Thrombin generation and fibrin clot structure

Generation of a hemostatic clot requires thrombin-mediated conversion of fibrinogen to fibrin. Previous in vitro studies have demonstrated that the thrombin concentration present at the time of gelation profoundly influences fibrin clot structure. Clots formed in the presence of low thrombin concentrations are composed of thick fibrin fibers and are highly susceptible to fibrinolysis; while, clots formed in the presence of high thrombin concentrations are composed of thin fibers and are relatively resistant to fibrinolysis. While most studies of clot formation have been performed by adding a fixed amount of purified thrombin to fibrinogen, clot formation in vivo occurs in a context of continuous, dynamic changes in thrombin concentration. These changes depend on the local concentrations of pro- and anti-coagulants and cellular activities. Recent studies suggest that patterns of abnormal thrombin generation produce clots with altered fibrin structure and that these changes are associated with an increased risk of bleeding or thrombosis. Furthermore, it is likely that clot structure also contributes to cellular events during wound healing. These findings suggest that studies explicitly evaluating fibrin formation during in situ thrombin generation are warranted to explain and fully appreciate mechanisms of normal and abnormal fibrin clot formation in vivo.     

Elevated prothrombin results in clots with an altered fiber structure: a possible mechanism of the increased thrombotic risk

Individuals with elevated prothrombin levels are at increased risk of venous thrombosis. To understand the mechanism behind this observation, we studied the effect of prothrombin concentration on thrombin generation and fibrin clot structure. The pattern of thrombin generation was directly related to the prothrombin level at all concentrations tested. From 0% to 300% of normal plasma levels of prothrombin, increasing the prothrombin concentration increased the initial rate, peak, and total amount of thrombin generated. Importantly, fibrin clot structure was also affected by the prothrombin concentration. Fibrin clots made from prothrombin concentrations less than 10% of plasma levels were weak and poorly formed. Fibrin clots made at 10% to 100% of plasma levels of prothrombin had similar fiber structures (mass-to-length ratio; mu). However, the fiber mass-to-length ratio decreased with increasing prothrombin levels more than 100% of plasma levels, in a dose-dependent manner. These results suggest that increased levels of prothrombin alter thrombin generation and clot structure. Specifically, elevated prothrombin levels produce clots with reduced fibrin mass-to-length ratios compared with normal clots. We hypothesize that this alteration in fibrin clot structure is an important determinant of the risk of thrombosis.     

Stabilization of fibrin clots by activated prothrombin complex concentrate and tranexamic acid in FVIII inhibitor plasma

Summary. Defective hemostasis in haemophilia patients with FVIII inhibitors results in a dramatic decrease in thrombin generation forming unstable fibrin clots that are susceptible to fibrinolyisis. In this study we tested whether the combination of plasma derived activated prothrombin complex concentrate (pd‐aPCC) with tranexamic acid (TXA) may improve fibrin clot stability in FVIII inhibitor plasma. A microplate assay for clot lysis time was used to assess clot stability in FVIII inhibitor plasma. The effect of pd‐aPCC on clot stability was first tested using the commercial FVIII inhibitor plasma. TXA (5 ～ 10 mg mL^?1) increased clot lysis time, but pd‐aPCC (0.25 ～ 1.0 U mL^?1) had no effect on it. The combination of pd‐aPCC and TXA significantly increased clot lysis time compared with TXA alone. The effect appeared to be limited to fibrin clot resistance to fibrinolysis, as TXA was found to have no effect on thrombin generation induced by pd‐aPCC. The effect of pd‐aPCC and TXA on clot stability was then tested and verified in plasma samples from ten patients with severe haemophilia A and inhibitors. The combination of TXA (10 mg mL^?1) and pd‐aPCC (0.5 U mL^?1) significantly increased clot lysis time compared to TXA alone. Our results suggest that the combination of pd‐aPCC with TXA improves clot stability in FVIII inhibitor plasma without additional increases in thrombin generation.     

