Journal of Thrombosis and Haemostasis, 7: 559–565
ORIGINAL ARTICLE
The pharmacokinetics of idraparinux, a long-acting indirect
factor Xa inhibitor: population pharmacokinetic analysis from
Phase III clinical trials
C . V E Y R A T - F O L L E T , * N . V I V I E R , * M . T R E L L U ,   C . D U B R U C   and G . J . S A N D E R I N K
*sanofi-aventis, Vitry sur Seine, Cedex; and  sanofi-aventis, Chilly-Mazarin, France
To cite this article: Veyrat-Follet C, Vivier N, Trellu M, Dubruc C, Sanderink GJ. The pharmacokinetics of idraparinux, a long-acting indirect factor
Xa inhibitor: population pharmacokinetic analysis from Phase III clinical trials. J Thromb Haemost 2009; 7: 559–65.
Summary. Background: Idraparinux, a long-acting synthetic
pentasaccharide, is a specific antithrombin-dependent inhibitor
of activated factor X that has been investigated in the treatment
and prevention of thromboembolic events. Objectives: To
characterize the population pharmacokinetic profile of idra-
parinux in patients enrolled in van Gogh and Amadeus Phase
III clinical trials. Patients and methods: Idraparinux was
administered once-weekly subcutaneously at a dose of 2.5 mg,
or 2.5 mg (first dose) and then 1.5 mg for patients with severe
renal insufficiency (creatinine clearance <30 mL min)1). A
population pharmacokinetic model was developed using data
from 704 patients with acute deep-vein thrombosis or pulmo-
nary embolism, 1310 patients suffering from atrial fibrillation,
and 40 healthy subjects. Potential covariates analyzed included
demographics (age, sex, weight and ethnicity), and serum
creatinine and creatinine clearance determinations. Results: A
three-compartment model best described idraparinux pharma-
cokinetics, with interindividual variability on clearance, central
volume of distribution, and absorption rate constant; residual
variability was low. Typical clearance, central volume of
distribution, absorption rate constant and volume of distribu-
tion at steady-state were 0.0255 L h)1, 3.36 L, 1.37 h and
30.8 L, respectively. Peak concentration was reached at 2.5 h.
The terminal half-life was 66.3 days and time to steady-state
was 35 weeks. At steady-state, exposures were similar for
patients without and with severe renal impairment receiving
adjusted-dose. Creatinine clearance was the most important
covariate affecting idraparinux clearance. The particular char-
acteristics of idraparinux – rapid onset of action and long-acting
anticoagulant effect – offer interesting clinical perspectives
currently under investigation with idrabiotaparinux, the revers-
ible biotinylated form of idraparinux.
Correspondence: Christine Veyrat-Follet, sanofi-aventis, Global
Metabolism and Pharmacokinetics, 13, Quai Jules Guesdes B.P. 14,
94403 Vitry sur Seine, Cedex, France.
Tel.: +33 1 58 93 85 71; fax: +33 1 58 93 85 67.
E-mail: christine.veyrat-follet@sanofi-aventis.com
Received 18 November 2008, accepted 12 January 2009
 2009 International Society on Thrombosis and Haemostasis
DOI: 10.1111/j.1538-7836.2009.03298.
 
