Thrombosis Research 127 Suppl.

3 (2011) S21S25

Contents lists available at ScienceDirect

Thrombosis Research
j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / t h r o m r e s

Thrombin generation tests

Elisabetta Castoldi*, Jan Rosing

Departement of Biochemistry, Maastricht University, Maastricht, The Netherlands

article info abstract

Keywords: The recent development of semi-automated methods has revived interest in the thrombin
Thrombin generation generation test, a global assay that measures the overall tendency of a plasma sample to form
Global assay thrombin after initiation of coagulation. The thrombin generation curve, which is characterised by a
Calibrated Automated Thrombography lag phase followed by the formation and subsequent inhibition of thrombin, reects all three phases
(CAT) of coagulation (initiation, propagation and termination). However, the specic contribution of each
coagulation factor or inhibitor to the assay outcome depends on the reaction conditions used (e.g.
tissue factor concentration used to trigger coagulation, addition of thrombomodulin or activated
protein C). Although several studies have shown a correlation between thrombin generation and
the risk of bleeding or venous thrombosis, the application of thrombin generation assays to clinical
decision-making is still hampered by standardisation problems. The present paper discusses these
issues with particular reference to Calibrated Automated Thrombography.
2011 Elsevier Ltd. All rights reserved.

Introduction by plasma protease inhibitors). The resulting thrombin generation

Given the complexity of the coagulation system, which involves
several inter-related procoagulant and anticoagulant pathways,
measuring the plasma levels and/or activities of individual proteins
has little utility in the risk assessment and clinical management
of hypo- and hypercoagulable states. In fact, patients with the
same coagulation factor deciency (e.g. haemophilia A) may present
with different bleeding tendencies and carriers of a particular
thrombophilic defect (e.g. factor V Leiden) often experience
different thrombosis risks, depending on additional genetic and
environmental modulators. Moreover, acquired conditions such as
pregnancy and oral contraceptive use modify the levels of several
coagulation factors and inhibitors at the same time. Hence the need
for global assays evaluating the overall coagulation function in order
to reliably estimate the bleeding/thrombosis risk in each individual
patient.
Thrombin generation assays measure the ability of a plasma
sample to generate thrombin following in vitro activation of
coagulation with tissue factor (TF) or another trigger. In contrast
to the classical clotting assays (PT and APTT), which only probe
the initiation phase of coagulation (where the rst traces of
thrombin, necessary for plasma to clot, are formed), thrombin
generation assays also probe the propagation phase (where the bulk
of thrombin is generated via feedback loops on factors V, VIII and XI)
and the termination phase (where thrombin formation is shut down
by the anticoagulant pathways and all thrombin activity is inhibited
* Corresponding author. Elisabetta Castoldi. Department of
Biochemistry, Maastricht University, P.O. Box 616, 6200 MD Maastricht,
The Netherlands. Tel.: +31 43 3884160; fax: +31 43 3884159.
E-mail address: e.castoldi@maastrichtuniversity.nl (E. Castoldi).
curve therefore reects and integrates all pro- and anticoagulant
reactions that regulate the formation and inhibition of thrombin.
Since thrombin is the nal enzyme and the major effector of the
coagulation process, thrombin generation assays show potential as
global tests of plasma coagulability [1].
The rst thrombin generation test was introduced in 1953. This
assay involved regular subsampling from the activated plasma
mixture and quantication of thrombin in each subsample by
measuring its ability to clot brinogen, a very laborious and
rather imprecise technique. Over time, brinogen was replaced by
synthetic thrombin substrates, and subsampling eventually gave
way to the continuous measurement of thrombin activity via low-
afnity chromogenic or uorogenic thrombin substrates added
directly to the plasma. As a result of successive improvements,
several different methods to measure thrombin generation have
been developed and used in different studies. Here, we are going
to focus on the method with which we have the most experience,
i.e. Calibrated Automated Thrombography (CAT).
The CAT assay and the thrombin generation curve
The CAT assay [2,3] relies on a low-afnity uorogenic substrate
(Z-Gly-Gly-Arg-AMC) to continuously monitor thrombin activity
in plasma. Since uorescence is not disturbed by the turbidity
changes associated with clot formation, the test can be performed
in whole (not debrinated) platelet-poor or platelet-rich plasma,
close to the in vivo situation. However, appropriate calibration is
needed to correct for inner-lter effects, i.e. the quenching of the
uorescence signal by substrate molecules that have already been
converted and by the plasma colour (which differs among individual

