Abstract
Calprotectin (S100A8/A9), a protein expressed in neutrophils and monocytes/macrophages in circulation and inflamed tissue, is associated with measures of disease activity in rheumatoid arthritis (RA) patients both when measured in ethylenediaminetetraacetic acid (EDTA)-plasma and in serum. We wanted to explore if EDTA-plasma or serum should be preferred for calprotectin as a marker of disease activity. Calprotectin was analysed in EDTA-plasma and serum by enzyme-linked immunosorbent assay (ELISA) at baseline in 141 RA patients, starting biologic disease-modifying anti-rheumatic drugs (bDMARDs), and after three months. Differences between plasma and serum levels of calprotectin were assessed by Wilcoxon signed rank test. Variability was assessed by quartile coefficient of dispersion. Spearman’s test explored correlations between calprotectin in plasma and serum and between calprotectin (plasma or serum) and clinical/ultrasound (US) measures of disease activity. Bland Altman plots were used for method comparisons. Conventional inflammatory markers were evaluated for comparison. Calprotectin had similar variability when measured in plasma and serum, but there was a significant difference in concentrations between plasma and serum (p < .001). The correlation coefficients at baseline between calprotectin measured in plasma/serum and measures of disease activity were rs = 0.62/0.46 for sum power Doppler score (PD), rs = 0.60/0.48 for assessor’s global visual analogue scale (VAS), rs = 0.59/0.43 for sum grey scale (GS) score and rs = 0.47/0.37 for swollen joint count of 32, all p < .001. Similar differences were found after three months. Calprotectin measured in plasma showed the strongest associations with assessments of disease activity, and EDTA-plasma should preferably be used when evaluating disease activity in RA patients.
Acknowledgement
We thank study nurses Britt Birketvedt and Anne-Katrine Kongtorp for the clinical evaluation of patients and Marianne Eidsheim for technical assistance with the calprotectin ELISA. We thank Magne Solheim for statistical help and advice, and Karl Albert Brokstad for help with the use of GraphPad Prism. We also thank CALPRO AS for the support of a proportion of the ELISA kits used in this study and for the CV analyses.
Disclosure statement
Hilde Berner Hammer has received fees for speaking and/or consulting from AbbVie, Pfizer, UCB, Roche, MSD, BMS and Novartis.
The other authors declare no conflicts of interest.