Keywords: atrial fibrillation, idraparinux, population pharma-
cokinetics, stroke, venous thromboembolism.
Introduction
The prevention of venous thromboembolism (VTE) recurrence
following initial treatment of deep-vein thrombosis (DVT) or
pulmonary embolism (PE), and of thromboembolic events
such as stroke in patients with atrial fibrillation (AF), usually
involves the long-term use of vitamin K antagonists (VKAs),
which is often problematic and inconvenient [1]. Idraparinux, a
long-acting synthetic pentasaccharide, is a selective and potent
antithrombin-dependent inhibitor of activated factor X (FXa),
with a very high affinity for antithrombin (Kd = 1.4 ±
0.3 nmol L)1) [2]. After subcutaneous (s.c.) administration
idraparinux is rapidly absorbed (tmax ranged between 2 and
4 h), with its bioavailability approaching 100% [2,3]. No
evidence of metabolism of the parent compound was observed
in humans after a single intravenous administration of [14C]-
idraparinux. Urine was the main excretion route of the
compound and all the radioactivity excreted in urine repre-
sented the unchanged compound. Due to its long half-life,
idraparinux is suitable for s.c. once-weekly dosing. Moreover,
the anticoagulant response produced by idraparinux is highly
predictable, as assessed by its linear pharmacokinetics with
dose-proportionality up to 14 mg [3], and low (<30%)
variability in healthy subjects. Thus, idraparinux offers the
potential for improved convenience in terms of dosing and
monitoring, in the long-term anticoagulant prophylaxis and
treatment of thromboembolic events.
The Phase II dose ranging PERSIST study demonstrated
that s.c. once-weekly administration of 2.5 mg idraparinux was
as effective as warfarin for the secondary prevention of DVT,
and was not associated with major bleeding [4]. Based on these
findings, s.c. once-weekly administration of idraparinux 2.5 mg
was investigated in Phase III clinical trials of patients with acute
symptomatic DVT or PE in the van Gogh trials [5,6] and with
AF in the Amadeus trial [7]. The objectives of the present
analysis were to characterize the population pharmacokinetic
profile of idraparinux in patients who participated in the Phase
560 C. Veyrat-Follet et al
III trials, and to identify covariates that affected the pharma-
cokinetics of idraparinux.
Methods
Trial protocols
The van Gogh trials were multicenter, international, random-
ized, open-label, assessor-blind, non-inferiority studies, which
compared the efficacy and safety of idraparinux with the
combination of heparin and VKA in the treatment of either
acute symptomatic DVT or PE [5]. Patients received either s.c.
once-weekly idraparinux 2.5 mg, or heparin (low-molecular-
weight heparin or unfractionated heparin) followed by
adjusted-dose VKA for 3 or 6 months. The van Gogh
extension trial compared the efficacy and safety of s.c. once-
weekly idraparinux with placebo in the long-term prevention of
symptomatic VTE in patients with symptomatic PE or DVT
who completed 6 months of treatment with VKA or idrapar-
inux [6]. Patients were treated for an additional 6 months
(26 weeks) with once-weekly s.c. idraparinux 2.5 mg or
placebo.
The Amadeus trial was a multicenter, multinational, ran-
domized, open-label, assessor-blind, non-inferiority study
comparing the efficacy and safety of s.c. once-weekly idrapar-
inux 2.5 mg with adjusted-dose VKA in the prevention of
thromboembolic events in patients with AF; patients were
treated for 6 months to 2 years [7]. Patients with severe renal
insufficiency, defined as creatinine clearance (CrCl)
<30 mL min)1, received 1.5 mg idraparinux once-weekly
following an initial dose of 2.5 mg, or 1.5 mg for patients in
the van Gogh extension trial who had previously received
idraparinux.
Data from healthy subjects in two Phase I trials who received
s.c. once-weekly single-dose idraparinux 2.5 mg were included
in this analysis in order to better characterize the structural