0049-3848 /$ see front matter 2011 Elsevier Ltd. All rights reserved.
S22 E. Castoldi, J. Rosing / Thrombosis Research 127 (2011) S21S25

Measurement well Calibration well

TF/PL Thrombin calibrator

CaCl2 / fluorogenic substrate CaCl2 / fluorogenic substrate

Plasma Plasma

Raw fluorescence data Raw fluorescence data

400 700
600
Fluorescence (RFU)

Fluorescence (RFU)
300 500
400
200
300
Correction for
100 inner-filter 200
effects and 100
substrate
0 consumption 0
0 10 20 30 40 0 10 20 30 40
Time (min) Time (min)

Correction for 2M-thrombin

First derivative

Thrombin generation curve

250

200
Thrombin (nM)

Peak
150 time

ETP Peak
height
100

50 Lag
time

0
0 5 10 15 20 25
Time (min)

Fig. 1. CAT assay principle and parameters of the thrombin generation curve. The CAT assay requires two uorescence measurements in the same plasma. In the measurement
well coagulation is triggered with tissue factor (TF), phospholipids (PL) and CaCl2 ; in the calibration well only the thrombin calibrator (i.e. a known amount of a2 M-thrombin)
and CaCl2 are added to plasma. Thrombin activity in both wells is monitored continuously via conversion of a low-afnity uorogenic substrate added to plasma. The
thrombin generation curve is obtained by taking the rst derivative the uorescence recorded in the measurement well after (i) correction for inner lter effects and
substrate consumption (based on the uorescence measured in the calibration well); and (ii) subtraction of the uorescence signal deriving from a2 M-thrombin (based on
a mathematical algorithm). The thrombin generation curve can be described in terms of lag time, peak time, peak height and area under the curve (endogenous thrombin
potential, ETP).

plasma samples). The assay is performed in a microtitre plate and
uorescence readings are automatically converted into thrombin
generation curves by dedicated software.
Technically, each CAT test requires two uorescence measure-
ments in the same plasma (Fig. 1). In one well (the measurement
well) TF and synthetic phospholipids vesicles are added to plasma
to initiate coagulation and induce thrombin formation; in a second
well (the calibration well) a known amount of substrate-converting
activity (the thrombin calibrator) is added to plasma without
activating coagulation. The thrombin calibrator consists of thrombin
bound to a2 -macroglobulin (a2 M-thrombin), a form of thrombin
that can not be inhibited by plasma protease inhibitors. A mixture
of CaCl2 and uorogenic substrate is subsequently dispensed in both
wells and the developing uorescence is recorded in real time by
a uorometer. The uorescence tracing in the measurement well is
produced by the thrombin generated in plasma following initiation
of coagulation and (after appropriate corrections) forms the basis
for the calculation of the thrombin generation curve. Differently,
in the calibration well the uorogenic substrate is converted at
a constant rate by the added thrombin calibrator. This yields an
(initially) linear uorescence tracing, whose slope provides the
calibration factor needed to convert raw uorescence units (RFU) to
thrombin concentration units (nM). As more substrate is converted,
the uorescence tracing in the calibration well progressively bends
off as a result of inner-lter effects and substrate consumption.
This deviation from linearity is quantied by the software and
used to correct the uorescence recorded in the measurement well.
Moreover, the signal deriving from spurious substrate conversion by
the a2 M-thrombin generated in the measurement well is subtracted
via a mathematical algorithm. The thrombin generation curve is
eventually calculated by taking the rst derivative of the fully
corrected uorescence tracing.
 E. Castoldi, J. Rosing / Thrombosis Research 127 (2011) S21S25 S23

Table 1
Major determinants of thrombin generation parameters (lag time and ETP) in healthy individuals

Determinant SD Beta
1 pM TF 1 pM TF + TM 13.6 pM TF 13.6 pM TF + TM
Dep. variable: Lag time (min), meanSD 5.841.65 5.571.85 2.670.38 4.340.88
Fibrinogen (g/L) 0.54 0.197 0.270 0.292 0.250
Factor V (U/dL) 18.6 n.s. n.s. n.s. 0.344
Factor VII (U/dL) 22.6 0.232 n.s. 0.287 0.406
Factor IX (U/dL) 11.5 0.409 0.414 n.s. 0.411
Free protein S (U/dL) 17.4 0.295 0.355 0.331 n.s.
Free TFPI (ng/mL) 3.05 0.536 0.479 0.254 0.265

