Magnetic resonance imaging (MRI) is considered an essential outcome tool for objective assessment of inflammation and damage in the sacroiliac joints (SIJ) of patients with juvenile spondyloarthritis (JSpA) which includes the juvenile idiopathic arthritis (JIA) subtypes of enthesitis related arthritis (ERA), psoriatic arthritis and some types of undifferentiated arthritis. An international multidisciplinary team of experts, the Outcome Measures in Rheumatology juvenile idiopathic arthritis MRI-SIJ (OMERACT JAMRIS-SIJ) working group, has developed standardized definitions for the inflammatory and structural lesions that may be observed on MRI of the SIJ and reported the first results of their validation in a reliability exercise of 30 patients with JSpA as a first step in the development of the OMERACT JAMRIS-SIJ score for quantifying inflammatory and structural lesions in the SIJ[1,2]. Although reliability for the assessment of the extent of bone marrow edema (BME) attained pre-specified proficiency targets, this was not the case for the assessment of structural lesions, especially erosion, backfill, and fat lesion.

Recent consensus-based deliberations conducted by imaging and methodology experts of the OMERACT consortium have resulted in the drafting of a framework of recommendations aimed at reducing sources of variability for imaging-based instruments and enhancing reliability of scoring instruments[3]. These are now considered essential for the application of an imaging instrument and include a clear description of the scoring framework, the availability of reference standards such as an atlas of images, and a systematic process for training and calibration using validated knowledge transfer (KT) tools. We have adhered to these recommendations in this second phase of development of the OMERACT JAMRIS-SIJ score by creating several KT tools which were then used for reader calibration prior to the second formal scoring exercise. First, an atlas of MRI inflammatory and structural lesions in the SIJ of patients with JSpA was developed by the JAMRIS-SIJ group[4,5] and is available at www.carearthritis.com/mriportal/jamrissij/index/. Second, the same website includes PowerPoint modules detailing the scoring methodology and DICOM-based reference images of MRI lesions in the SIJ from pediatric cases that were selected according to an international consensus of rheumatologists and radiologists[6]. Third, we also used validated web-based online KT tools created by developers of the Spondyloarthritis Research Consortium of Canada (SPARCC) MRI scoring instruments, SPARCC-SIJRETIC-INF and SPARCC-SIJRETIC-STR, for calibration of readers evaluating MRI lesions in the SIJ of adults with axSpA[7,8].

The aims of the OMERACT 15 JAMRIS SIG were twofold: 1. To determine whether the use of these KT tools could reduce sources of variability and enhance scoring proficiency to meet prespecified targets, especially for MRI structural lesions in the SIJ. 2. To compare fat-suppressed and gadolinium (Gd)-augmented MRI sequences to determine whether the latter offered any advantages as regards sensitivity or reliability for detection of MRI lesions.