Mol Diag Ther 2007; 11 (2): 123-128

ORIGINAL RESEARCH ARTICLE 1177-1062/07/0002-0123/$44.95/0

© 2007 Adis Data Information BV. All rights reserved.

The Influence of NSAIDs on Coumarin

Sensitivity in Patients with CYP2C9
Polymorphism After Total Hip
Replacement Surgery
Maarten J. Beinema,1 Petra H. de Jong,2 Har J.M. Salden,3 Merel van Wijnen,4 Jan van der Meer5 and
Jacobus R.B.J. Brouwers2
1 Deventer Hospital Thrombosis Center, Deventer, The Netherlands
2 Department Pharmacotherapy and Pharmaceutical Care, University of Groningen, Groningen Research Institute for
Pharmacy, Groningen, The Netherlands
3 Clinical Chemical Laboratory, Deventer Hospital, Deventer, The Netherlands
4 Clinical Chemical Laboratory, Meander Hospital, Amersfoort, The Netherlands
5 Division of Hemostasis, Academical Hospital Groningen, Thrombosis and Rheology, Groningen, The Netherlands

Abstract Objective: To determine the influence of NSAIDs on the international normalized ratio (INR) in patients with
cytochrome P450 (CYP)-2C9 enzyme variants starting acenocoumarol (an oral coumarin) therapy during the
first 7 days after total hip replacement surgery.
Methods: In this prospective study, an age-dependent protocol was used for the initiation of the acenocoumarol
dose. Low-molecular-weight heparin was given for 5 days. The study included 100 patients undergoing total hip
replacement surgery. After the inclusion of the last patient, polymerase chain reaction CYP2C9 mutation testing
was performed for all patients. Drug-use evaluation of NSAIDs and other potential coumarin-drug interactions
was also performed.
Results: Eleven patients had an INR on 1 or more days >4.9. There were 52 patients who were using NSAIDs.
Patients with a CYP2C9 mutation had a mean INR curve similar to patients without the mutation when NSAIDs
were not coadministered. Within the group of patients heterozygous for a CYP2C9 mutation (n = 30) only
concomitant use of a NSAID resulted in an INR >4.9 (0% vs 38.9%, p < 0.05).
Conclusion: In the group of patients with a CYP2C9 variant (*2 or *3 alleles), only concomitant use of a NSAID
resulted in INRs >4.9. The cost effectiveness of CYP2C9 screening before elective surgery has yet to be
determined.

Both low-molecular-weight heparins (LMWH)[1-4] and couma-
rins have been shown to reduce venous thrombo-embolism (VTE)
after total hip replacement (THR) surgery.[5-7] Coumarins inhibit
the formation of vitamin K-dependent proteins in the liver and
achieve their optimal anticoagulant effect after the first 3 or 4 days
of use. Therefore, the initial dosage should be combined with
LMWH during at least the first 5 days for the prophylaxis or
treatment of deep vein thrombosis (DVT). The international nor-
malized ratio (INR) must be frequently monitored because of the
small therapeutic range and the inter-individual variability of the
required daily dosage.[8] To refine the optimal starting dose for
coumarins, an age-adjusted dosage has been shown to be effec-
tive.[9-11]
Coumarins are metabolized in the liver by the cytochrome P450
(CYP) system, especially by the enzyme CYP2C9.[12] Polymorph-
isms of this enzyme exist within the population. The CYP2C9
allele most often found is *1; the genotype *1*1 being the wild-
type form. The variant alleles *2 (found in 20–30% of the Cauca-
124 Beinema et al.

