Background Rheumatoid arthritis (RA) is a chronic inflammatory disease with elevated extracellular matrix (ECM) turnover. It is important to determine the efficacy of a therapy quickly to ensure that patients are given the right treatment. Biomarkers of ECM turnover could be such possible biomarkers. Matrix metalloproteinase degraded fragments of type III and IV collagen and C-reactive protein (C3M, C4M and CRPM) have previously found to be dose-dependently inhibited to tocilizumab (TCZ) and could identify ACR50 responders.
Objectives 1) Confirm that TCZ have a dose dependent effect on the circulating biomarkers C3M, C4M, and CRPM and 2) investigate if these biomarkers are associated with disease activity and pain in RA patients undergoing TCZ treatment.
Methods RA patients (n=217) fulfilling the American College of Rheumatology (ACR) criteria were included from a phase III randomized controlled trial (RADIATE). Patients received either TCZ+MTX (4mg/kg; TCZ4 or 8mg/kg; TCZ8) or placebo+MTX (PBO) every 4 weeks. C3M, C4M and CRPM were assessed in serum at baseline (BL) and week 16. Difference of a biomarker level over time was tested by paired t-test. Difference between the treatment groups was examined by one-way ANOVA. Correlation between biomarkers and clinical observations was investigated with Spearman's correlation. Bonferroni correction was used to find the critical p-value<0.0071 for correlations.
Results C3M significantly decreased (p=0.027) at week 16 compared to BL for TCZ8. TCZ8 and TCZ4 was decreased in both C4M (p=0.0001) and CRPM (p<0.0001) at week 16 compared to BL. A dose dependent decrease was observed for C4M and CRP as the difference between TCZ8 and TCZ4 (p=0.031 and p=0.02), TCZ8 and PBO (p<0.0001 and p<0.0001), and TCZ4 and PBO (p=0.002 and p=0.004)Changes to week 16 in the biomarker levels of all patients were found to correlate with the changes in HAQ-QI (C3M: ρ=0.26, C4M: ρ=0.33, CRPM: ρ=0.33), DAS28 (C3M: ρ=0.36, C4M: ρ=0.38, CRPM: ρ=0.39), VAS pain (C3M: ρ=0.32, C4M: ρ=0.35, CRPM: ρ=0.32), CRP (C3M: ρ=0.5, C4M: ρ=0.55, CRPM: ρ=0.55) and ESR (C3M: ρ=0.51, C4M: ρ=0.54, CRPM: ρ=0.58). The CRPM BL level correlated negatively with the changes in HAQ-QI (ρ=-0.21). C3M, C4M and CRPM BL levels correlated negatively with changes to week 16 in VAS pain (C3M: ρ=-0.23, C4M: ρ=-0.28, CRPM: ρ=-0.26), CRP (C3M: ρ=-0.29, C4M: ρ=-0.36, CRPM: ρ=-0.34) and ESR (C3M: ρ=-0.17, C4M: ρ=-0.23, CRPM: ρ=-0.27).
The changes in C4M and CRPM levels were significantly higher in ACR20 and ACR50 responders (C4M: p=0.011 and p=0.009, respectively and CRPM: p=0.004 and p=0.025, respectively) compared to non-responders.
Conclusions C4M and CRPM decreased in a dose dependent manner in response to TCZ, while C3M was significantly reduced by TCZ8. Changes to week 16 in C3M, C4M and CRPM levels correlated with changes in disease activity and VAS pain. These findings support that ECM biomarkers are predictive biomarkers of treatment benefit. This was furthermore supported by the finding that there were difference between ACR20 and ACR50 responders in C4M and CRPM levels. Overall, these data confirm that ECM biomarkers are applicable as predictive tools in RA.
A.C. Bay-Jensen et al. SAR 43 (2014) 470–478
Disclosure of Interest None declared