MADRID – There is no added benefit of performing spinal MRI in the diagnosis of spondyloarthritis, the results of an international, multicenter study suggest.
Around one-quarter of patients with nonradiographic axial spondyloarthritis (nr-axSpA) who had a negative MRI scan of the sacroiliac joints (SIJs) were reclassified as being positive for SpA by a combined evaluation of SIJ MRI and spinal MRI scans.
However, 17.5% of healthy volunteers and up to 26.8% of patients with mechanical back pain who had a negative SIJ MRI were also reclassified as having SpA. This false-positive result balances out the value of combined spinal and SIJ MRI.
"Combined MRI added little incremental value compared to SIJ MRI alone for diagnosis of nr-axSpA," said Dr. Ulrich Weber, who presented the findings at the annual European Congress of Rheumatology (Ann. Rheum. Dis. 2013;72:145).
"Although you get about 20% more patients – which is the good news – we found about the same magnitude of false-positive controls," he said in an interview. He added that he "was very disappointed with this result, and these data need confirming." Dr. Weber collected data for the study while at the Balgrist University Clinic in Zurich, and also as a visiting professor in the rheumatology department at the University of Alberta, Edmonton.
SIJ MRI is a major criterion of the Assessment of SpondyloArthritis classification criteria for ankylosing spondylitis (AS). If a patient is strongly suspected of having early SpA, but this is not yet visible on radiographs, then an SIJ MRI is often the next step. If this is negative, however, it may be unclear what to do next.
"Our question was, ‘Would it help to order an additional spinal MRI in this situation?’ " Dr. Weber explained. These data suggest that it is not.
The collaborative study included 130 individuals with newly diagnosed back pain and aged 50 years or younger who were recruited from clinics based in Canada, China, Denmark, and Switzerland, as well as 20 healthy control individuals in whom MRI scans of the SIJ and spine were available.
The investigators used a clinical examination and pelvic radiography to stratify patients into groups, with 50 found to have nr-axSpA, 33 with AS, and 47 with mechanical back pain.
Three separate researchers blinded to the initial stratification read the MRI of the SIJ. An MRI of the spine was then performed 6 months later, with a combined SIJ/spinal scan done 1-12 months later. The presence or absence of SpA was determined in these scans, and comparisons were made between the results for MRI of the SIJ alone versus spinal MRI alone, as well as for the SIJ alone versus a combined read of both the spinal and SIJ MRI scans.
Dr. Weber noted that he would not recommend changing current practice as a result of this study. Further data are eagerly awaited from an ongoing Danish initiative that hopes to scan and assess around 2,000 whole-body MRIs in patients with suspected spondyloarthropathy by the end of the year. "This study will be very informative and very important for us because this is a large sample size. Preliminary data on about 1,000 MRIs point in the same direction," he observed.
Other data from the study, which Dr. Weber presented separately at the meeting, looked at the frequency and possible reasons for false-positive results with spinal MRI in the control groups (Ann. Rheum. Dis. 2013;72:125).
"Patients with mechanical back pain and healthy volunteers may show spinal MRI lesions suggestive of spondyloarthritis, such as corner inflammatory lesions or corner fat lesions," he explained. "We found that about 30% of those controls were misclassified as having spondyloarthritis by evaluation of the spinal MRI alone, so without SIJ MRI," said Dr. Weber. Bone marrow edema and fat infiltration were the MRI lesions largely responsible for this misclassification.
Dr. Weber said that, at least in Switzerland, general practitioners were more likely to order a spinal MRI than an SIJ MRI to assess a young patient with nonspecific back pain. "We think that caution is needed if a classification of SpA is based on MRI of the spine alone," he concluded.
Dr. Weber had no disclosures.