“In spring 2009, MRI was accepted by the international research organization on spondyloarthritis (ASAS) as equivalent to X-ray” for diagnosis, lead author Dr. Ulrich Weber told Reuters Health in e-mail. “The big disadvantage of X-rays is that it takes up to 10 years until definite lesions are seen. A diagnosis by MRI can be made within 3-4 months after symptom onset.”
Dr. Weber, from Balgrist University Hospital, Zurich and colleagues evaluated the sacroiliac joints of 187 subjects, using standardized definitions of active inflammatory and structural MRI lesions developed by the Canada-Denmark MRI Working Group. They focused on four types of lesions: bone marrow edema on short tau inversion recovery (STIR) sequences, and joint erosion, marrow fat infiltration, and ankylosis on T-1 weighted images.
Seventy-five subjects (mean age, 31) had ankylosing spondylitis, another 27 (mean age, 29) had “pre-radiographic” inflammatory back pain, 26 patients (mean age, 34) had non-specific back pain, and 59 healthy individuals (mean age, 31) volunteered to serve as controls.
The researchers first developed a reference set of images to serve as a benchmark and improve the concordance rate among the 5 physicians who interpreted the study images. For assessment of the images, they followed guidelines set by the Spondyloarthritis Research Consortium of Canada.
They found, Dr. Weber said, that “the diagnostic utility of MRI for spondyloarthritis is much higher than reported in the literature.”
Based on their standardized definitions of a positive MRI, the team estimates 90% sensitivity and 97% specificity for ankylosing spondylitis, and 51% sensitivity and 97% specificity for inflammatory back pain.
All 4 lesions – bone marrow edema, joint erosion, marrow fat, and ankylosis – gave complementary information, Dr. Weber said, but bone marrow edema and joint erosions were most important. “Ankylosis is not so important because we want to diagnose spondyloarthritis before this advanced disease stage,” he said, while “marrow fat infiltration probably is the least specific MR lesion.”
He recommends MRI for confirming a spondyloarthritis diagnosis suspected on clinical grounds, using both T1 and STIR sequences. “Without sufficient clinical suspicion, MRI can be misleading due to the relatively frequent unspecific findings in healthy volunteers and mechanical back pain patients.”
On the other hand, he doesn’t believe patients should be followed with serial MRI exams.
“In conclusion, this cross-sectional study adopting a systematic and standardized evaluation of acute and structural lesions of the sacroiliac joint showed that MRI has much greater diagnostic utility in spondyloarthritis that documented previously,” the researchers write.
“There is a need for greater awareness and training of rheumatologists in recognizing structural lesions on sacroiliac joint MRI,” they add.
Arthritis Rheum 2010.