Table I. Age-adjusted initial dose for acenocoumarol in study participants Methods

following total hip replacement surgery. The same does was given on days
1 and 2. The first international normalized ratio was measured on day 3
Study Population
Age (y) Dose (mg)
25–35 3 The population consisted of consecutive eligible patients
35–45 4 scheduled for THR surgery. Exclusion criteria were femoral neck
45–55 4 fracture and current active bleeding disorders, which prohibited
55–65 3 the use of coumarins. Written informed consent was obtained from
65–75 3 the patients, according to the protocol.
75–85 2
Setting
>85 2
At least 6 hours after THR surgery, postoperative nadroparin
sian population) and *3 (5–10% of the Caucasian population) are calcium (Fraxiparine®; Sanofi-Synthélabo, Maassluis, The
associated with a different pharmacokinetic profile for couma- Netherlands)1 was administered in a dosage of 0.3mL (5.700 IE
anti-factor Xa) and was continued for a minimum of 5 days until
rins.[13-15] The *3 variant, particularly the homozygous genotype,
the INR was >2.0. Acenocoumarol was commenced on the same
lowers the required daily dose to 20% of that in people with the
day as nadroparin according to an age-dependent initial-dose
wild-type form.[16] Patients with a CYP2C9 gene variant have a schedule (table I). The first INR was measured on day 3 and the
higher risk of bleeding complications.[17] frequency of the next INRs depended upon the stability of the
The peri-operative management of coumarin therapy is often anticoagulant therapy, defined by whether the following INRs
were within the target range (1.8–3.5). We used the linear interpo-
complicated by the concomitant use of interacting drugs.[18]
lation method as described by Rosendaal et al.[24] in order to
NSAIDs are effective analgesic agents and are often used for the
estimate the daily INR on the days between the two INR measure-
prevention of heterotopic ossification after THR surgery.[19] ments. INR testing was performed on ACL™ Futura (IL, using
Cyclo-oxygenase (COX)-1/COX-2-type NSAIDs have anti-plate- HemosIL™ RecombiPlasTin; Instrumentation Laboratory, Milan,
let properties, so their potential interaction with coumarin is docu- Italy).
mented as a pharmacodynamic interaction,[20] which may lead to The surgeon performing the THR surgery prescribed the initial
acenocoumarol dose, conforming to the starting dose protocol in
bleeding complications. COX-2-specific NSAIDs have no anti-
table I. The physicians at the thrombosis center of our hospital
platelet properties. This has been established in a retrospective prescribed acenocoumarol after the initial dose. A blood sample
study by Visser et al.[21] Recently, significant increases in INRs (EDTA) was collected for the detection of CYP2C9 mutations.
were described in patients receiving stable coumarin therapy after The PCR testing was performed on a LightCycler® (Roche, Al-
adding a ‘classical’ NSAID or COX-2 inhibitor.[22,23] Coumarins mere, The Netherlands) with probes for the *2 and *3 variants of
may potentially have a pharmacokinetic interaction with NSAIDs,
Table II. Patient demographics
because both drugs are metabolized by CYP2C9 and patients with
Characteristic
a CYP2C9 variant have a lower metabolizing capacity. Age (y) [mean (range)] 70.9 (33–95)
The primary hypothesis for this study was that during the first Men (n) 27
week after THR surgery patients with a CYP2C9 gene variant Women (n) 73

would have higher INRs, especially when the prescribed NSAIDs Days in hospital [mean (range)] 11.2 (5–35)
Body mass index [mean (range)] 26.9 (18.2–44.1)
are metabolized by the same enzyme.
Days receiving low-molecular-weight 6.4 (5–13)
The goal of our study was to determine the influence of heparin [mean (range)]
NSAIDs on the INR in patients with CYP2C9 gene variants Drugs prescribed [median (range)] 13 (6–26)
starting with acenocoumarol therapy during the first 7 days after Drop in hemoglobin level [mean (range)] –2.6 (1.1–3.9) mmol/L
(= –4.2 g/dL)
THR surgery.