Dep. variable: ETP (nMmin), meanSD 995290 299143 1490224 152107

Fibrinogen (g/L) 0.54 0.400 0.258 0.280 n.s.
Prothrombin (U/dL) 18.2 n.s. n.s. 0.420 n.s.
Factor V (U/dL) 18.6 n.s. 0.241 0.224 n.s.
Factor X (U/dL) 15.5 n.s. n.s. n.s. 0.286
Factor XII (U/dL) 20.2 0.302 0.317 n.s. n.s.
Antithrombin (U/dL) 10.1 0.311 0.247 0.455 n.s.
Protein C (U/dL) 19.8 n.s. 0.229 n.s. n.s.
Free protein S (U/dL) 17.4 n.s. n.s. n.s. 0.470
Free TFPI (ng/mL) 3.05 0.399 0.391 0.216 0.393

Standardised regression coefcients are shown. SD, standard deviation of the coagulation factor/inhibitor level in the study
population; TFPI, tissue factor pathway inhibitor; ETP, endogenous thrombin poteintial; n.s., not signicant. Determinants of
peak height (not shown) almost completely overlap with ETP determinants. Table adapted from ref [8].

A typical thrombin generation curve is shown at the bottom of
Fig. 1. The curve is characterised by a short lag phase (initiation),
followed by an explosive burst of thrombin (propagation) that
subsequently disappears due to inhibition by (mainly) antithrombin
(termination). Several parameters can be derived from the thrombin
generation curve, including lag time, peak time, peak height
and area under the curve (endogenous thrombin potential, ETP)
(Fig. 1). The lag time is dened as the time needed for thrombin
concentration to reach 1/6 of the peak concentration and shows a
good correlation with the plasma clotting time. The ETP represents
the total enzymatic work performed by thrombin during the time
that it was active and is generally considered the most predictive
parameter of bleeding/thrombosis risk. Although ETP and peak
height are usually well correlated, peak height is sometimes a more
sensitive indicator of the plasma thrombin-generating capacity,
as it is less readily saturated than ETP at increasing plasma
concentrations of coagulation factors or trigger [4]. Prolonged lag
times and decreased ETP/peak heights indicate a hypocoagulable
(prohaemorrhagic) state. Vice versa, short lag times and high ETP/
peak heights point at a hypercoagulable (prothrombotic) state.
With intra- and inter-assay variabilities of <10% for all read-out
parameters, the thrombin generation assay is quite reproducible,
both in platelet-poor [5] and in platelet-rich plasma [6]. Moreover,
it was shown to yield reasonably consistent results (%CV 1015%)
when the same individuals were tested repeatedly over a one-year
period [7].
Assay determinants
One of the assets of the thrombin generation assay is that
experimental conditions can be easily modied for specic
purposes. For example, coagulation can be initiated by different TF
concentrations (or even a different trigger) and the concentration
and composition of the added lipids can be varied. Corn trypsin
inhibitor (CTI, a specic inhibitor of FXIIa) can also be added to
block contact activation. Moreover, soluble thrombomodulin (TM)
or activated protein C (APC) can be included in the plasma mixture
to challenge the protein C anticoagulant pathway.
Experimental conditions should be chosen carefully, as they
determine which pro- and anticoagulant pathways contribute to the
assay outcome. For example, the TF concentration used to trigger
coagulation determines to what extent the intrinsic coagulation
pathway contributes to thrombin generation and what role the
tissue factor pathway inhibitor (TFPI)/protein S system plays in
inhibiting thrombin formation. In a similar way, the addition of
TM or APC extends the sensitivity of the assay to the protein C
anticoagulant pathway. In other words, the coagulation factors and
inhibitors that actually contribute to shape the thrombin generation
curve (the so-called assay determinants) change according to the
reaction conditions used. To illustrate this concept, we recently
measured thrombin generation under four different conditions
(1 pM TFTM and 13.6 pM TFAPC) as well as the plasma levels
of 15 coagulation-related proteins in 140 healthy individuals,
and used multiple regression analysis to sort out the effects of
individual proteins on thrombin generation parameters [8]. This
study indicated that the strongest determinants of the ETP are
brinogen and TFPI at low TF, but prothrombin and antithrombin at
high TF (Table 1). When APC was included in the reaction mixture,
the ETP was dependent on protein S, TFPI and factor X, whereas the
addition of TM made the system sensitive to protein C (Table 1).
Another study also showed that, when the thrombin generation