High dose factor VIIa improves clot structure and stability in a model of haemophilia B

Factor IX (FIX) deficiency results in haemophilia B and high dose recombinant activated factor VII (rFVIIa) can decrease bleeding. Previously, we showed that FIX deficiency results in a reduced rate and peak of thrombin generation. We have now used plasma and an in vitro coagulation model to examine the effect of these changes in thrombin generation on fibrin clot structure and stability. Low FIX delayed the clot formation onset and reduced the fibrin polymerisation rate. Clots formed without FIX were composed of thicker fibrin fibres than normal. rFVIIa shortened the clot formation onset time and improved the fibre structure of haemophilic clots. We also examined clot formation in the presence of a fibrinolytic challenge by including tissue plasminogen activator or plasmin in the reaction milieu. In these assays, normal FIX levels supported clot formation; however, clots did not form in the absence of FIX. rFVIIa partially restored haemophilic clot formation. These results were independent of the effects of the thrombin-activatable fibrinolysis inhibitor. Our data suggest that rFVIIa enhances haemostasis in haemophiliacs by increasing the thrombin generation rate to both promote formation of a structurally normal clot and improve clot formation and stability at sites with high endogenous fibrinolytic activities.     

Polyphosphate enhances fibrin clot structure

Polyphosphate, a linear polymer of inorganic phosphate, is present in platelet dense granules and is secreted on platelet activation. We recently reported that polyphosphate is a potent hemostatic regulator, serving to activate the contact pathway of blood clotting and accelerate factor V activation. Because polyphosphate did not alter thrombin clotting times, it appeared to exert all its procoagulant actions upstream of thrombin. We now report that polyphosphate enhances fibrin clot structure in a calcium-dependent manner. Fibrin clots formed in the presence of polyphosphate had up to 3-fold higher turbidity, had higher mass-length ratios, and exhibited thicker fibers in scanning electron micrographs. The ability of polyphosphate to enhance fibrin clot turbidity was independent of factor XIIIa activity. When plasmin or a combination of plasminogen and tissue plasminogen activators were included in clotting reactions, fibrin clots formed in the presence of polyphosphate exhibited prolonged clot lysis times. Release of polyphosphate from activated platelets or infectious microorganisms may play an important role in modulating fibrin clot structure and increasing its resistance to fibrinolysis. Polyphosphate may also be useful in enhancing the structure of surgical fibrin sealants.     

Alteration of fibrin network by activated protein C

The antithrombotic plasma enzyme, activated protein C (APC), may play a role in thrombolysis. In vitro, acceleration of clot lysis by APC depends on its ability to inhibit the activation of prothrombin. The effect of APC on the assembly and dispersion of fibrin network was studied using turbidimetry, plasmin digestion of fibrin, and electron microscopy of plasma clots. The addition of APC before clotting but not after clotting accelerated clot lysis. The rate of increase in the turbidity of clotting plasma was reduced by APC. The turbidity of plasma clots containing APC was directly related to the clot lysis time. Fibrin from plasma clots that were formed in the presence of APC yielded less fibrin degradation products than fibrin from clots without added APC. Furthermore, APC reduced the diameter and relative number of fibrin fibers in plasma clots during gel assembly. We propose that APC may enhance the efficacy of thrombolysis by reducing the relative mass of fibrin within maturing thrombi.     

A novel mutation (deletion of Aalpha-Asn 80) in an abnormal fibrinogen: fibrinogen Caracas VI. Consequences of disruption of the coiled coil for the polymerization of fibrin: peculiar clot structure and diminished stiffness of the clot.

An abnormal fibrinogen was identified in a 10-year-old male with a mild bleeding tendency; several years later, the patient developed a thrombotic event. Fibrin polymerization of plasma from the propositus and his mother, as measured by turbidity, was impaired. Plasmin digestion of fibrinogen and thrombin bound to the clot were both normal. The structure of clots from both plasma and purified fibrinogen was characterized by permeability, scanning electron microscopy and rheological measurements. Permeability of patients' clots was abnormal, although some measurements were not reliable because the clots were not mechanically stable. Consistent with these results, the stiffness of patients' clots was decreased approximately two-fold. Electron microscopy revealed that the patients' clots were very heterogeneous in structure. DNA sequencing of the propositus and his mother revealed a new unique point mutation that gives rise to a fibrinogen molecule with a missing amino acid residue at Aalpha-Asn 80. This new mutation, which would disrupt the alpha-helical coiled-coil structure, emphasizes the importance of this part of the molecule for fibrin polymerization and clot structure. This abnormal fibrinogen has been named fibrinogen Caracas VI.     