model. One study was a single-center, single ascending dose,
double-blind, randomized, placebo-controlled, sequential
group study to assess the tolerability, safety and pharmacoki-
netics of idraparinux in young healthy males, and elderly
healthy male and female subjects. The second study was a
multicenter, single dose, open-label, parallel group, pharma-
codynamic study.
Blood sampling schedules
In the van Gogh DVT and PE trials, samples were collected at
2–4 h after the first idraparinux injection, just prior to the next
idraparinux injection and at the end of week 13, plus at the end
of week 26 for those in the 6-month stratum. No schedule was
prespecified in the van Gogh extension trial. In addition, in all
the van Gogh trials, blood samples were collected in case of
safety or efficacy events. In the Amadeus trial, blood samples
were collected at 3, 12 and 18 months, and at the end of
treatment, with additional collections in case of a suspected
stroke, non-central nervous system systemic embolism, bleed-
ing, myocardial infarction, or VTE event. In the Phase I studies
rich sampling was performed over the 0–648 h time interval.
Blood samples of 5 mL were collected in tubes containing
trisodium citrate and were centrifuged at 3000 ·g for 15 min at
4 C. Plasma was divided into two aliquots and stored at
)20 C until analyzed.
Study assessments
Idraparinux assay Idraparinux concentrations in plasma
were measured by a fully validated biological assay that used
the antithrombin-mediated FXa inhibitory activity of
idraparinux [8; Data on file]. Idraparinux concentrations in
plasma were determined by a chromogenic assay using the
Stachrom Heparin kit (Diagnostica Stago, Inc., Parsippany,
NJ, USA) in an excess of exogenous antithrombin. The
determination of the concentration in a sample in gravimetric
units is achieved through comparison of measured anti-FXa in
the sample with a calibration curve spiked with known
amounts of idraparinux. Plasma concentrations are expressed
as free acid. Quantification of idraparinux in plasma was
unaffected by the presence of EDTA as a blood anticoagulant,
the presence of hemolyzed blood in plasma (3%), or by the
presence of acetaminophen, aspirin, ibuprofen, caffeine,
nicotine, ketoconazole, digoxin, racemic warfarin, or
levonorgestrel. The limit of quantification was 71.4 ng mL)1;
total precision and accuracy were better than 5.47% and
7.11%, respectively.
Clinical covariates Potential clinical covariates in this
pharmacokinetic analysis were subject demographics (age, age
index, i.e. <75 years or ‡75 years, sex, weight, and ethnicity),
and laboratory determinations (serum creatinine and CrCl
computed from serum creatinine concentration, weight, age,
and sex according to the Cockcroft and Gault formula [9]).
Pharmacokinetic analysis
The population pharmacokinetic analysis was performed in
three steps using non-linear mixed-effect modeling (version V
NONMEM, GloboMax, Hanover, MD, USA). A basic
structural pharmacokinetic model was obtained based on
idraparinux concentration data from the Phase III and Phase I
clinical trials detailed above [Data on file]. The main pharma-
cokinetic parameters determined were: clearance, which indi-
cates the rate of elimination from plasma; distribution volumes;
and elimination half-lives. Then an initial exploratory analysis
for covariate effects on individual pharmacokinetic parameters
was performed using SAS software (version 8.2 SAS Institute
Inc, Cary, NC, USA). Univariate analysis was also used.
The relationship between pharmacokinetic parameters and
covariates was investigated based on nonlinear mixed-effect
modeling using the likelihood ratio test to discriminate among
alternative models. Model building was accomplished by a
backward elimination procedure [10]. Finally, the model was
evaluated using visual predictive check.
 2009 International Society on Thrombosis and Haemostasis
 Population pharmacokinetics of idraparinux 561

