WASHINGTON, DC—Ankylosing spondylitis (AS) research reported at the 70th annual meeting of the American College of Rheumatology (ACR) suggests that serum matrix metalloproteinase 3 (MMP3) levels are a good predictor of progressive joint damage, but that magnetic resonance imaging (MRI) of the sacroiliac joints is a less sensitive marker of AS than had been assumed.^1,2

Serum MMP3 Levels Especially Accurate in Patients with Pre-Existing Damage

Walter P. Maksymowych, MD, professor of medicine at the University of Alberta in Edmonton, in Canada, and colleagues, analyzed a panel of biomarkers in 100 AS patients to determine which ones predicted structural damage over a subsequent 2-year study period.

The panel included cartilage oligomeric matrix protein; human cartilage glycoprotein-39 (YKL-40); C2C epitope; C1,2C epitope; 846 epitope; CPII epitope; osteoprotegerin; and MMP3. Researchers used baseline clinical and radiographic data plus 2-year radiographic progression data to draw their conclusions. One observer who was aware of the time order of the x-rays scored radiographic progression with the modified Stoke Ankylosing Spondylitis Spinal Score (mSASSS), (range: 0-72).

"As we have new treatments for all of our diseases including AS, it's critical to identify some of the things that might predict this kind of damage," Dr. Maksymowych said. "AS is a disease that occurs very early in life, and 30% to 40% have a particularly aggressive course. The most severe outcome is the spine becoming rigid, and as of now there are no predictors of which patients will develop the most severe variant of the disease."

Of all the proteins assessed, MMP3 proved to be most strongly associated with progressive joint and bone damage, particularly in those patients with pre-existing damage visible on baseline radiograph. MMP3 levels correctly identified two-thirds of patients who progressed. YKL-40) showed some correlation with joint damage, but not after researchers adjusted for sex, age, disease duration, C-reactive protein, and baseline mSASSS.

"Measuring MMP3 is [a means] to predict the majority of AS patients who are likely to progress, and it could be readily implemented in the clinic at this time," Dr. Maksymowych said. "This is a test that should be done in patients with AS." MMP3 is also known to be a good predictor of structural damage in rheumatoid arthritis (RA).

MRI of SI joints in AS

In a related presentation,² researchers showed MRI of the sacroiliac (SI) joints in patients with axial undifferentiated spondyloarthritis (uSpA) is not as sensitive or as specific a predictor of AS as previously believed.

"There is very little data on the sensitivity or specificity of MRI findings in AS, which have been assumed to be about 90%," said lead researcher Martin Rudwaleit, MD, of the rheumatology department at Charité-Campus Benjamin Franklin in Berlin.

Instead, the new study showed sensitivity to be 67% to 85%, Dr. Rudwaleit said. The specificity was very high, which had been expected, given that active inflammatory lesions were rare in mechanical low back pain.

The researchers compared MRI and radiographs of the SI joints in 209 patients who reported chronic back pain as a leading symptom between January 2002 and June 2003 (cohort 1) and 342 patients who presented at a specific inflammatory back pain referral clinic with yet undiagnosed chronic back pain in July 2004 (cohort 2).

Nine of 11 rheumatologists made a diagnosis of AS, axial uSpA, ‘possible' axial uSpA, and back pain of mechanical origin (mLBP), which served as the gold standard. The majority of patients without definite radiographic sacroiliitis underwent MRI of SI joints. Only patients for whom MRI was available were analyzed for the presence or absence of active inflammatory lesions without further grading.

In cohort 1, active inflammatory lesions of SI joints on MRI were found in 32 of 48 patients (66.7%) with early axial uSpA, in 8.6% of 58 patients with mechanical LBP, and in 45.5% of patients with ‘possible' axial uSpA.

In cohort 2, active inflammatory lesions on MRI  were found in 41 of 48 patients (85.4%) with early axial uSpA, 1.5% of 68 patients with mechanical LBP, and 14.3% of 21 patients with ‘possible' uSpA.

The presence of chronic lesions of the SI joints that were compatible with axial SpA were only assessed systematically in cohort 2. In that group, 43.8% of 48 patients with early axial uSpA had chronic lesions compared to 7.5% of 67 mechanical LBP patients and 28.6% of patients with ‘possible' axial uSpA.

Eighteen of 23 patients (16 from cohort 1 and 7 from cohort 2) with no active inflammatory MRI lesions had the combination of inflammatory back pain, positive HLA-B27, and good response to nonsteroidal anti-inflammatory drugs (NSAIDs), and 14 of these 18 had at least one extraspinal manifestation. Moreover, six of the 23 patients had imaging evidence of sacroiliits.

"Patients with early axial uSpA, but normal MRI of SI joint, were largely HLA-B27 positive, had inflammatory back pain, and had good response to NSAIDs," Dr. Rudwaleit said. "The sensitivity may be quite high if performed by an experienced radiologist, but you can only have high likelihood ratios if you are sure the interpretation of MRI is correct."

E-mail any comments to .


Reference