A systematic critical review on MRI in spondyloarthritis

Introduction: Magnetic resonance imaging (MRI) has been proven capable of showing inflammatory and structural changes in patients with spondyloarthritis (SpA) and has become widely used in the diagnosis of SpA. Despite this, no systematic reviews evaluate the diagnostic utility of MRI for SpA. Therefore, the objective of this systematic review was to determine the evidence for the utility of MRI in the clinical diagnosis of SpA. The aims were to identify which MRI findings are associated with the diagnosis of SpA and to quantify this association. Methods: MEDLINE and EMBASE were electronically searched. Inclusion criteria were cross-sectional or longitudinal case-control or cohort MRI studies. The studies required a group with either SpA or inflammatory back pain (IBP) and a non-case group without SpA or IBP. Each group required a minimum of 20 participants. The included articles had to report results containing raw numbers suitable for the construction of two-by-two tables or report results by sensitivity and specificity for cross-sectional studies or odds ratios, relative risk ratios, or likelihood ratios for longitudinal studies. Method quality was assessed by using criteria based on the QUADAS tool. Results: In total, 2,395 articles were identified in MEDLINE and EMBASE before November 2011. All articles were reviewed by title and abstract. Seventy-seven articles were reviewed by full text, and 10 met the inclusion criteria. Two were considered of high quality: one evaluated the sacroiliac joints, and the other, the spine. Because of the small number of high-quality studies, a meta-analysis was not performed. The two high-quality studies found a positive association between MRI findings (bone marrow edema, erosions, fat infiltrations, global assessment of sacroiliitis, and ankylosis) and the diagnosis of IBP and SpA. Conclusion: In this review, several MRI findings were found to be associated with SpA. However, because of the small number of high-quality studies, the evidence for the utility of MRI in the diagnosis of SpA must be considered limited. Therefore, caution should be taken to ensure that inflammatory and structural MRI findings are not interpreted as being more specific for SpA than is supported by research.