Results 2000 Van Gogh

1500
Demographics
1000

Concentration (ng mL–1)

In all, idraparinux plasma concentration data from 2027 500
subjects evaluable for PK analysis were included, comprising
0
1987 patients from the Phase III trials in which the model was 0 12 24 36 48 60 72 84 96 108
2000
evaluated, and 40 healthy subjects from the Phase I studies. The Amadeus
demographics of these patients are shown in Table 1. Overall, 1500
the majority of individuals included in this analysis were male, 1000
weighed between 50 kg and <100 kg (81%), were aged 500
<75 years, and were Caucasian. Most individuals had either
0
normal or mildly impaired renal function (83% overall), 15% 0 12 24 36 48 60 72 84 96 108
had moderate and 2% severe renal impairment. The percent- Time (week)
ages of patients aged ‡75 years and average patient weight Fig. 1. Idraparinux concentration time profiles in the van Gogh and
were higher in the Amadeus trial than in the van Gogh trials. Amadeus Phase III clinical trials.

on residual error (16.0% vs. 20%) compared with the two-

Population pharmacokinetic model building
Basic model development was performed using a data set
including 2027 subjects, corresponding to 4112 concentrations,
of which 88.2% were from the Phase III trials. The idraparinux
time-concentration data of the van Gogh and Amadeus trials
populations are shown in Fig. 1.
An average of 2.6 and 1.35 blood samples per patient were
obtained in the Phase III van Gogh and Amadeus trials,
respectively, and 12 blood samples per subject were obtained in
the Phase I trials. Initially, a two-compartment model was
tested with idraparinux time-concentration data. However, a
three-compartment model provided the best fit to the data with
a decreased objective function (‡821 point drop), and a lower
interindividual variability on clearance (20.4% vs. 25.7%) and
compartment model. A more complex residual error model,
involving both a proportional and an additive component,
significantly improved the fit, with a drop of 96 points of the
objective function. Consequently, the three-compartment
model, with interindividual variability on clearance, central
volume of distribution (V2), and absorption rate constant with
mixed-proportional and additive-residual errors, was retained
as the best basic model. The parameters were well estimated
and the observed data were well described by the model as
illustrated in Fig. 2. Furthermore, the predictive check showed
that the developed model was able to simulate data in close
agreement with the actual studiesÕ observations.
In the final model, peak concentration was rapidly achieved
at 2.5 h, typical clearance was 0.0255 L h)1, V2 was 3.36 L,

Table 1 Characteristics of the patient study population

Number of patients (%) Mean ± SD Median (5–95%)

van Gogh Amadeus

(N = 692) (N = 1295) van Gogh Amadeus van Gogh Amadeus

Patient demographics
Age, years 61 ± 17 70 ± 9 63 (28–83) 71 (54–83)
Age index
< 5 years 530 (76.6) 851 (65.7)
‡75 years 162 (23.4) 444 (34.3)
Weight, kg 81 ± 19 86 ± 18 80 (54–115) 84 (60–116)
Sex, Male 367 (53.0) 898 (69.3)
Ethnicity
Caucasian 672 (97.1) 1258 (97.1)
Black 10 (1.5) 9 (0.7)
Oriental 5 (0.7) 7 (0.5)
Other 5 (0.7) 21 (1.6)
Laboratory measurements
Creatinine clearance, mL min)1 88.6 ± 41.7 77.3 ± 30.6 82 (32–172) 73 (37–134)
Renal insufficiency
Severe (10 to <30 mL min)1) 30 (4.3) 13 (1.0)
Moderate (30 to <50 mL min)1) 89 (12.9) 217 (16.8)
Mild (50–80 mL min)1) 215 (31.1) 558 (43.1)
Normal (>80 mL min)1) 358 (51.7) 507 (39.2)

SD, standard deviation.

 2009 International Society on Thrombosis and Haemostasis

562 C. Veyrat-Follet et al

volume of distribution at steady state was 30.8 L, and
absorption rate constant was 1.37 h. The half-lives of the three
phases of the model were 6.51 h for the first half-life, 95.1 h for
the second half-life, and 66.3 days for the terminal half-life
(Fig. 3). Interindividual variability of idraparinux clearance
was significantly related to subject weight, CrCl, sex, and age
1800
Individual predictions (ng mL–1)
index (whether <75 years or ‡75 years), whereas interindivid-
ual variability on central volume was significantly related to
age, weight and sex. The remaining interindividual variabil-
ity—not explained by covariate effects—was 18.6% on clear-
ance, 20.2% on central volume of distribution, and 129% on
absorption rate constant. Residual variability was low, with a
coefficient of variation of 16.5% for the proportional part and
34.2 ng mL)1 for the additive part, which was below the lower
limit of quantification of 71.4 ng mL)1.