assay is performed in platelet-rich plasma, platelet function also
importantly contributes to the assay outcome [9].
Differences in the plasma levels of the assay determinants
account for the large inter-individual variation in thrombin gen-
eration parameters observed in population studies [6,8]. Moreover
they explain why thrombin generation is inuenced by biological
variables like age [8,10], sex [8], body mass index [11], genetic
factors [12] and acquired conditions (e.g. pregnancy [13], oral
contraceptive use [14], etc.), all of which affect the levels of certain
coagulation factors and inhibitors.
Although the existence of several assay variants with different
determinants is often considered an annoying complication, there
are advantages as well. For example, the thrombin generation
curve obtained in a patients plasma may be similar to that
S24 E. Castoldi, J. Rosing / Thrombosis Research 127 (2011) S21S25
Normal control
Thrombin (nM) 400 400
Thrombin (nM)
300 300
200 200
100 100
0 0
0 5 10 15 20
Time (min)
Fig. 2. Thrombin generation with and without APC in different plasma samples. Thrombin generation was determined at 6.8 pM in the absence (closed symbols) and presence
(open symbols) of 5 nM puried APC in platelet-poor plasma from a normal control, a factor V Leiden heterozygous carrier and a pill user. In the absence of APC, thrombin
generation is the same in the normal control and the factor V Leiden carrier, and only slightly elevated in the pill user. Differently, in the presence of APC thrombin generation
is considerably higher in the factor V Leiden carrier and in the pill user than in the normal control. This example illustrates the concept that not all experimental conditions
are equally suitable to demonstrate the presence of a hypercoagulable state in an individuals plasma.
of control plasma under certain experimental conditions, but
abnormal under other conditions, depending on the determinants
involved. Comparing thrombin generation curves obtained under
different experimental conditions in the same patients sample may
therefore help pinpoint which coagulation pathway or protein is
responsible for the patients phenotype. We have recently used
this approach to identify residual platelet factor V and low plasma
TFPI levels as rescue mechanisms in patients with undetectable
plasma factor V and a paradoxically mild bleeding tendency [15,16].
Furthermore, different research or clinical questions may be best
addressed with different assay variants, and knowledge of the
assay determinants may guide the choice of the most suitable
experimental conditions.
Research or clinical tool?
Thrombin generation assays have enjoyed ever increasing popu-
larity since the recent introduction of semi-automated methods,
and they are currently performed in numerous laboratories all
over the world. Although they are mainly being employed for
research purposes, potential clinical applications have also been
explored (reviewed in refs [1,17].). In particular, several studies
have investigated the correlation between thrombin generation
parameters and the risk of bleeding or venous thrombosis.
On the haemorrhagic side it has been reported that, in
several bleeding disorders, thrombin generation in platelet-poor
plasma correlates well with coagulation factor levels and bleeding
tendency [17,18], whereas the ETP determined in platelet-rich
plasma may even discriminate between mild and severe bleeders
among haemophilic patients with equally undetectable factor VIII or
IX levels [19]. In addition, thrombin generation tests may be useful
in monitoring haemophilia treatment with inhibitor bypassing
agents (reviewed in ref [17].).
On the thrombotic side, the ETP and/or peak height of thrombin
generation determined under various experimental conditions were
found to be elevated in carriers of virtually all thrombophilic
defects, including factor V Leiden [20] (Fig. 2), the prothrom-
bin G20210A mutation [21,22], antithrombin deciency [23] and
protein S deciency [24], as well as in acquired conditions that
predispose to venous thrombosis, such as pregnancy [13] and oral
contraceptive use [14] (Fig. 2). Therefore, it is not surprising that
several epidemiological studies have found a correlation between
elevated thrombin generation and the risk of venous thrombosis,
both retrospectively for the rst thrombotic event [25,26] and
prospectively for recurrences after the rst event [27,28]. Thrombin
generation tests performed at low TF in the presence of TM were the
