Background/Purpose: The OMERACT Juvenile Idiopathic Arthritis MRI (JAMRIS) Working group has defined a spectrum of MRI lesions in the sacroiliac joint (SIJ) in youth with spondyloarthritis and axial disease (axJSpA) that can be included in a scoring method. Previous reports have described the application of Spondyloarthritis Research Consortium of Canada (SPARCC) methodology to quantify MRI SIJ lesions in axJSpA, where assessment of inflammation is limited to bone marrow edema (BME) and lesions are assessed in a limited number of semicoronal slices. We aimed to assess the impact of including all inflammatory lesions and SPARCC methodology versus an all-slice method on the metrology of JAMRIS in two longitudinal cohorts of youth with axJSpA treated with a tumor necrosis factor inhibitor (TNFi).

Methods: The two cohorts recruited youth with axJSpA who had a rheumatologist diagnosis of axJSpA. All subjects underwent MRI as part of a diagnostic evaluation for axial disease. MRI lesions were assessed on semicoronal slices through the SIJ from the most anterior (≥1cm vertical height of cartilaginous portion) to the most posterior (≥1cm visible cartilaginous portion) slice. Structural lesions were assessed first on the T1-weighted scan followed by inflammatory lesions on a fluid-sensitive sequence. Scans were obtained at baseline and 3 months-1 year follow up after initiation of treatment with TNFi and assessed blinded-to timepoint and clinical details by 7 readers (4 pediatric and 1 adult MSK radiologists, 1 adult and 1 pediatric rheumatologist) after standardized calibration using a CARE Arthritis online module. MRI lesions were recorded based on standardized lesion definitions derived from ASAS and the JAMRIS group in a standardized eCRF (ASAS MRImagine). Analyses included descriptive statistics, reliability (intraclass correlation coefficient (ICC), smallest detectable difference (SDD) or change (SDC)), and responsiveness (standardized response mean (SRM)).

Results: Pre- and post-treatment scans (mean(SD) duration between scans 46.6(47.1) weeks) were available from 60 cases, 30 from each cohort. An inflammatory lesion in addition to BME was recorded in 41(68.3%) cases by a majority of readers. But only 1 case had an inflammatory lesion without any BME. For change in structural lesion scores after TNFi treatment, minimal differences were evident between SPARCC and all slice methodologies (Figure 1A). The major contributor to change in inflammatory lesion scores was BME followed by inflammation in an erosion cavity (IEC), though differences between SPARCC and all slice methodologies were minimal (Figure 1B). Reliability for detection of change scores was comparable between SPARCC and all slice methodologies (Table). Responsiveness was comparable between methodologies and combining scores for inflammatory lesions was not superior to BME alone (Figure 2).

Conclusion: BME is the most responsive lesion to treatment with TNFi in axJSpA followed by IEC and then erosion. Responsiveness is not enhanced by assessment of all slices as compared to SPARCC nor by scoring additional inflammatory lesions beyond BME.

Figure 1. Change from baseline values (in boxes) in A. MRI structural lesion scores (SIJ quadrants for erosion, sclerosis, and fat lesion, SIJ halves for backfill and ankylosis) and B. MRI inflammation scores (SIJ quadrants for BME and inflammation in an erosion cavity (IEC)), after TNFi assessed across all SIJ slices or limited to the slices selected in the SPARCC methods.

Table. Reliability of change scores for MRI SIJ inflammatory and structural lesions for all 7 readers, MSK radiologists (n=5), and rheumatologists (n=2). SDC smallest detectable change.

Figure 2. Standardized response means for MRI SIJ inflammatory and structural lesions scores after TNFi assessed across all SIJ slices or limited to the slices selected in the SPARCC method. IEC inflammation in an erosion cavity

« Back to ACR Convergence 2025

ACR Meeting Abstracts - https://acrabstracts.org/abstract/the-omeract-juvenile-idiopathic-arthritis-mri-score-for-active-and-structural-lesions-in-the-sacroiliac-joints-development-and-validation-in-two-longitudinal-cohorts/