Rapid separation of serum does not avoid artificially higher matrix metalloproteinase (MMP)-9 levels in serum versus plasma

Objectives: To examine whether the time between blood drawing and centrifugation (TBDC) affects the levels of matrix metalloproteinase (MMP)-9 and MMP-2 levels in serum and in plasma samples, and to assess whether there is correlation between MMP-9 and MMP-2 levels in serum and plasma samples. Design and methods: Serum and plasma samples (N=8) were separated from venous blood collected into citrate, heparin, and EDTA tubes, which were either centrifuged immediately or after 5, 10, 20, or 30 min after blood drawing. We assessed the correlation between MMP-9/MMP-2 in serum and citrate, heparin, and plasma samples (N=20), which were assayed for gelatine zymography of MMP-2 and MMP-9. Results: MMPs are released by platelets or leukocytes during platelet activation or sampling process, thus leading to artificially higher MMP-9 levels in serum compared with citrate, heparin, or EDTA plasma samples, independently of TBDC. Citrate and heparin plasma samples had the lowest Pro-MMP-9 and MMP-9 levels, which correlated with each other. Pro-MMP-9 levels in serum correlated with Pro-MMP-9 levels in EDTA or citrate plasma, but not with heparin plasma. While no significant correlations were found between MMP-9 levels in serum and those found in plasma samples, the total MMP-9 levels (Pro-MMP-9+MMP-9) in serum and in plasma samples correlated with each other. No significant differences were found in pro-MMP-2 levels. Conclusions: These results suggest that the circulating levels of MMP-9 should be assessed in citrate or heparin plasma samples, but not in serum samples because of artificially higher MMP-9 levels in serum, independently of TBDC, and because they do not correlate with the MMP-9 levels in plasma samples. (c) 2006 The Canadian Society of Clinical Chemists. All rights reserved.