Hyperhomocysteinemia in patients with pulmonary embolism is associated with impaired plasma fibrinolytic capacity

Hyperhomocysteinemia (HHcy) affects haemostasis and shifts its balance in favour of thrombosis. In vitro and in vivo studies suggested that HHcy may impair fibrinolysis either by influencing the plasma levels of fibrinolytic factors or by altering the fibrinogen structure. We investigated the influence of mild HHcy levels on plasma fibrinolytic potential by using clot lysis time (CLT) and fibrin susceptibility to plasmin-induced lysis in 94 patients with previous pulmonary embolism and no pulmonary hypertension. CLT was measured as lysis time of tissue factor induced clots exposed to exogenous tissue plasminogen activator (t-PA). The rate of in vitro plasmin-mediated cleavage of fibrin β-chain was assessed over a 6-h period on fibrin clots, which were obtained by exposition to thrombin of purified fibrinogen. Homocysteine plasma levels were measured by Abbott Imx immunoassay and we considered as altered the values above 15 μmol/L according to the literature. In 68 patients homocysteine levels were below 15 μmol/L (NHcy) and in 26 they were above (HHcy). Significant differences were observed between the two groups regarding plasma fibrinolytic potential (p = 0.016), TAFIact (expressed as clot lysis ratio) (p = 0.02), t-PA (0.008) and PLG (0.037), but not for the other assessed components. The HHcy-patients had a threefold higher risk to have an impaired fibrinolysis. Instead, a multivariate logistic regression analysis adjusted for significances of univariate showed that HHcy (OR 5.2 95 % CI 1.7–15.9; p = 0.003) and BMI (OR 5.0 95 % CI 1.6–15.9; p = 0.006) resulted independently associated with impaired fibrinolytic activity. HHcy affects TAFI-mediated hypofibrinolysis but not fibrin(ogen) structure or function as documented by fibrin degradation analysis.     

The influence of fibrin crosslinking on the kinetics of urokinase-induced clot lysis

Summary . The effect of fibrin crosslinking on the lysis of plasma clots was investigated with plasma from a patient congenitally deficient in plasma factor XIII (fibrin stabilizing factor). The thrombin-activated plasma factor XIII was found to render clots more resistant to fibrinolysis when urokinase (UK) was used to induce plasminogen activation. Incorporation of UK in the clot by addition to plasma immediately before clotting resulted in a log-log relationship when lysis time was plotted against UK concentration, with greater differences between normal and factor-XIII deficient clots at lower UK concentrations. Addition of UK to the clot externally, after preincubation of the clot, gave linear plots on rectangular coordinates when either plasma or euglobulin fraction was used; and the difference in lysis time between factor-XIII deficient and normal clots was nearly constant over the range of UK concentrations tested. When the fluorescent amine, dansylcadaverine, was used to measure factor-XIII activity quantitatively, plasma clot lysis times were found to be directly proportional to amine-incorporating activity over a wide range of activities at low UK concentrations. The range of proportionality was reduced when UK concentration was increased. Addition of purified factor XIII to the patient's plasma restored the resistance of these clots to within the normal range.     

Patients with paroxysmal nocturnal hemoglobinuria demonstrate a prothrombotic clotting phenotype which is improved by complement inhibition with eculizumab

Paroxysmal nocturnal hemoglobinuria (PNH) is a rare hematological disorder, characterized by complement-mediated intravascular hemolysis and thrombosis. The increased incidence of PNH-driven thrombosis is still poorly understood, but unlike other thrombotic disorders, is thought to largely occur through complement-mediated mechanisms. Treatment with a C5 inhibitor, eculizumab, has been shown to significantly reduce the number of thromboembolic events in these patients. Based on previously described links between changes in fibrin clot structure and thrombosis in other disorders, our aim was to investigate clot structure as a possible mechanism of thrombosis in patients with PNH and the anti-thrombotic effects of eculizumab treatment on clot structure. Clot structure, fibrinogen levels and thrombin generation were examined in plasma samples from 82 patients from the National PNH Service in Leeds, UK. Untreated PNH patients were found to have increased levels of fibrinogen and thrombin generation, with subsequent prothrombotic changes in clot structure. No link was found between increasing disease severity and fibrinogen levels, thrombin generation, clot formation or structure. However, eculizumab treated patients showed decreased fibrinogen levels, thrombin generation and clot density, with increasing time spent on treatment augmenting these antithrombotic effects. These data suggest that PNH patients have a prothrombotic clot phenotype due to increased fibrinogen levels and thrombin generation, and that the antithrombotic effects of eculizumab are, in-part, due to reductions in fibrinogen and thrombin generation with downstream effects on clot structure.     