1 The use of trade names is for product identification purposes only and does not imply endorsement.

© 2007 Adis Data Information BV. All rights reserved. Mol Diag Ther 2007; 11 (2)
CYP2C9 Polymorphism and Effect of NSAIDs on Coumarin Sensitivity 125

Table III. CYP2C9 genotypesa among patients (n = 100) in this study
Genotype No. of patients
*1*1 65
*1*2 22
*1*3 8 distribution of CYP2C9 alleles in our patients is presented in table
*2*2 4 III.
*2*3 1 During the first 7 days after the initial dose of acenocoumarol,
a Alleles tested were the wild-type (CYP2C9*1) and the variants *2
and *3.
CYP2C9.[25] The analysis was performed 2 weeks after the inclu-
sion of the last patient and, therefore, coumarin dosage was not
influenced by knowledge of CYP2C9 polymorphism.
We collected patient demographics at baseline (i.e. age, sex,
medical history, weight, height, co-morbidities) as well as data
occurring during the hospital stay (i.e. blood loss, fever, renal
function, liver function, concomitant diseases). We used the infor-
mation system of the hospital pharmacy in order to collect data
about concomitant drugs, especially NSAIDs.
Outcomes
The target range was defined as 1.8–3.5 INR. The primary
outcome of the study was the number of patients with an INR >4.9.
Complications
Bleeding complications were indexed as major bleeding (lead-
ing to death or re-operation, bleedings in vital organs [intracranial,
retroperitoneal], bleeding index >2) or minor bleeding. VTE
events, if clinically suspected, were confirmed with compression
Results
One hundred patients were included over a period of 6 months.
Baseline patient’s characteristics are presented in table II. The
the average INR was within target range (1.8–3.5) for 66% of the
INR values measured, 26% were below an INR of 1.8, and 8%
were above an INR of 3.5. The average duration of LMWH
therapy was 6.4 days (5–13 days). The duration of LMWH use was
prolonged if the INR on day 5 was below 2. Eleven patients had an
INR >4.9 on 1 or more days. One patient had INRs >5.9.
The total number of prescription drugs in the peri- and post-
operative period ranged from 6 to 26 per patient, with a mean of
13. We specifically monitored comedication with CYP2C9-inhib-
iting drugs and other drugs that are metabolized by CYP2C9.
During the observed period, 52 patients had a prescription for one
or more NSAIDs in a standard daily dosage (once or twice daily)
for 2 or more days. Usually the combination with a gastro-
protective drug (e.g. a proton pump inhibitor) was given.
We analyzed the patients who had INRs >4.9. Of 65 patients
with the wild-type enzyme (genotype *1*1), four patients (6%)
had an INR >4.9, compared with 7 of 35 patients (20%) with the
variant-type enzyme (p < 0.05) [table IV]. In the subgroup of 34
wild-type patients who did not use NSAIDs, three patients had an
INR >4.9, compared with one patient in the group of 31 NSAID
users (p > 0.05). In the variant enzyme group, all seven patients
Table IV. Proportion of patients with an international normalized ratio (INR)
>4.9 on 1 or more days during the first 7 days after total hip replacement
surgery, grouped by genotype and NSAID use

ultrasonography (CUS) for DVT and spiral CT for pulmonary CYP2C9 INR >4.9 [n/total (%)]
embolism (PE). In both cases the d-dimer level (Tina-quant® D- genotype whole group no NSAID subgroup NSAID subgroup
dimer; Roche) was used as a negative predictive test in case the *1*1 4/65 (6) 3/31 (10) 1/34 (3)a
CUS or spiral CT was negative. *1*2 5/22 (23) 0/6 (0) 5/16 (31)
*1*3 2/8 (25) 0/6 (0) 2/2 (100)
*2*2 0/4 (0) 0/3 (0) 0/1 (0)
Statistical Analysis
*2*3 0/1 (0) 0/0 (0) 0/1 (0)
*3*3 0/0 (0) 0/0 (0) 0/0 (0)
Standard descriptive methods and statistics were used to char-
*1*2 + *1*3 7/30 (23)b 0/12 (0) 7/18 (39)c
acterize the study population. Continuous variables were de-
All variants 7/35 (20) 0/15 (0) 7/22 (32)d
scribed with mean and standard deviation, and categorical vari-
All genotypes 11/100 (11) 3/48 (6) 8/52 (15)a
ables as frequency and percentage. Continuous variables were
a No significant difference between ‘NSAID’ and ‘no NSAID’ groups.
compared using the Fisher exact test. For all tests, a two-tailed p-
b p < 0.05 vs the *1*1 group.
value ≤ 0.05 was considered statistically significant.
c p = 0.0156 vs the ‘no NSAID’ group.
The study was approved by the hospital medical ethical com-
d p = 0.0407 vs the ‘no NSAID’ group.
mittee.