1400
Assessment of covariates effects

1000 The effects of covariates on idraparinux pharmacokinetic

parameters in the final model are shown in Table 2. The most
important covariate effect on idraparinux clearance was CrCl,
600 with a decrease of 20% in a typical subject with moderate renal
Unit line impairement (CrCl ‡30 to <50 mL min)1) and a decrease of
Smooth
43% in a subject with severe renal impairment (CrCl
200
<30 mL min)1). Idraparinux clearance decreases by 16% in
200 600 1000 1400 1800 female subjects and by 5.3% in those aged ‡75 years; ethnicity
Observed concentrations (ng mL–1) appeared to have no effect on clearance.
Fig. 2. Goodness-of-fit plot: individual predictions of concentration vs.
observed concentration over one dosing interval. Individual pharmacokinetic parameter estimates

1200 Overall, pharmacokinetic parameters values in subjects across

the various clinical studies were in the same range, and were
Concentration (ng mL–1)

1000 t 1 α = 6.51 h
2 comparable between the Phase III and Phase I studies
800 (Table 3). At week 1 and at steady state, similar pharmacoki-
t 1 β = 95.1 h
netic parameters were observed in patients across clinical
600 2 studies, both in patients without (idraparinux dose of 2.5 mg)
400 t
1
γ = 66.3 days
and with (first dose of 2.5 mg then adjusted dose of 1.5 mg)
2 severe renal insufficiency. Mean parameter values ± SD at
200
steady state for patients without and with severe renal
0 insufficiency, respectively, were: Cmax ss 1.19 ± 0.278 lg mL)1
1.00 4.00 7.00 10.00 13.00 16.00 and 1.09 ± 0.261 lg mL)1; Ctrough ss 0.493 ± 0.166 lg mL)1
Time (week) and 0.571 ± 0.166 lg mL)1; and AUCss 110 ± 30.3
Fig. 3. Idraparinux steady state concentration time profile in a typical lg*h mL)1 and 115 ± 30.1 lg*h mL)1. The mean accumu-
patient. t1/2, half-life. lation ratio (AUCss/AUCsweek1) was 2.9 ± 0.5 and 2.5 ± 0.4

Table 2 Assessment of covariate effects on idraparinux pharmacokinetic parameters according to the final model
5 and Clearance % Distribution %
Covariate 95% (CL), L h)1 Change* volume (V2) Change*

Mean patient  0.0255 3.36

Creatinine clearance 35.1 0.0195 )23.5 N/A
(CrCl), mL min)1 150 0.0322 26.4
Weight (WT), kg 57 0.0231 )9.41 2.75 )18.1
116 0.0279 9.67 4.05 20.5
Sex (SEX) Female 0.0214 )16 2.73 )18.7
Age index (IAGE) ‡75 years 0.0241 )5.3 N/A
Age (AGE), years 35 N/A 3.68 9.67
83 3.27 )2.48

N/A, not applicable.

*Theoretical effect (% change with respect to the mean) of the covariate considered alone, the other covariates being set to their median value.
 
Male, aged <75 years, weighing 82.3 kg, and with a CrCl of 76.2 mL min)1.
^ ¼ ðh1 ðWT =82:3Þ h11  ðCrCl=76:2Þ h12  hSEX  hLAGE Þ
CL 13 14
^ ¼ ðh2 ðAGE=69Þ h15  ðWT =82:3Þ h16  hSEX Þ
V2 17

 2009 International Society on Thrombosis and Haemostasis

 Population pharmacokinetics of idraparinux 563

65.9 ± 13.0 (19.7)

Predicted concentration (ng mL–1)

0.0242 ± 0.00638
1200

97.2 ± 9.12 (9.4)

[0.0142–0.0347]
1000

[51.5–89.5]
[86.9–112]
800

(26.4)
Total

2027
600
Estimated pharmacokinetic
concentration equivalent to Ctrough
400 week 1, at week 36

64.1 ± 10.1 (15.7)

200
0.0245 ± 0.00616

96.2 ± 6.04 (6.3)

[0.0147–0.0347]
0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45

[84.1–52.0]
[87.5–106]
Amadeus

Time (week)
(25.2)
Table 3 Descriptive statistics of individual pharmacokinetic parameter estimates across studies – mean ± standard deviation (coefficient of variation %) [5–95%]

1295

Estimated Ctrough pharmacokinetic

concentration at week 1 (7 days after the
first administration): 109 ng mL–1

Fig. 4. Concentration-time course in a typical patient, and decrease of the

66.9 ± 10.4 (15.6)
0.0235 ± 0.00530

concentration to levels equal to the Ctrough at week 1.