Fluid phase coagulation events have minimal impact on plasma fibrin structure

Reports of altered fibrin structure in clots formed from factor VIII-deficient plasma have raised the possibility that plasma clots mediated by activation of the fluid phase coagulation system might differ from clots formed by the direct addition of thrombin to plasma. In this study, turbidity measurements were used to compare the assembly and structure of clots formed from platelet-poor plasma by either the addition of thrombin or the exposure of recalcified plasma to glass. When clotted by recalcification, the lag phase before initial increase in turbidity was 10 to 25 times longer than when clotted by the addition of thrombin. Decreasing the ionic strength or increasing the calcium concentration shortened the lag phase. At high calcium concentrations (greater than 25 mM) polymerization was delayed and precipitation was noted. pH had a minimal impact over the range of 7.0 to 7.4. Fibrin fiber mass-length ratios for plasma gels formed by activation of the intrinsic cascade were virtually identical to those in gels formed by the direct addition of thrombin. These studies indicate that fluid phase coagulation events before the production of thrombin have a minimal impact on plasma fibrin structure.     

Fibrinogen gamma-chain splice variant gamma' alters fibrin formation and structure

Fibrinogen gammaA/gamma' results from alternative splicing of mRNA. This variant, which constitutes approximately 8% to 15% of plasma fibrinogen, contains FXIII and thrombin binding sites. Our objective was to investigate whether gammaA/gamma' differs in fibrin formation and structure from the more common variant gammaA/gammaA. Both variants were separated and purified by anion-exchange chromatography. Fibrin formation and clot structure of the variants and unfractionated fibrinogen were investigated by turbidity and scanning electron microscopy (SEM). Thrombin cleavage of fibrinopeptides was analyzed by high-performance liquid chromatography (HPLC). Turbidity analysis showed significantly altered polymerization rates and overall fiber thickness in gammaA/gamma' clots compared with gammaA/gammaA and unfractionated fibrinogen. This finding was consistent with a range of thrombin concentrations. HPLC demonstrated reduced rates of fibrinopeptide B (FpB) release from gammaA/gamma' fibrinogen compared with gammaA/gammaA. Delayed FpB release was associated with delayed lateral aggregation of protofibrils and significant differences were found on SEM, with gammaA/gamma' clots consisting of smaller diameter fibers and increased numbers of branch points compared with both gammaA/gammaA and unfractionated fibrinogen. These results demonstrate that the gammaA/gamma' splice variant of fibrinogen directly alters fibrin formation and structure, which may help to explain the increased thrombotic risk associated with this variant.     

The factor XIII V34L polymorphism accelerates thrombin activation of factor XIII and affects cross-linked fibrin structure

Factor XIII on activation by thrombin cross-links fibrin. A common polymorphism Val to Leu at position 34 in the FXIII A subunit is under investigation as a risk determinant of thrombosis. Because Val34Leu is close to the thrombin cleavage site, the hypothesis that it would alter the function of FXIII was tested. Analysis of FXIII subunit proteolysis by thrombin using sodium dodecyl sulfate-polyacrylamide gel electrophoresis and high-performance liquid chromatography showed that FXIII 34Leu was cleaved by thrombin more rapidly and by lower doses than 34Val. Mass spectrometry of isolated activation peptides confirmed the predicted single methyl group difference and demonstrated that the thrombin cleavage site is unaltered by Val34Leu. Kinetic analysis of activation peptide release demonstrated that the catalytic efficiency (k(cat)/K(m)) of thrombin was 0.5 for FXIII 34Leu and 0.2 (micromol/L)(-1) x sec(-1) for 34Val. Presence of fibrin increased the catalytic efficiency to 4.8 and 2.2 (micromol/L)(-1) x sec(-1), respectively. Although the 34Leu peptide was released at a similar rate as fibrinopeptide A, the 34Val peptide was released more slowly than fibrinopeptide A but more quickly than fibrinopeptide B generation. Cross-linking of gamma- and alpha-chains appeared earlier when fibrin was incubated with FXIII 34Leu than with 34Val. Fully activated 34Leu and 34Val FXIII showed similar cross-linking activity. Analysis of fibrin clots prepared using plasma from FXIII 34Leu subjects by turbidity and permeability measurements showed reduced fiber mass/length ratio and porosity compared to 34Val. The structural differences were confirmed by electron microscopy. These results demonstrate that Val34Leu accelerates activation of FXIII by thrombin and consequently affects the structure of the cross-linked fibrin clot.     

