Comparison of sacroiliac CT findings in patients with and without ankylosing spondylitis aged over 50 years.

Diagnosis of axial spondyloarthritis (axSpA) is nowadays commonly made with the help of pelvic radiography or magnetic resonance imaging (MRI). However, there is an important inter-observer variability in radiography, and MRI is subject to possible false positives and is not the best modality for studying structural lesions. Conversely, pelvic computed tomography (CT) has excellent specificity and appears to be more effective than radiography for the diagnosis of ankylosing spondylitis (AS). However, its findings in patients over 50 years of age have not yet been studied. The objectives of this study were to describe the CT characteristics of sacro-iliac joints (SIJ) and the presence of intra-articular gas in patients with AS aged over 50 years and to compare them with controls of the same age and sex. This two-center, cross-sectional, observational study was performed using the medical records of the rheumatology departments of two University Hospitals. We included patients with a clinical diagnosis of axSpA, who had both definite radiographic sacroiliitis according to the modified New York criteria and met the ASAS 2009 criteria for axSpA (that is, AS), and who had undergone any CT scan including the whole SIJ. Each patient was matched for age and sex to a control randomly selected on the Picture Archiving and Communication System (PACS), symptomatic or asymptomatic, and without spondyloarthritis. For each individual, CT scans were interpreted blindly by two independent rheumatologists and scored for joint space narrowing (JSN), erosions, sclerosis, intra-articular gas, and diffuse idiopathic skeletal hyperostosis (DISH). Ninety patients and 90 controls were included in the study. The rates of positive JSN, erosion, and sclerosis scores were higher in the AS group (91% vs. 21%, p < 0.0001; 31% vs. 2%, p < 0.0001; 27% vs. 13%, p = 0.03, respectively), but the rates of intra-articular gas and DISH were higher in the control group (24% vs. 68%, p < 0.0001; 7% vs. 33%, p < 0.0001, respectively). 58% of patients had complete bilateral ankylosis. A total of 83 (92.2%) patients had a CT scan considered positive for AS, compared with only seven controls (7.8%). Sclerosis and erosions were predominantly on the anterosuperior part and iliac side of the joint in controls and were more diffuse in patients with AS. CT findings in patients with AS over 50 years of age are mostly represented by changes in the joint space; patients with AS have more erosions and sclerosis changes, but less intra-articular gas than controls.