Ultrasonography as a useful modality for documenting sacroiliitis in radiographically negative inflammatory back pain: a comparative evaluation with MRI

Objectives. The aims of this study were to identify and characterize features of sacroiliitis in patients with non-radiographic inflammatory low back pain by ultrasonography (USG) and to correlate the findings with that of MRI. Methods. MRI and USG of SI joints were performed on 29 patients who fulfilled the definition of inflammatory low back pain according to the Assessment of SpondyloArthritis International Society 2009 criteria for axial SpA but were X-ray negative for sacroiliitis. Increased vascularity, low resistive index (RI) and hyperechogenicity of the joint space were considered USG features of sacroiliitis. The findings were compared with those of 32 controls. USG features of sacroiliitis were compared with MRI by kappa statistics. Results. Receiver operating characteristic analysis revealed cut-off values for flow signals and RI of 3 and 0.605, respectively. There was a significant difference in the number of flow signals, RI and echogenicity of the SI joint between MRI-proven cases and controls. The Cohen's kappa for flow signals, RI and hyperechogenicity when compared with MRI were 0.816 (95% CI 0.676, 0.937) and 0.821 (95% CI 0.662, 0.965) and 0.403 (95% CI 0.108, 0.695). Taking both flow signals and RI parameters as criteria for determining sacroiliitis, comparison with MRI returned a kappa of 0.816 (95% CI 0.601, 0.963). Conclusion. Three or more flow signals and a RI <= 0.605 can be applied as USG criteria for sacroiliitis. USG can be a cost-effective and non-inferior modality compared with MRI in documenting sacroiliitis in early SpA.