Incorporation of fibrin molecules containing fibrinopeptide A alters clot ultrastructure and decreases permeability

Previous studies have shown that a heterozygous mutation in the fibrinogen Aalpha chain gene, which results in an Aalpha R16C substitution, causes fibrinolytic resistance in the fibrin clot. This mutation prevents thrombin cleavage of fibrinopeptide A from mutant Aalpha R16C chains, but not from wild-type Aalpha chains. However, the mechanism underlying the fibrinolytic resistance is unclear. Therefore, this study investigated the biophysical properties of the mutant fibrin that contribute to fibrinolytic resistance. Fibrin clots made from the mutant fibrinogen incorporated molecules containing fibrinopeptide A into the polymerised clot, which resulted in a 'spiky' clot ultrastructure with barbed fibrin strands. The clots were less stiff than normal fibrin and were cross-linked slower by activated FXIII, but had an increased average fiber diameter, were more dense, had smaller pores and were less permeable. Protein sequencing showed that unclottable fibrinogen remaining in the supernatant consisted entirely of homodimeric Aalpha R16C fibrinogen, whereas both cleaved wild-type alpha chains and uncleaved Aalpha R16C chains were in the fibrin clot. Therefore, fibrinolytic resistance of the mutant clots is probably a result of altered clot ultrastructure caused by the incorporation of fibrin molecules containing fibrinopeptide A, resulting in larger diameter fibers and decreased permeability to fibrinolytic enzymes.     

Comparison of polymerization of ancrod and thrombin fibrin monomers.

The polymerization of thrombin and ancrod fibrin monomers was studied with a standardized technique that evaluated turbidity changes and protein incorporation into the clot. Ancrod fibrin monomers were found to polymerize more slowly and form less turbid clots (at identical protein concentrations). Changes in ionic strength and pH influences ancrod fibrin monomer polymerization to a greater extent than thrombin fibrin monomer polymerization. Benzyltriethylammonium chloride was shown to be a potent inhibitor of fibrin monomer polymerization, with a greater inhibitory effect on ancrod fibrin monomers than on thrombin fibrin monomers. The differences between ancrod and thrombin fibrin may play a role in the infrequent thrombotic complications reported with ancrod therapy.     

Aβ delays fibrin clot lysis by altering fibrin structure and attenuating plasminogen binding to fibrin

Alzheimer disease is characterized by the presence of increased levels of the β-amyloid peptide (Aβ) in the brain parenchyma and cerebral blood vessels. This accumulated Aβ can bind to fibrin(ogen) and render fibrin clots more resistant to degradation. Here, we demonstrate that Aβ(42) specifically binds to fibrin and induces a tighter fibrin network characterized by thinner fibers and increased resistance to lysis. However, Aβ(42)-induced structural changes cannot be the sole mechanism of delayed lysis because Aβ overlaid on normal preformed clots also binds to fibrin and delays lysis without altering clot structure. In this regard, we show that Aβ interferes with the binding of plasminogen to fibrin, which could impair plasmin generation and fibrin degradation. Indeed, plasmin generation by tissue plasminogen activator (tPA), but not streptokinase, is slowed in fibrin clots containing Aβ(42), and clot lysis by plasmin, but not trypsin, is delayed. Notably, plasmin and tPA activities, as well as tPA-dependent generation of plasmin in solution, are not decreased in the presence of Aβ(42). Our results indicate the existence of 2 mechanisms of Aβ(42) involvement in delayed fibrinolysis: (1) through the induction of a tighter fibrin network composed of thinner fibers, and (2) through inhibition of plasmin(ogen)-fibrin binding.     

A new type of congenital dysfibrinogenaemia with defective fibrin lysis—Dusard syndrome: possible relation to thrombosis

S ummary . Congenital dysfibrinogenaemia is described in three members of a family presenting with recurrent thrombosis and in two other young members not yet affected. An abnormality in the polymerization of fibrin monomers was noted. In addition, the pathological fibrin clots were found to be less sensitive to degradation by a post venous occlusion euglobulin solution than normal fibrin. After fibrin clot incubation with lys-plasminogen at different concentrations, the biological activity of plasminogen in patient fibrin clot on S 2251 after SK-addition, was less than that observed with normal fibrin. It is speculated that defective in vivo thrombolysis might explain the recurrent thrombosis observed in this family. This finding represents a new concept in understanding thromboembolic diseases.     

Pharmacologic modulation of thrombin generation associated with human clots by human purified antithrombin alone or in the presence of low molecular weight heparin or unfractionated heparin.

