Preanalytical Conditions and Circulating Matrix Metalloproteinases

We appreciate the interest of Klaus Jung in our study and his response to the manuscript on serum matrix metalloproteinases (MMP)-2 and MMP-9 after human liver transplantation. We share his concern regarding the effects of blood sample collection on the accuracy and diagnostic value of serological MMP determinations in pathological processes. We were aware of the fact that the preanalyte used might have a considerable impact on the results obtained. Exactly this concern made us perform a preevaluation of different blood samples from seven liver transplant patients, collected at six time points from pretransplantation to one year of follow-up, before we executed the complete study. We routinely collected simultaneously serum (without clotting activator), sodium-heparin, and K2-EDTA blood samples through unforced blood sampling of the patients in the morning, using the Becton Dickinson Vacutainer system. The samples are also processed simultaneously by centrifugation and stored at -70°C until analysis. Similar to Jung et al. (1, 2) we also found considerable differences in the MMP levels, as determined by ELISA, depending on the type of analyte used (see Table 1).TABLE 1: MMP levels in different analytes of liver transplant patientsOverall, EDTA-plasma samples were found to have a two- to threefold lower MMP level compared to the serum and heparin-plasma levels. However, the latter two samples had higher MMP-9 levels in serum but comparable MMP-2 levels. In addition, the correlation in the MMP levels between the different analytes were highly significant (P< 0.0005; i.e., MMP-9: 0.53 < r < 0.57; MMP-2: 0.73 < r< 0.89). More importantly, however, were the levels over time with the different analytes. As shown in Figure 1, the patterns of the MMP levels over time are very similar for the different analytes, peaking for MMP-9 (Fig. 1A) and decreasing for MMP-2 (Fig. 1B); this was less impressive for the EDTA-levels, which were consistently low for both MMPs.FIGURE 1. (A) MMP-9 and (B) MMP-2 levels in the different analytes of liver transplant patients. Each point represents the mean ± SEM of seven patients/samples at different time points during follow-up (: 1 , pretransplantation; 2 , within 5 days posttransplantation; 3 , at 1 month posttransplantation; 4 , at 3 months posttransplantation; 5 , at 6 months posttransplantation; and 6 , at 1 year of follow-up).These observations do indicate, in our opinion, that the type of analyte does affect the level of the MMPs; with appropriate processing, the dynamics of the levels over time within the different analytes do show a similar pattern. In addition, the consistency in both the sample preparations and MMP results of the different analytes seems to exclude a variable release of MMP-9 from blood leukocytes due to sampling as the cause of raised levels in serum, obtained without clotting activator, of our liver transplant patients with allograft rejection. Jung also suggested that the raised MMP-9 level in the patients with rejection might be related to increased numbers of white blood cells and/or platelets in the circulation, which release MMP-9. Although we did find a significantly higher serum MMP-9 level at one week after transplantation in the patients with rejection compared to the patients without rejection [233±29 (n = 13) versus 158±16 (n = 20) ng MMP-9/ml, P< 0.05], at the same time point the white blood cell and platelet counts were almost the same for both groups (respectively, 11.9±1.3 versus 11.0±1.3 and 156±24 versus 136±15; all 109/L). These findings support our conviction that the raised serum MMP-9 levels with liver rejection are most likely, though probably not exclusively, derived from infiltrating inflammatory cells within the liver (3). Interestingly, in another study we observed a relation between citrate-plasma MMP-9 levels and severity of ischemia/reperfusion injury of the transplanted liver (4). Thus, the suggestion that sodium-citrate plasma might be the best analyte to be used for MMP assessments, as suggested by Jung and by Mannello et al. (5), might be valid, although we found relatively low preoperative levels for MMP-9 (55±6 ng/ml) and for MMP-2 (782±76 ng/ml) in the 24 liver transplantation patients included in this international study (4), a phenomenon also described in the other studies (2, 5). Since most of our studies on MMPs also include evaluation of their activity, either by zymography or immunosorbent activity assays, we prefer to use serum obtained without clotting activator or citrate-plasma because of the known interference of EDTA and heparin in these kinds of analyses. In conclusion, we concur with Jung that consistency in blood sample preparation is essential in the assessment of the clinical and diagnostic value of circulating MMP levels in relation to pathological disease processes. We do think, however, that when this is executed accurately the type of preanalyte used might be more relevant to the obtained absolute MMP level and to the type of analysis to be subsequently performed (e.g., activity assay versus ELISA) but less to the dynamics of the changes in the MMP parameter under study. Hein W. Verspaget Johan Ph. Kuyvenhoven Bart van Hoek Department of Gastroenterology and Hepatology Leiden University Medical Center Leiden, The Netherlands