Background/Purpose: Better objectives measures for evaluating disease activity and anti-TNFα response in patients with ankylosing spondylitis (AS) are needed. MMP-3 seems to be the most promising biomarker but published data are not conclusive. The aim of this study was to investigate the association between serum biomarkers (MMP-3, DKK-1 and sclerostin) levels with disease activity parameters and to evaluate if these biomarkers are useful to predict anti-TNFα response in patients with AS.

Methods: From November 2010 to July 2011, consecutive patients with AS (New York criteria) who initiated anti-TNFα therapy in a University hospital were included. Before and after 3 months of therapy, disease activity was measured using BASDAI, ASDAScrp, CRP, patient’s VAS of pain and patient’s and physician’s VAS of global disease activity (GDA). Blood samples for determination of serum levels of biomarkers by enzyme immunoassay were also collected at both visits. Spearman correlation test was used to evaluate association between biomarkers and disease activity parameters. Biomarkers change was compared in responders versus non-responders, based on BASDAI50 and ASDAS response using Mann-Whitney U test. Accuracy to predict response (ROC analysis) was performed.

Results: Twenty AS patients were included; 80% received adalimumab and 20% received etanercept. Median (IQR) age and disease duration were 42.4 (31-49) years and 6.8 (3-10) years, respectively; 86% were men, and 83% HLA-B27 positive. Inverse correlation between patient’s VAS pain and MMP-3 at baseline was the only significant observed correlation between investigated biomarkers and disease activity parameters (table 1).  MMP-3 levels decreased only in patients who responded to anti-TNF therapy but increased in patients who did not response (table 2). Baseline biomarkers serum levels (including CRP) for both groups were similar except for MMP-3 (122.9 vs 58.9; p<0.05). The area under the curve for MMP-3 to predict BASDAI50 and ASDAS response was 0.73 and 0.78, respectively. The best cut-off was established for levels higher than 59.5, with sensibility of 79-85% and specificity of 50-57%.

Conclusion: No correlation was observed between serum levels of MMP-3, DKK-1 and sclerostin and disease activity parameters in patients with AS. Serum levels of MMP-3 may be useful to predict response to anti-TNFα therapy in patients with AS.

Table 1: Correlation between serum levels of biomarkers with disease activity parameters.

              * p<0.05 ** p<0.01

Table 2: Change in serum levels of MMP-3, DKK-1, sclerostin and CRP after 3 months of anti-TNFα therapy based on the clinical response.