most predictive, which is not surprising considering the fact that
these experimental conditions are sensitive to most known genetic
and acquired risk factors of venous thrombosis.
FV Leiden heterozygote Pill user
400
Thrombin (nM)
300
200
100
0
0 5 10 15 20 0 5 10 15 20
Time (min) Time (min)
Overall, these ndings are encouraging and suggest that thrombin
generation assays might provide useful information to guide
clinical decisions with respect to prevention and treatment of
coagulation disorders in individual patients. However, several
standardisation problems still prevent the widespread application
of thrombin generation assays in clinical settings [17]. First of
all, blood collection, plasma preparation and other pre-analytical
variables may signicantly inuence the assay outcome (e.g.
by promoting contact activation and/or the release of platelet
microparticles), especially when thrombin generation is measured
at low concentrations of TF and lipids. In addition, the heterogeneity
of assay conditions and protocols, employing different reagents
at different concentrations, makes it presently difcult (if not
impossible) to establish reference ranges for thrombin generation
parameters and to compare results between laboratories [29].
However, assay variability deriving from the random effects of
contact activation can be signicantly reduced by the addition
of CTI, which is recommended whenever thrombin generation
is initiated with a low TF concentration [29,30]. Moreover, the
inter-laboratory variability can be considerably improved by
standardising reagents and conditions and by normalising the ETP
of each sample against the ETP of a reference plasma measured in
parallel [29].
Conclusions
Thrombin generation assays probe pro- and anticoagulant pathways
and may be useful as indicators of overall plasma coagulability,
both in haemorrhagic and bleeding disorders. Although their
clinical application is still hampered by poor standardisation and
excessive inter-laboratory variability, standardisation of protocols
and reagents and normalisation of results against a reference
plasma may help minimise these problems in the future. Since
no assay variant is optimally sensitive to all coagulation factors
and inhibitors, different clinical questions will probably require
to measure thrombin generation under different experimental
conditions.
Acknowledgements
E. Castoldi is supported by a VIDI grant (nr. 91776312) from the
Dutch Organisation for Scientic Research (NWO).
Conict of Interest Statement
The authors have no conicts of interest to report.
References
[1] Baglin T. The measurement and application of thrombin generation. Br J
Haematol 2005;130:65361.
 E. Castoldi, J. Rosing / Thrombosis Research 127 (2011) S21S25
[2] Hemker HC, Giesen P, AlDieri R, Regnault V, de Smed E, Wagenvoord R, et al. The
calibrated automated thrombogram (CAT): a universal routine test for hyper-
and hypocoagulability. Pathophysiol Haemost Thromb 2002;32:24953.
[3] Hemker HC, Al Dieri R, De Smedt E, Beguin S. Thrombin generation,
a function test of the haemostatic-thrombotic system. Thromb Haemost
2006;96:55361.
[4] Duchemin J, Pan-Petesch B, Arnaud B, Blouch MT, Abgrall JF. Inuence of
coagulation factors and tissue factor concentration on the thrombin generation
test in plasma. Thromb Haemost 2008;99:76773.
[5] Spronk HM, Dielis AW, De Smedt E, van Oerle R, Fens D, Prins MH, et al.
Assessment of thrombin generation II: Validation of the Calibrated Automated
Thrombogram in platelet-poor plasma in a clinical laboratory. Thromb Haemost
2008;100:3624.
[6] Gerotziafas GT, Depasse F, Busson J, Leem L, Elalamy I, Samama MM. Towards
a standardization of thrombin generation assessment: the inuence of tissue
factor, platelets and phospholipids concentration on the normal values of
Thrombogram-Thrombinoscope assay. Thromb J 2005;3:16.
[7] Dargaud Y, Francillon S, Negrier C. Intraindividual thrombin generation
measurement variability in healthy adults over a one year period. Thromb Res
2009;124:2378.
[8] Dielis AW, Castoldi E, Spronk HM, van Oerle R, Hamulyak K, Ten Cate H, et al.
Coagulation factors and the protein C system as determinants of thrombin
generation in a normal population. J Thromb Haemost 2008;6:12531.
[9] Vanschoonbeek K, Feijge MA, Van Kampen RJ, Kenis H, Hemker HC, Giesen PL,
et al. Initiating and potentiating role of platelets in tissue factor-induced
thrombin generation in the presence of plasma: subject-dependent variation
