ of the cartilage at the posterior condyles only, that could be mistaken for true synovial inflammation.