Background/Purpose: The ultimate goal for clinicians caring for children with Juvenile idiopathic arthritis (JIA) is to attain disease remission off medications. Correct identification of remission is crucial in the decision to withdraw medications. Ultrasound (US) is a non-invasive and sensitive technique used to detect the presence of active synovitis. Between 30-40% of JIA patients in physician diagnosed clinical remission demonstrated persistent active synovitis with ultrasound in published experiences, and withdrawal of medications in these patients may precipitate a disease relapse. The purpose of this study is to assess whether or not the use of ultrasound can better enable pediatric rheumatologists to determine clinical remission in children with JIA.

Methods: We recruited 41children with either oligoarticular or polyarticular JIA, 17 with clinically active disease and 24 in clinical remission. Three to 6 joints were scanned using US in each child, including the currently or previously most active joints. Joint effusion, synovial thickening, erosions and synovial hyperemia by Power Doppler signal were recorded within each joint. These four variables were scored on a scale from 0-3 and a score of 2 or 3 was considered to represent significant clinical change. Univariate and multivariate logistic regression analyses were performed, to best distinguish the two groups. Inter-reader reliability was calculated using Cohen’s kappa statistics between two independent ultrasonographers.

Results: Of the 41 subjects, 73.2% were female, 90.3% were Caucasian and 48.8% were polyarticular. 156 joints were scanned; 63 in the active group and 93 in the remission group. The most commonly scanned joints were MCP’s (50), PIP’s (50) and knees (29). Overall 57.7% of joints had significant US abnormalities: 74.6% in the active group, and 46.2% in the remission group (p=0.003). 64% of MCP joints had abnormal findings on US: 84.2% in the active group, and 51.6% in the remission group (p=0.053). 33.3% of ankle joints had abnormal findings: 62.5% in active group, and 0.0% in remission group (p=0.03). The calculated kappa coefficient of 0.68 indicates a high degree of inter-reader agreement. The most common abnormal finding was Power Doppler signal 47.4%. PIPs were the joints most commonly abnormal in the remission group. None of the patients in clinical remission who had normal ultrasound findings had elevated ESRs, but 42.1% of the patients with abnormal ultrasound findings had elevated ESRs (P = 0.5).

Table 1: Frequency of ultrasound abnormalities by group
	All patients (N= 41)	Active (N=17)	Remission (N=24)	p-value	Odds Ratio
All joints	90/156 (57.7%)	47/63 (74.6%)	43/93 (46.2%)	0.003	3.51 (1.53, 8.04)
Elbow	3/4 (75.0%)	2/2 (100.0%)	1/2 (50.0%)	0.25
Wrist	4/6 (66.7%)	2/2 (100.0%)	2/4 (50.0%)	0.47
MCPs	32/50 (64.0%)	16/19 (84.2%)	16/31 (51.6%)	0.053	5.07 (0.98, 26.3)
PIPs	35/50 (70.0%)	16/19 (84.2%)	19/31 (61.3%)	0.12	3.66 (0.67, 19.9)
DIPs	1/2 (50.0%)	1/1 (100.0%)	0/1 (0.0%)	0.16
Knee	10/29 (34.5%)	5/12 (41.7%)	5/17 (29.4%)	0.53	1.81 (0.20, 16.3)
Ankle	5/15 (33.3%)	5/8 (62.5%)	0/7 (0.0%)	0.03

Conclusion: There were higher than expected findings of active synovitis (46%) in children believed to have been in clinical remission, with the most common active joints being the PIP, MCP, and ankle joints. Use of US with Power Doppler of these specific joints may prove to be helpful in assisting clinicians to more accurately define remission in children with JIA.