© 2007 Adis Data Information BV. All rights reserved. Mol Diag Ther 2007; 11 (2)
126 Beinema et al.

who had an INR >4.9 were in the group using NSAIDs (n = 20),
while none were seen in the non-NSAID group (n = 15) [p < 0.05].
There was no significant difference in the average daily INRs
between the patients with a wild-type enzyme and patients with a
variant enzyme (figure 1a). However, there was a difference in
patients with a variant CYP2C9 allele who used concomitant
NSAIDs (figure 1b and figure 1c). It could not be attributed to the
use of one specific NSAID (naproxen, indometacin, ibuprofen,
a
5.0
4.5
4.0
3.5
INR
diclofenac, acetylsalicyl acid [aspirin], meloxicam, propyfenazon,
rofecoxib).
The mean decrease in hemoglobin level in all patients was 2.6
mmol/L after surgery; 47 patients were transfused blood. One
patient had rectal bleeding at day 7, with a bleeding index >2 and
who subsequently received a transfusion of two units of red blood
cells. The INR of this patient was within target range at the time of
the bleeding, and was being treated with an NSAID at the time.
None of the patients developed clinical signs of DVT or PE.
*1*1
Variant Discussion
In this study we found that, as a pooled group, patients with a
variant CYP2C9 enzyme were at greater risk than patients with the

3.0
2.5 wild-type form for INRs >4.9 following an initial dose of ace-
2.0 nocoumarol. Analysis of subgroups (NSAID use vs no NSAID
1.5 use) within the wild-type and variant groups showed no difference
1.0
1 2 3 4 5 6 7
between the wild-type subgroups, as we suspected, because pa-
tients with the wild-type enzyme have the full capacity for elimi-
b nating drugs. The subgroup of patients with a variant CYP2C9
6.0 *1*1
*1*2
enzyme without NSAID use had the same mean INR curve as
5.5
*1*3 patients with the wild type. In the subgroup analysis, patients with
5.0
4.5
a CYP2C9 variant only showed an increase in the measured INRs
4.0 when NSAIDs were coadministered with acenocoumarol.
3.5 CYP2C9 polymorphic variants have a reduced capacity to hydrox-
INR

3.0 ylize substrate drugs; thus the coadministeration of two substrate

2.5
drugs may have pharmacokinetic effects on the elimination on one
2.0
or both drugs. NSAIDs are frequently prescribed for inflammatory
1.5
1.0
processes, yet the effect of the combination of common CYP2C9
1 2 3 4 5 6 7 variants and NSAID use in the outpatient population is unclear.
Inflammation alone can lead to an increase in the INR; this may be
c
6.0 *1*1 mediated by interleukin-6 via inhibition of the CYP receptor. In
5.5 *1*2 this study the population was homogeneous, without clinical signs
*1*3
5.0 of inflammation.
4.5 Dose determination of coumarins after THR surgery is often
4.0
considered as ‘walking the tightrope’; without expert care and
3.5
follow-up, patients are at high risk of falling off. In other words,
INR

