Background/Purpose: The lack of pediatric classification criteria for axial disease is a major impediment to the conduct of clinical trials for juvenile spondyloarthritis (SpA). Classification criteria for axial disease in children with SpA are being developed. The objective of this study was to identify and grade a set of items that maximizes the likelihood of accurate classification of children and adolescents with axial disease.

Methods: First, we conducted an item-generation exercise with an international group of pediatric rheumatologists to generate a comprehensive list of clinical, laboratory, and imaging features used to decide whether or not a child/adolescent has clinical evidence of axial involvement. Second, we performed a literature review to formulate a list of features used in juvenile and adult SpA clinical study inclusion criteria and adult axial SpA classification criteria. Next, we completed a grading exercise in which respondents in the item generation survey rated the features generated from the first 2 steps for specificity (-3 to +3), with anchors “much less likely to have axial disease” and “much more likely to have axial disease”, respectively. Items with a mean score of zero or close to zero and redundant items were removed. The final list of items was reviewed by a panel of juvenile SpA experts and split into “entry criteria”, which would be required for classification applicability, potential “additive criteria”, and potential “exclusion” or “negatively weighted criteria”. Additive and negatively weighted criteria were grouped into domains by an iterative process.

Results: One hundred and ninety-nine physicians from 45 countries proposed 108 items useful in the determination of axial involvement or not. Responses showed broad coverage but also high consistency. Responses were grouped into 15 domains with a range of 38-141 unique respondents noting ≥1 item in a domain. Most items were related to imaging, location of pain or stiffness, pattern of pain, genetics, range of motion, laboratory markers, and comorbidities.  Literature review did not identify any additional clinical features to include in the item reduction exercise. In the grading exercise, 87 items had a mean score of ≥1.5 and 5 items ≤1.5. The expert panel reviewed the candidate items, grouped them into 6 independent domains, and identified 2 entry criteria, 2 exclusion criteria, 23 potential additive criteria, and 4 negatively weighted criteria (Table). The expert panel also voted to align criteria within the imaging domains with definitions provided by the OMERACT Juvenile Arthritis MRI Score (JAMRIS) international Working Group (1) when possible and to exclude radiographs and spinal MRI. Final additive and negatively-weighted criteria will be determined after evaluation of the sensitivity and specificity of potential criteria in a development cohort of juvenile SpA cases.

Conclusion: The item-generation exercise, grading exercise, and expert panel’s iterative work resulted in inclusion, exclusion, candidate additive and negatively weighted criteria. Refinement and weighting of criteria as well as determination of threshold for classification will be ascertained in the next phase.

*Peripheral arthritis should be present for ≥6 weeks. ^≥4 of the following: improvement with exercise, pain at night, insidious onset, age at onset < 40 years, and no improvement with rest (3). # modified NY criteria for radiographs(4) and ASAS MRI working group definitions of active sacroiliitis (5).

« Back to ACR Convergence 2020

ACR Meeting Abstracts - https://acrabstracts.org/abstract/development-of-candidate-criteria-for-axial-disease-in-juvenile-spondyloarthritis-an-international-collaboration/