Imaging Reliability Varies in Ankylosing Spondylitis

Nancy A. Melville

September 19, 2016

GHENT, Belgium — The detection of key structural changes in ankylosing spondylitis depends on the mode of imaging used, new research shows.

"We unexpectedly found that agreement between imaging methods was limited, with fewer erosions detected by MRI and x-ray than with CT. However, CT and MRI are more sensitive than x-ray for ankyloses," said investigator Xenofon Baraliakos, MD, from Ruhr-University Bochum, Germany.

Dr Baraliakos presented results from the first direct comparison of the three modes of imaging in ankylosing spondylitis here at the International Congress on Spondyloarthritides 2016.

MRI is considered the gold standard for detection of the inflammation and fat metaplasia characteristic in the sacroiliac joints and spine of patients with axial spondyloarthritis. In contrast, structural changes — such as erosion, sclerosis, and ankylosis — are more characteristic of ankylosing spondylitis.

The study involved 69 patients with ankylosing spondylitis and 49 age- and sex-matched control subjects who had undergone imaging for reasons other than back pain. All had complete sets of MRI, CT, and x-ray imaging available.

In the ankylosing spondylitis group, two readers assessed 552 sacroiliac joint quadrants for erosion and sclerosis and 276 pairs for ankylosis. In the control group, the readers assessed 392 sacroiliac joint quadrants.

First Direct Comparison

For erosion, agreement was strongest between CT and MRI (100 quadrants), followed by x-ray and MRI (70 quadrants) and then x-ray and CT (64 quadrants). Specifically, 48.9% of the changes detected on x-ray were detected on CT, and 45.4% of the changes detected on CT were detected on x-ray.

For sclerosis, agreement was strongest between x-ray and MRI (45 quadrants), followed by x-ray and CT (31 quadrants) and then CT and MRI (22 quadrants). Specifically, 52.3% of the changes detected on x-ray were also detected on MRI, and 71.4% of the changes detected on MRI were also detected on x-ray.

For ankylosis, agreement was strongest between x-ray and MRI (94 pairs), followed by x-ray and CT (87 pairs) and then CT and MRI (72 pairs). Specifically, 79.1% of the changes detected on x-ray were detected on MRI, and 67.9% of the changes detected on MRI were detected on x-ray.

From this analysis, "we learned that erosions and ankyloses are more common than sclerosis in the sacroiliac joints of AS patients, and they are rarely seen in controls," Dr Baraliakos reported.

The stage of disease could play a role in the detection of changes by the different modes, he suggested.

"The large potential of MRI to detect erosions may be related to early disease stages, while erosions seen on CT seem to be due to detection in later stages of AS," he explained.

"MRI seems superior in early stages, when the effects are more in cartilage, but not so much the bone. CT seems superior in the later stages, when both the cartilage and bone are affected," Dr Baraliakos said.

Surprise Finding

Although the limited agreement between modes of imaging "was a surprise," this study helps confirm the different roles of each, said session comoderater Atul Deodhar, MD, from the division of arthritis and rheumatic diseases at Oregon Health & Science University in Portland.

"This was a very well-done study, and it confirms what we already knew — that CT scan is better than plain x-ray or MRI in finding erosions. It finds cortical bony edge better than MRI and x-ray," Dr Deodhar told Medscape Medical News.

Nevertheless, the drawbacks of CT, specifically radiation exposure, need to be weighed against the benefits, he pointed out.

"We need a low-dose CT scan method to reduce the radiation dose associated with conventional CT," he said. "These are in development and we need to then see if the low dose will still have good accuracy and maintain this advantage."

These findings reflect the wide array of factors that can make a difference in imaging accuracy, said Robert Lambert, MD, from the Department of Radiology and Diagnostic Imaging at the University of Alberta in Edmonton, Canada.

"It's still early days in this very dynamic field of research," he told Medscape Medical News.

"One problem is that the research is moving faster than the natural history of the disease, so new data are coming out all the time, before we've had a chance to follow-up on the new work from a year or two ago," he explained.

Dr Lambert, who spoke at the meeting on challenges related to the MRI assessment of disease activity, said he agrees that methodology and stage of disease influence assessment.

"At a very detailed level, we would expect to see significant differences between imaging methods. At the macro level, how much difference we see depends on the study population — early disease equals greater difference, and late disease equals little difference," he said.

But the level of the clinician's experience is a major factor in the assessment of structural changes. "It depends a lot on background and training," he pointed out.

Dr Baraliakos has disclosed no relevant financial relationships. Dr Lambert reports that he consults for BioClinica.

International Congress on Spondyloarthritides (SPA) 2016. Presented September 16, 2016.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.