3.0
2.5 patients are at high risk for supratherapeutic INRs leading to
2.0 bleeding, or subtherapeutic INRs leading to an increased risk for
1.5 VTE events if they are not closely monitored and educated. In this
1.0 study we used a protocol for the initial dose of acenocoumarol,
1 2 3 4 5 6 7
Day taking into account the patient’s age. The required daily dosage
Fig. 1. (a) Mean international normalized ratio (INR) [and standard devia- decreases with age and although the cause of this decrease is
tion] of patients with and without a cytochrome P450-2C9 (CYP2C9) vari- uncertain, it is probably due to decreased enzyme activity. There is
ant (*2 or *3 alleles). There is no significant difference in INRs between the
no need for a high initial acenocoumarol dose using our protocol
two groups when NSAID use is not taken into account. (b) Mean INR and
standard deviation for CYP2C9 variants in patients who used NSAIDs. (c) because the risk of thrombosis is reduced by LMWH during the
Mean INR and standard deviation for CYP2C9 variants in patients who did first days post-surgery. The majority of patients reached a thera-
not use NSAIDs. peutic INR within 5 days after surgery. Patients requiring more

© 2007 Adis Data Information BV. All rights reserved. Mol Diag Ther 2007; 11 (2)
CYP2C9 Polymorphism and Effect of NSAIDs on Coumarin Sensitivity
oral anticoagulant drugs are protected by an extended prescription
of LMWH. No clinically manifest VT or PE was reported. Only
one major bleeding (bleeding index >2) occurred during our study,
resulting in a 2.6 mmol/L decrease in the patient’s blood hemoglo-
bin level. This was precisely the same mean decrease in hemoglo-
bin level for the total study population.
We performed CYP2C9 PCR testing only after the last patient
was discharged from the hospital, eliminating bias by both dose
determination and dose adjustment, as well as the use of concomi-
tant drugs. The frequencies of the variant polymorphisms studied
are similar to those found in other studies within the Caucasian
population.
The low incidence of bleeding complications in our subpopula-
tion can be explained by the relatively low number of patients with
an INR >4.9 compared with other clinical trials. Only one patient
had an INR >5.9. We believe that INRs >4.9 as a primary endpoint
is justified because it has been established that there is a direct
relationship between INR >4.9 and the incidence of bleeding
complications in patients administered coumarin.[26,27]
The influence of VKORC1 (vitamin K expoxide reductase
complex, subunit 1) polymorphisms is also of great importance in
coumarin dose finding,[28] and the combination with CYP2C9
polymorphisms explains up to 50% of the interindividual variabili-
ty in coumarin sensitivity.[29] We did not perform VKORC1 geno-
type testing, because this was not available at the time (2002–3).
Only the combination of both VKORC1 and CYP2C9*3
polymorphisms is associated with an increased risk of severe over-
anticoagulation.[30] The combination of a CYP2C9 *1*1 genotype
with a VKORC1 polymorphism does not confer an increased risk.
Therefore, the VKORC1 genotype may not greatly influence the
outcomes of our study, because in our study only the combination
of CYP2C9 *2 or *3 polymorphisms with NSAID use led to over-
anticoagulation.
Recently, newer prophylactic anticoagulants such as
fondaparinux have been introduced into clinical practice. Because
of the necessity of extended prophylaxis in orthopedic surgery,
oral anticoagulants (e.g. coumarins, direct thrombin inhibitors)
may be preferred instead of parenteral anticoagulants (e.g.
LMWH, fondaparinux).
Optimizing coumarin prophylaxis by computer dosage has
proven effective.[31] Our protocol, which has been previously