in thrombogram characteristics. J Thromb Haemost 2004;2:47684.
[10] Haidl H, Cimenti C, Leschnik B, Zach D, Muntean W. Age-dependency
of thrombin generation measured by means of calibrated automated
thrombography (CAT). Thromb Haemost 2006;95:7725.
[11] Fritsch P, Kleber M, Rosenkranz A, Fritsch M, Muntean W, Mangge H,
et al. Haemostatic alterations in overweight children: Associations between
metabolic syndrome, thrombin generation, and brinogen levels. Atheroscle-
rosis 2010:212;6505.
[12] Segers O, van Oerle R, ten Cate H, Rosing J, Castoldi E. Thrombin generation
as an intermediate phenotype for venous thrombosis. Thromb Haemost 2010;
103:11422.
[13] Rosenkranz A, Hiden M, Leschnik B, Weiss EC, Schlembach D, Lang U,
et al. Calibrated automated thrombin generation in normal uncomplicated
pregnancy. Thromb Haemost 2008;99:3317.
[14] Tchaikovski SN, van Vliet HA, Thomassen MC, Bertina RM, Rosendaal FR,
Sandset PM, et al. Effect of oral contraceptives on thrombin generation
measured via calibrated automated thrombography. Thromb Haemost 2007;
98:13506.
[15] Duckers C, Simioni P, Spiezia L, Radu C, Gavasso S, Rosing J, et al. Low plasma
levels of tissue factor pathway inhibitor in patients with congenital factor V
deciency. Blood 2008;112:361523.
[16] Duckers C, Simioni P, Spiezia L, Radu C, Dabrilli P, Gavasso S, et al.
Residual platelet factor V ensures thrombin generation in patients with
severe congenital factor V deciency and mild bleeding symptoms. Blood
2010;115:87986.
S25
[17] van Veen JJ, Gatt A, Makris M. Thrombin generation testing in routine clinical
practice: are we there yet? Br J Haematol 2008;142:889903.
[18] Al Dieri R, Peyvandi F, Santagostino E, Giansily M, Mannucci PM, Schved JF,
et al. The thrombogram in rare inherited coagulation disorders: its relation to
clinical bleeding. Thromb Haemost 2002;88:57682.
[19] Santagostino E, Mancuso ME, Tripodi A, Chantarangkul V, Clerici M,
Garagiola I, et al. Severe hemophilia with mild bleeding phenotype:
molecular characterization and global coagulation prole. J Thromb Haemost
2010;8:73743.
[20] Simioni P, Castoldi E, Lunghi B, Tormene D, Rosing J, Bernardi F. An
underestimated combination of opposites resulting in enhanced thrombotic
tendency. Blood 2005;106:23635.
[21] Kyrle PA, Mannhalter C, Beguin S, Stumpen A, Hirschl M, Weltermann A,
et al. Clinical studies and thrombin generation in patients homozygous or
heterozygous for the G20210A mutation in the prothrombin gene. Arterioscler
Thromb Vasc Biol 1998;18:128791.
[22] Castoldi E, Simioni P, Tormene D, Thomassen MC, Spiezia L, Gavasso S,
et al. Differential effects of high prothrombin levels on thrombin generation
depending on the cause of the hyperprothrombinemia. J Thromb Haemost
2007;5:9719.
[23] Alhenc-Gelas M, Canonico M, Picard V. Inuence of natural SERPINC1 mutations
on ex vivo thrombin generation. J Thromb Haemost 2010;8:8458.
[24] Castoldi E, Maurissen LF, Tormene D, Spiezia L, Gavasso S, Radu C, et al.
Similar hypercoagulable state and thrombosis risk in type I and type III protein
S-decient individuals from families with mixed type I/III protein S deciency.
Haematologica 2010;95:156371.
[25] Dargaud Y, Trzeciak MC, Bordet JC, Ninet J, Negrier C. Use of calibrated auto-
mated thrombinography +/ thrombomodulin to recognise the prothrombotic
phenotype. Thromb Haemost 2006;96:5627.
[26] van Hylckama Vlieg A, Christiansen SC, Luddington R, Cannegieter SC,
Rosendaal FR, Baglin TP. Elevated endogenous thrombin potential is associated
with an increased risk of a rst deep venous thrombosis but not with the risk
of recurrence. Br J Haematol 2007;138:76974.
[27] Tripodi A, Legnani C, Chantarangkul V, Cosmi B, Palareti G, Mannucci PM.
High thrombin generation measured in the presence of thrombomodulin is
associated with an increased risk of recurrent venous thromboembolism.
J Thromb Haemost 2008;6:132733.
[28] Besser M, Baglin C, Luddington R, van Hylckama Vlieg A, Baglin T. High rate
of unprovoked recurrent venous thrombosis is associated with high thrombin-
generating potential in a prospective cohort study. J Thromb Haemost 2008;6:
17205.
[29] Dargaud Y, Luddington R, Gray E, Negrier C, Lecompte T, Petros S, et al. Effect
of standardization and normalization on imprecision of calibrated automated
thrombography: an international multicentre study. Br J Haematol 2007;139:
3039.
[30] Luddington R, Baglin T. Clinical measurement of thrombin generation
by calibrated automated thrombography requires contact factor inhibition.
J Thromb Haemost 2004;2:19549.