Abstract

Introduction: We have recently shown that low TFPI is a weak risk factor for deep vein thrombosis (DVT)(

Dahm et al
Blood
2003
;
101
;
4387
–92
). Plasma contains free, full-length TFPI (FL-TFPI) and truncated and lipoprotein associated TFPI. Since free, FL-TFPI has a much stronger ability to prolong clotting time in diluted prothrombin time (dPT) assay than other types of TFPI, it has been suggested that FL-TFPI biologically plays a more important role than other forms of TFPI.

Aims: To determine the fraction of free, FL-TFPI in plasma and to determine the role of TFPI anticoagulant activity in the prevention of DVT.

Materials and Methods:Normalized TFPI Anticoagulant Activity (n-TFPIac) Ratio was assayed using a dPT assay after incubation of plasma in the absence and the presence of neutralizing anti-TFPI antibodies. Results were expressed as a ratio with dPT in the presence of anti-TFPI as the denominator. The ratio was normalized against a ratio obtained with a reference plasma within each run. TFPI chromogenic substrate (TFPIcs) activity assay was determined by the quantification of residual TF/FVIIa catalytic activity after the incubation of diluted plasma (containing TFPI) with TF, FVIIa, and FXa. TFPI antigen assays: TFPI free antigen (full-length TFPI) and TFPI total antigen (full-length + truncated TFPI) were assayed with commercial kits from Stago, France. Bound TFPI was calculated as the difference between TFPI total antigen and TFPI free antigen. Study population: Individuals included in The Leiden Thrombophila Study (LETS), which is a case-control study of 474 patients with DVT and 474 controls. 363 controls and 362 cases were available for the OR calculations and 473 controls for the estimation of the fraction of free, FL-TFPI. Statistics: The fraction of free, FL-TFPI was calculated by dividing TFPI free antigen by TFPI total antigen. Odds ratios (OR) for DVT were calculated for individuals with TFPI values below the 10th percentile as compared with those above.

Results: The fraction of free, FL-TFPI in plasma was 19%, but with large variations due to hormonal state (table 1). Normalized TFPIac ratio below the 10th percentile gave an OR of 1.5 (95% CI 0.97-2.4) for DVT, which was comparable to the ORs obtained with other TFPI assays. Individuals with low TFPI in both activity assays had an OR of 5.9 (95% CI 1.7–20) for DVT (table 2).

Conclusion: Approximately 20% of TFPI in plasma was free, FL-TFPI. Low n-TFPIac ratio was a weak risk factor for DVT, but was not a stronger risk factor than low TFPI in the other assays. However, combined low n-TFPIac ratio and low TFPIcs activity seemed to be a strong risk factor for DVT.

Table 1. Mean (95% CI) plasma fraction of TFPI free antigen in controls

OC users (n=54)OC nonusers (n=99)Postmenopausal women (n=89)Men (n=201)All (n=473)
TFPI Free Antigen/TFPI total TFPI antigen 0.125 (0.115–0.135) 0.175 (0.165–0.184) 0.197 (0.186–0.208) 0.206 (0.199–0.209) 0.187 (0.182–0.192) 
OC users (n=54)OC nonusers (n=99)Postmenopausal women (n=89)Men (n=201)All (n=473)
TFPI Free Antigen/TFPI total TFPI antigen 0.125 (0.115–0.135) 0.175 (0.165–0.184) 0.197 (0.186–0.208) 0.206 (0.199–0.209) 0.187 (0.182–0.192) 

Table 2. OR (95% CI) for DVT for different TFPI parameters below the 10th percentile

nTFPIac ratioTFPI free antigenTFPI total antigenTFPIcs activityBound TFPIn-TFPIac ratio + TFPIcs activity
1.5 (0.97–2.4) 1.3 (0.83–2.1) 1.3 (0.85–2.1) 1.2 (0.75–1.9) 1.4 (0.90–2.2) 5.9 (1.7–20) 
nTFPIac ratioTFPI free antigenTFPI total antigenTFPIcs activityBound TFPIn-TFPIac ratio + TFPIcs activity
1.5 (0.97–2.4) 1.3 (0.83–2.1) 1.3 (0.85–2.1) 1.2 (0.75–1.9) 1.4 (0.90–2.2) 5.9 (1.7–20) 

Author notes

Corresponding author