Plasma Serum Amyloid A Levels Are Increased In Venous Thrombosis Patients and Are Correlated with Blood Coagulability

Abstract 155

Inflammation is well correlated with cardiovascular diseases, but its relationship to venous thromboembolic disease (VTE) is less well established. One well studied inflammatory marker, C-reactive protein (CRP), has been inconsistently linked to VTE and this association is controversial. First, we explored the relationship between serum amyloid A (SAA), an inflammatory biomarker, and VTE because SAA can induce tissue factor (TF) expression and release pro-inflammatory cytokines, activities that might imply a linkage between SAA elevation and thrombotic disease. Second, we probed the influence of SAA on blood coagulability. Plasma SAA levels were determined by ELISA for Scripps VTE registry subjects (113 VTE patients and 113 matched controls). Inclusion criteria for this study included age at thrombosis < 55 years, > 3 months since diagnosis of VTE, a life expectancy of > 3 years, no cancer, and no lipid lowering medications. VTE patients had significantly higher median levels of SAA than controls (3.70 vs. 2.12 μ g/ml, p<0.0001). Odds ratio (OR) calculation based on SAA values > 90th percentile of controls showed that elevated SAA was associated with VTE with an OR value of 6.9 (95% CI: 3.3–14). Since SAA is associated with other inflammation parameters, we measured two other acute-phase inflammatory proteins. SAA was correlated with CRP and fibrinogen (r=0.32, p<0.0001 and r=0.37, p<0.0001, respectively). To test if the association of elevated plasma SAA levels with VTE was primary or, alternatively, only secondary to the association of SAA with other acute-phase proteins (e.g., CRP or fibrinogen), we adjusted the OR value (SAA > 90th percentile of controls) for CRP and fibrinogen. However, even after this adjustment for these two inflammation biomarkers, the OR for association of SAA with VTE was not altered (adjusted OR=6.8, 95%CI: 3.1–15). Thus, elevated plasma SAA (> 90th percentile of controls) is a strikingly strong biomarker for VTE that appears to be independent of two other well studied inflammatory biomarkers. To assess the potential direct relationship between SAA in plasma and blood coagulation reactions, we obtained data using endogenous thrombin potential assays and other coagulation assays. Plasma SAA levels from individual donors were very well correlated with parameters of TF-induced thrombin generation in plasma, showing that endogenous plasma SAA levels were associated with plasma coagulability. When added to plasma, purified recombinant SAA dose-dependently enhanced TF-induced thrombin generation. This procoagulant activity of endogenously added SAA that was seen in TF-induced thrombin generation assays was not affected by corn trypsin inhibitor, a specific inhibitor of factor XIIa, implying that contact activation was not mediating this SAA procoagulant activity. However, the procoagulant activity of purified recombinant SAA added to plasma was not seen when factor XI deficient plasma was used for assays, suggesting that SAA exerts procoagulant effects in plasma that require factor XI, i.e., effects that involve factor XI activation and/or expression of factor XIa activity. Further studies are needed to clarify details for the procoagulant effects of SAA on coagulation pathways. In conclusion, these results show that elevated plasma levels of SAA are strongly linked to VTE in adults (< 55 yrs old) and imply that SAA itself is a potential enhancer of thrombin generation in plasma acting via factor XI and the intrinsic coagulation pathway.