97.5 ± 8.61 (8.8)
[0.0136–0.0331]

[54.3–86.4]
van Gogh

[86.6–115]
Extension

Predicted concentration (ng mL–1)

(22.6)

1200
24

1000

800
CL, clearance; DVT, deep-vein thrombosis; PE, pulmonary embolism; t1/2b, second half-life; t1/2c, terminal half-life.
99.3 ± 13.0 (13.1)

69.7 ± 16.8 (24.2)

van Gogh PE and

Estimated pharmacokinetic
0.0234 ± 0.00684

600 concentration equivalent to Ctrough

van Gogh DVT

[0.0132–0.0352]

week 1, at week 39
400
[102–50.7]
[85.4–124]

200
(29.2)
668

0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45
Time (week)
Estimated Ctrough pharmacokinetic
concentration at week 1 (7 days after the
98.5 ± 12.4 (12.6)

67.9 ± 16.5 (24.2)

–1
0.0241 ± 0.00674

first administration): 143 ng mL

van Gogh DVT

[0.0135–0.0352]

Fig. 5. Concentration-time course in a patient with severe renal impair-

[50.2–99.3]
[84.7–124]

ment, and decrease of the concentration to levels equal to the Ctrough at

week 1.
(28)
335

in patients without and with severe renal insufficiency,

respectively.
71.4 ± 17.1 (23.9)
100 ± 13.5 (13.4)
0.0227 ± 0.00688

Time to steady state was 35 weeks in a typical patient

[0.0129–0.0359]
van Gogh PE

(defined as one without renal insufficiency and aged

<75 years), which corresponds to approximately 8 months.
[85.4–125]

[51.5–105]
(30.3)

The estimated decrease of idraparinux concentration to levels

333

equal to the Ctrough at the end of week 1 occurs at week 36,

which is 12 weeks after cessation of treatment (Fig. 4). In
patients with severe renal insufficiency this corresponds to
15 weeks (Fig. 5). The dose adjustment performed in
60.4 ± 8.06 (13.3)
0.0283 ± 0.00304

96.6 ± 8.25 (8.5)

patients with severe renal insufficiency – at 1.5 mg after

[0.0234–0.033]

the first dose at 2.5 mg – resulted in an exposure at steady

[47.6–75.1]
[80.9–109]

state similar to that of patients without severe renal

Phase I

(10.7)

insufficiency (Fig. 6).

