Prevalence and anatomical distribution of MRI lesions in axSpA and differences between patients with and without peripheral involvement: results from the ASAS classification cohort

Farisogullari B., Wichuk S., Baraliakos X., Eshed I., De Hooge M., Lambert R. G., ...More

RMD Open, vol.12, no.1, 2026 (SCI-Expanded, Scopus) identifier identifier

  • Publication Type: Article / Article
  • Volume: 12 Issue: 1
  • Publication Date: 2026
  • Doi Number: 10.1136/rmdopen-2025-006266
  • Journal Name: RMD Open
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, EMBASE, MEDLINE, Directory of Open Access Journals
  • Keywords: Axial Spondyloarthritis, Inflammation, Magnetic Resonance Imaging, Spondylitis, Ankylosing
  • Hacettepe University Affiliated: Yes

Abstract

Objective To evaluate the prevalence and anatomical distribution of inflammatory and structural MRI lesions in axial spondyloarthritis (axSpA) and compare these between patients with isolated axial involvement and those with peripheral manifestations. Methods Data from the Assessment of SpondyloArthritis International Society (ASAS) Classification Cohort were analysed. Peripheral involvement was defined as past or current arthritis/dactylitis/enthesitis. Sacroiliac joint (SIJ) and spinal MRI lesions typical of axSpA were classified per ASAS lesion definitions and centrally read with multi-reader majority agreement (lesion present if called by the majority; SIJ ≥4/7, spine ≥5/9 readers). Comparisons between patients with and without peripheral manifestations were made. Results Among 199 axSpA patients with SIJ MRI, 67 also had spinal MRI. Subchondral SIJ bone marrow oedema (BMO) was observed in 49%, without quadrant preference or subgroup differences. Other SIJ inflammatory lesions ranged from 4%–18%. Erosions (35%) and fat lesions (22%) were the most frequent structural lesions. In the spine, BMO, fat lesions and syndesmophytes/ankylosis were detected in 38%, 25% and 5%, respectively, with similar subgroup frequencies. Among 40 patients with both SIJ and whole spine MRI, inflammatory lesions were observed in both sites in 18%, SIJ only in 38%, and spine only in 20%. Structural lesions occurred in both sites in 19%, SIJ only in 30%, and spine only in 5%, with no subgroup differences. Conclusion The prevalence and anatomical distribution of ASAS-defined MRI lesions was similar across axSpA subgroups. Notably, 20% exhibited spine-only inflammation, suggesting potential added diagnostic and monitoring value of spinal MRI, warranting further study.