validated using the combination of LMWH and coumarins with an
age-adjusted dosage,[11] appears safe and effective in achieving
INRs within the target range for the majority of patients. To further
optimize coumarin prophylaxis in surgical patients, a pre-opera-
tive CYP2C9 genotype test may be a valuable tool. The cost-
effectiveness of this test strategy before elective surgery has yet to
be determined.
© 2007 Adis Data Information BV. All rights reserved.
127
These conclusions may also be valid for warfarin.[32] Warfarin
is also mainly metabolized by the CYP2C9 iso-enzymes and the
initial loading dose for warfarin is also age dependent.
Conclusion
During the first week after THR surgery, the combination of
acenocoumarol and NSAIDs leads to over-anticoagulation in pa-
tients with a variant CYP2C9 allele, while a variant alone is not
associated with a significant increase in INR relative to patients
with the wild-type enzyme form. CYP2C9 genetic screening may
help to prevent severe over-anticoagulation, although the cost-
effectiveness of this test strategy will need to be determined.
Acknowledgments
No sources of funding were used to assist in the preparation of this study.
The authors have no conflicts of interest that are directly relevant to the
content of this study.
References
1. O’Donnell M, Linkins LA, Kearon C, et al. Reduction of out-of-hospital sympto-
matic venous thromboembolism by extended thromboprophylaxis with low-
molecular weight heparin following elective hip arthroplasty. Arch Intern Med
2003; 163: 1362-6
2. Strebel N, Prins M, Agnelli G, et al. Preoperative or postoperative start of
prophylaxis for venous thromboembolism with low-molecular-weight heparin
in elective hip surgery? Arch Intern Med 2002; 162: 1451-6
3. Hull RD, Pineo GF, MacIsaac S. Low-molecular-weight heparin prophylaxis:
preoperative versus postoperative initiation in patients undergoing elective hip
surgery. Thromb Res 2001; 101: 155-62
4. Samama CM, Vray M, Barré J, et al. Extended venous thromboembolism prophy-
laxis after total hip replacement: a comparison of low-molecular-weight heparin
with oral anticoagulant. Arch Intern Med 2002; 162: 2191-6
5. Douketis JD, Eikelboom JW, Quinlan DJ, et al. Short-duration prophylaxis against
venous thromboembolism after total hip or knee replacement: a meta-analysis
of prospective studies investigating symptomatic outcomes. Arch Int Med
2002; 162: 1465-71
6. Palmer AJ, Koppenhagen K, Kirchhof B, et al. Efficacy and safety of low
molecular weight heparin, unfractioned heparin and warfarin for thrombo-
embolism prophylaxis in orthopaedic surgery: a meta-analysis of randomised
clinical trials. Haemostasis 1997; 27 (2): 75-84
7. Dunn AS, Turpie AG. Perioperative management of patients receiving oral antico-
agulants: a systematic review (a provisional record). Arch Intern Med 2003;
163: 901-8
8. Bates SM, Ginsberg JS. Bleeding after initiation of anticoagulant therapy for deep
vein thrombosis [editorial]. Am J Med 1999; 107: 517-8
9. Roberts GW, Helboe T, Nielsen CBM, et al. Assessment of an age-adjusted
warfarin initiation protocol. Ann Pharmacother 2003; 37: 799-803
10. O’Connell MB, Kowal PR, Allivato CJ, et al. Evaluation of warfarin initiation
regimes in elderly inpatients. Pharmacotherapy 2000; 20: 923-30
11. Van den Bemt PMLA, Beinema MJ, Van Roon EN, et al. Initiation of oral
anticoagulant therapy in orthopedic patients: an algorithm compared with
routine dosing. Eur J Clin Pharmacol 2002; 58: 203-8
12. Higashi MK, Veenstra DL, Kondo LM, et al. Association between CYP2C9
genetic variants and anticoagulation-related outcomes during warfarin therapy.
JAMA 2002; 287: 1690-8