40

Discussion
CL, L h)1

The model developed in this study permitted the population

t1/2b, h

t1/2c,
days

pharmacokinetic analysis of 1987 of the 5473 patients with

DVT, PE or AF at-risk of thromboembolic events who were
N

 2009 International Society on Thrombosis and Haemostasis

564 C. Veyrat-Follet et al
Predicted concentration (ng mL–1) 1200
Typical patient
1000 Patient with severe renal insufficiency
800
600
400
200
0
0 14 28 42 56 70 84 98 112 126 140 154 168
Time (h)
Fig. 6. Idraparinux steady state concentration-time profile in a typical
patient and in a patient with severe renal impairment receiving adjusted-
dose idraparinux.
enrolled in the van Gogh and Amadeus Phase III clinical trials
of idraparinux. The pharmacokinetics of idraparinux were
found to be best described by a three-compartment model, with
interindividual variability on clearance, central volume of
distribution, and absorption rate constant with mixed residual
error. In the final model, the peak concentration was rapidly
achieved at 2.5 h, the typical clearance was 0.0255 L h)1, the
central volume of distribution was 3.36 L, and the absorption
rate constant was 1.37 h. Terminal half-life was 66.3 days and
time to steady state, estimated in patients without severe renal
insufficiency, was 35 weeks. The volume of distribution at
steady state was equal to 30.8 L.
To explain the long half-life and volume of distribution of
idraparinux, qualitative examination of whole body autora-
diography was performed in rats after administration of
radiolabelled 14C-idraparinux. It showed a long residence
time of radioactivity in the whole body, particularly in the
main fibrous structures. The non-lymphoid compartment of
secondary lymphoid organs (spleen and lymph nodes) also
displayed a prolonged radiolabeling. These observations
suggest that the connective tissues (including fibrous struc-
tures) may represent a fairly large storage compartment of
the compound, as illustrated in the proposed model (Fig. 7).
In addition, because draining lymph nodes are clearly
involved, lymphatic draining may participate in the elimina-
tion of idraparinux from this large interstitial compartment
(Fig. 7), as demonstrated by the slow and continuous release
from tissues over large observation periods of at least
1.5 month (data not shown). These findings are consistent
Idraparinux
ATIII degradation/renewal
ATIII
Via
lymphatics
Lymph nodes
Idraparinux ATIII
Fig. 7. Model of the distribution of idraparinux in the whole body.
with the moderate distribution volume determined in animals
and in humans as well as with the hydrophilic properties of
idraparinux, the volume of distribution calculated in this
study (Vss 30.8 L) being between the volume of extracellular
water and of total body water.
The only covariate effect that was potentially clinically
relevant was the 43% decrease in idraparinux clearance
observed in severe renally impaired patients. These findings
are in keeping with the predominantly renal excretion of
idraparinux. The adjusted dosing regimen (1.5 mg idraparinux
weekly) used for patients with severe renal impairment in the
van Gogh and Amadeus trials led to idraparinux exposures
similar to those observed in subjects with normal renal
functions at steady state. These findings could also have
clinical implications in elderly subjects as renal function
declines with age even in healthy individuals, and so clearance
of renally excreted drugs may become impaired in the elderly
patient. Of note, age was only found to have a minor effect on
idraparinux clearance, with a decrease of 5.3% in elderly
patients (those aged ‡75 years). Weight and sex were also
identified as significant covariates on clearance and volume of
distribution in the current study but ethnicity did not appear to
have an effect on idraparinux clearance. However, although
covariates may have an impact on pharmacokinetic parame-
ters, they may not predict clinical outcomes.
From Phase I clinical trials, Ma and Fareed reported an
elimination half-life of 120 h for idraparinux [11]. This value is
lower than the estimated value in this analysis. As non-
compartmental analyses and sampling design implemented in
Phase I studies do not allow accurate estimates of this
parameter, the half-life captured by Ma and Fareed could
correspond to the half-life of the second phase—approximately
100 h in the present population pharmacokinetic analy-
sis—instead of the third and terminal half-life of 66.3 days
that could be accurately determined after repeated doses, at
steady state.
The rapid onset of action, long half-life and predictable
pharmacokinetic properties of idraparinux confirmed by this
study offer the advantage of a once-weekly dosing schedule, as
well as a potentially extended protective effect against VTE for
an additional 6 months after cessation of therapy, as observed
in the van Gogh extension trial [6,12,13]. Studies have
demonstrated that single s.c. administration of equimolar
doses of idraparinux and idrabiotaparinux—the biotinylated
form of idraparinux—result in equivalent pharmacodynamic
and pharmacokinetic parameters [14]. The pharmacokinetic
profile of idrabiotaparinux and its reversal agent avidin will be
determined based on the ongoing Phase III trials EQUINOX,
Blood flow
Recycling CASSIOPEA and BOREALIS-AF.
The pharmacokinetics of idraparinux were best described by
a three-compartment model and include a rapid absorption
Parenchymal Clearance
interstitial fluid
with peak concentration reached at 2.5 h, and a typical
clearance of 0.0255 L h)1 corresponding to a terminal half-life
ATIII/idraparinux complex
of 66.3 days. These particular characteristics of idraparinux, a
rapid onset of action together with a long-acting anticoagulant
effect, offer interesting clinical perspectives that are under
 2009 International Society on Thrombosis and Haemostasis