Procoagulant activities associated with human clots may contribute to thrombus extension. We investigate the inhibition of clot-associated factor Xa and thrombin activities by purified human antithrombin either alone or as combination with a low molecular weight heparin (enoxaparin) as compared with unfractionated heparin (UFH). The standard clots were prepared by recalcification of frozen platelet-poor human plasma. Clot-associated thrombin was measured on the clot after clot incubation in recalcified buffer or recalcified prothrombin solution. The enzymatic reaction was measured using a specific substrate for thrombin (CBS 3447). The thrombin concentration was determined both on the clots and in the reaction mixtures. In parallel, prothrombin fragment 1.2 and thrombin-antithrombin complexes (TAT) were measured using enzyme-linked immunosorbent assay methods. We demonstrated that in the presence of purified human prothrombin and antithrombin (AT), a partial inhibition of clot associated thrombin activity correlated with an increase of TAT complexes. However, antithrombin was unable to inhibit thrombin generation induced by the clot-associated factor Xa. Enoxaparin (low molecular weight heparin) and UFH did not enhance clot-bound thrombin inhibition induced by AT. We conclude that clot-bound thrombin is accessible to human antithrombin alone. AT is also able to inhibit thrombin generated by factor Xa-associated clot. However, neither a low molecular weight heparin or UFH enhanced the effect of AT alone.     

Ultrastructure of the interaction between human platelets and polymerizing fibrin within the first minutes of clot formation

In the presence of fibrinogen, thrombin induces the aggregation of platelets, the formation of fibrin, and the retraction of the fibrin network. Since these phenomena are initiated in the first minutes after stimulation, the platelet-fibrin interaction in the early stages of clot formation was investigated. To avoid the formation of high fibrin masses, which may obstruct the morphological evaluation, suspensions of washed platelets were diluted with homologous platelet-poor plasma and stimulated with thrombin. The incubation was stopped by fixation after different times and the clots were studied in series of ultra-thin sections. Between 30 and 60 s after stimulation, polymerizing fibrin was found to be situated predominantly within the contact spaces in aggregates of degranulating platelets. Assembling fibres were seen also in plasmalemmal invaginations located in the close vicinity of the constricting contractile cytoskeleton. Between 5 and 10 min after stimulation, fibres with increasing thickness were observed between the platelets and were internalized into deep-surface invaginations, which remained attached to the outer rim of the contractile cytoskeleton. Surface areas without connection to the cytoskeleton revealed no binding or internalization of fibres. Clots obtained after addition of cytochalasin (36 microM) showed no retraction. Fibrin was found to be bound on the whole surface of discoid and degranulated platelets but no internalization occurred. These results suggest that the association of the clustered fibrin(ogen)-receptor complexes to the contractile cytoskeleton and the formation of a constricting sphere leads to the retraction of the clot by internalization of the fibres.     

Plasma levels of factor XIIa and factor VIIa are increased but not related in primary hyperlipidaemia.

The lipolysis of triglyceride-rich lipoproteins may provide a surface that supports the activation of factor XII (FXII) with subsequent activation of factor VII (FVII). Plasma levels of activated FVII (FVIIa) but not activated FXII (FXIIa) are increased in the post-prandial state when there is a transient increase in triglyceride levels. We compared plasma levels of FXIIa antigen in control subjects (n = 33) and in patients with chronically elevated lipids (primary hyperlipidaemia, n = 49), with FVIIa and markers of thrombin generation. Results are given as median (first and third quartiles). Plasma levels of FXIIa [2.34 (1.68-3.32) ng/ml versus 1.53 (0.93-1.86) ng/ml, P = 0.0002], FVIIa [3.02 (2.15-4.64) ng/ml versus 2.20 (1.66-2.56) ng/ml, P = 0.0004], thrombin-antithrombin complexes [3.08 (2.16-5.54) microg/I versus 2.13 (1.46-2.84) microg/l, P = 0.005] and prothrombin fragment 1 + 2 (Pro F1 + 2) [1.28 (1.08-1.50) nmol/l versus 0.92 (0.65-1.08) nmol/l, P = 0.0001] were increased compared with controls irrespective of the type of hyperlipidaemia. In hyperlipdaemic subjects, levels of Pro F1 + 2 were correlated with FVIIa (r = 0.56, P = 0.0002) and FXIIa (r = 0.31, P = 0.03). These results suggest increased activation of both FVII and FXII in hyperlipidaemic subjects, which correlates with increased thrombin generation. Given the lack of correlation between levels of FXIIa and FVIIa, it remains to be established whether the increase in FXIIa is responsible for increased FVIIa activity in this subject group.