13. Thijssen HH, Flinois JP, Beaune PH. Cytochrome P4502C9 is the principal catalyst
of racemic acenocoumarol hydroxylation reactions in human liver microsomes.
Drug Metab Dispos 2000; 28: 1284-90
Mol Diag Ther 2007; 11 (2)
128
14. Takahashi H, Echizen H. Pharmacogenetics of warfarin elimination and its clinical
implications. Clin Pharmacokinet 2001; 40: 587-603
15. Margaglione M, Colaizzo D, Dándrea G, et al. Genetic modulation of oral an-
ticoagulation with warfarin. Thromb Haemost 2000 Nov; 84 (5): 775-8
16. Verstuyft C, Morin S, Robert A, et al. Early acenocoumarol overanticoagulation
among cytochrome P450 2C9 poor metabolizers. Pharmacogenetics 2001; 11:
735-7
17. Aithal GP, Day CP, Kesteven JL, et al. Association of polymorphisms in the
cytochrome P450 CYP2C9 with warfarin dose requirements and risk of bleed-
ing complications. Lancet 1999; 353: 717-79
18. Brouwers JR, de Smet PA. Pharmacokinetic and pharmacodynamic drug interac-
tions with non-steroidal and anti-inflammatory drugs. Clin Pharmacokinet
1994; 27: 462-85
19. Fijn R, Koorevaar RT, Brouwers JRBJ. Prevention of heterotopic ossification after
total hip replacement with NSAIDs. Pharm World Sci 2003; 25: 138-45
20. Connelly CS, Panush RS. Should nonsteriodal ani-inflammatory drugs be stopped
before elective surgery? Arch Intern Med 1991; 151: 1963-6
21. Visser LE, van Schaik RH, van Vliet M, et al. Allelic variants of cytochrome P450
2C9 modify the interaction between non-steroidal anti-inflammatory drugs and
coumarin anticoagulants. Clin Pharmacol Ther 2005; 77: 479-85
22. Schaeffer MG, Plowman BK, Morreale AP, et al. Interaction of rofecoxib and
celecoxib with warfarin. Am J Health Syst Pharm 2003; 60: 1319-23
23. Van Dijk KN, Plat AW, van Dijk AAC, et al. Potential interaction between
acenocoumarol and diclofenac, naproxen and ibuprofen and the role of
CYP2C9 genotype. Thromb Haemost 2004; 91: 95-101
24. Rosendaal FR, Cannegieter SC, Van der Meer FJM, et al. A method to determine
the optimal intensity of oral anticoagulant therapy. Thromb Haemost 1993; 69:
236-9
© 2007 Adis Data Information BV. All rights reserved.
Beinema et al.
25. Burian M, Grösch S, Tedeger I, et al. Validation of a new fluorogenical real-time
PCR assay for detection of CYP2C9 allelic variants and CYP2C9 allelic
distribution in a German population. J Clin Pharmacol 2002; 54: 518-21
26. Van der Meer FJM, Roosendaal FR, Vandenbroucke JP, et al. Bleeding complica-
tions in oral anticoagulant therapy: an analysis of risk factors. Arch Intern Med
1993; 153: 1557-62
27. Azar AJ, Cannegieter SC, Deckers JW, et al. Optimal intensity of oral anticoagu-
lant therapy after myocardial infarction. J Am Coll Cardiol 1996 May; 27 (6):
1349-55
28. Rieder M, Reiner AP, Gage BF, et al. Effect of VKORC1 haplotypes on transcrip-
tional regulation and warfarin dose. N Engl J Med 2005; 352: 2285-93
29. Bodin L, Versuyft C, Tregouet DA, et al. Cytochrome P450 2C9 (CYP2C9) and
vitamin K epoxide reductase (VKORC1) genotypes as determinants of ace-
nocoumarol sensitivity. Blood 2005; 106: 135-40
30. Schalekamp T, Brassé BP, Roijers JFM, et al. VKORC1 and CYP2C9 genotypes
and acenocoumarol anticoagulation status: interaction between both genotypes
affects overanticoagulation. Clin Pharm Ther 2006; 80: 13-22
31. MacDonald D, Bhalla P, Cass W, et al. Computerized management of oral
anticoagulant therapy: experience in major joint arthroplasty. Can J Surg 2002;
45: 47-52
32. Gage BF, Eby C, Milligan PE, et al. Use of pharmacogenetics and clinical factors
to predict the maintanance dose of warfarin. Thromb Haemost 2004; 91: 87-9
Correspondence: Dr Maarten J. Beinema, Deventer Hospital Thrombosis
Center, PO Box 5001, 7400 GC Deventer, The Netherlands.
E-mail: beinemam@dz.nl
Mol Diag Ther 2007; 11 (2)