investigation with idrabiotaparinux, the reversible biotinylated
form of idraparinux.
Disclosure of Conflict of Interests
The authors are employees of sanofi-aventis. The authors
received editorial support in the preparation of this manuscript,
funded by sanofi-aventis. The authors were fully responsible
for content of and editorial decisions for this manuscript.
References
1 Kearon C, Kahn SR, Agnelli G, Goldhaber SZ, Raskob GE,
Comerota AJ. Antithrombotic therapy for venous thromboembolic
disease: American college of chest physicians evidence-based clinical
practice guidelines (8th edition). Chest 2008; 6(Suppl): 454S–545S.
2 Herbert JM, Hérault JP, Bernat A, van Amsterdam RG, Lormeau JC,
Petitou M, van Boeckel C, Hoffmann P, Meuleman DG. Biochemical
and pharmacological properties of SANORG 34006, a potent and
long-acting synthetic pentasaccharide. Blood 1998; 91: 4197–205.
3 Organon, Sanofi-Synthélabo. Idraparinux sodium. Drug Data Rep
2003; 25: 332.
4 PERSIST investigators. A novel long-acting synthetic factor Xa
inhibitor (SanOrg34006) to replace warfarin for secondary prevention
in deep vein thrombosis. A Phase II evaluation. J Thromb Haemost
2004; 2: 47–53.
5 van Gogh Investigators, Büller HR, Cohen AT, Davidson B,
Decousus H, Gallus AS, Gent M, Pillion G, Piovella F, Prins MH,
Raskob GE. Idraparinux versus standard therapy for venous
thromboembolic disease. N Engl J Med 2007; 357: 1094–104.
 2009 International Society on Thrombosis and Haemostasis
Population pharmacokinetics of idraparinux 565
6 van Gogh Investigators, Büller HR, Cohen AT, Davidson B,
Decousus H, Gallus AS, Gent M, Pillion G, Piovella F, Prins MH,
Raskob GE. Extended prophylaxis of venous thromboembolism with
idraparinux. N Engl J Med 2007; 357: 1105–12.
7 Amadeus Investigators, Bousser MG, Bouthier J, Büller HR,
Cohen AT, Crijns H, Davidson BL, Halperin J, Hankey G, Levy S,
Pengo V, Prandoni P, Prins MH, Tomkowski W, Thorp-Pedersen C,
Wyse DG. Comparison of idraparinux with vitamin K antagonists for
prevention of thromboembolism in patients with atrial fibrillation: a
randomised, open-label, non-inferiority trial. Lancet 2008; 371: 315–
21.
8 Teien AN, Lie M. Evaluation of an amidolytic heparin assay method:
increased sensitivity by adding purified antithrombin III. Thromb Res
1977; 10: 399–410.
9 Cockcroft DW, Gault MH. Prediction of creatinine clearance from
serum creatinine. Nephron 1976; 16: 31–41.
10 Wählby U, Jonsson EN, Karlsson MO. Assessment of actual signifi-
cance levels for covariate effects in NONMEM. J Pharmacokinet
Pharmacodyn 2001; 28: 231–52.
11 Ma Q, Fareed J. Idraparinux sodium. sanofi-aventis. IDrugs 2004; 7:
1028–34.
12 Harenberg J, Vukojevic Y, Mikus G, Joerg I, Weiss C. Long-elimi-
nation half-life of idraparinux may explain bleedings and recurrent
events of patients from the van Gogh trials. J Thromb Haemost 2008; 6:
890–2.
13 Harenberg J, Jörg I, Vukojevic Y, Mikus G, Weiss C. Anticoagulant
effects of idraparinux after termination of therapy for prevention of
recurrent venous thromboembolism: observations from the van Gogh
trials. Eur J Clin Pharmacol 2008; 64: 555–63.
14 Trellu M, Perez Y, Ortiz J, Cheng S, Paty I. Bioequipotency of
idraparinux and biotinylated idraparinux after single dose in healthy
subjects. J Thromb Haemost 2007; 5(Suppl. 2): P-T-678.