Associations Between Spondyloarthritis Features and Magnetic Resonance Imaging Findings A Cross-Sectional Analysis of 1,020 Patients With Persistent Low Back Pain

Objective The Assessment of SpondyloArthritis international Society (ASAS) has previously published criteria for spondyloarthritis (SpA). In the Spines of Southern Denmark cohort, which included patients with persistent low back pain and an unknown proportion of patients with SpA, our objectives were 1) to estimate the prevalence of magnetic resonance imaging (MRI) findings and clinical features included in the ASAS criteria for SpA and 2) to explore the associations between MRI findings and clinical features. MethodsWe included patients ages 18-40 years with persistent low back pain who had been referred to the Spine Centre of Southern Denmark. We collected information on clinical features (including HLA-B27 and high-sensitivity C-reactive protein) and MRI findings in the spine and sacroiliac (SI) joints. ResultsOf 1,020 included patients, 537 (53%) had at least 1 of the clinical features included in the ASAS criteria for SpA. Three clinical features were commoninflammatory back pain according to the ASAS criteria, a good response to nonsteroidal antiinflammatory drugs (NSAIDs), and family history of SpA. The prevalence of these features ranged from 15% to 17%. Sacroiliitis on MRI according to the ASAS definition was present in 217 patients (21%). Of those 217 patients, 91 (42%) had the minimum amount of bone marrow edema required according to the ASAS definition (a low bone marrow edema score). The presence of HLA-B27, peripheral arthritis, a good response to NSAIDs, and preceding infection were independently positively associated with MRI findings in the SI joints (odds ratios [ORs] of 1.9-9.0). The remaining 8 clinical features were not positively associated with MRI findings. Importantly, only age was independently associated with low bone marrow edema score at the SI joints (OR of 1.1 per year). ConclusionIn this population, 53% had at least 1 clinical feature included in the ASAS criteria for SpA, and 21% had sacroiliitis according to the ASAS definition; furthermore, the associations between the clinical and imaging domains were inconsistent. The results indicate a need for further investigation of the importance of these findings in SpA, including investigation of the minimum requirements for defining sacroiliitis on MRI.