Pharmacogenetic aspects of anticoagulant therapy with warfarin in patients after heart valve replacement under conditions of polypharmacotherapy

The dissertation is devoted to the study of pharmacogenetic factors that determine individual sensitivity to warfarin in patients after heart valve replacement under conditions of intensive concomitant drug therapy. It examines the influence of polymorphisms of the CYP2C9 and VKORC1 genes, the activity of the CYP2C9 biotransformation enzyme assessed by means of the losartan test, as well as drug-drug interactions on the dosing regimen and the anticoagulant effect of the indirect anticoagulant in the early and late postoperative periods.

The study was carried out in a sample of cardiac surgery patients with implanted artificial heart valves; genotyping was performed for polymorphic markers of the CYP2C9 and VKORC1 (G3673A) genes, the concentration of losartan and its metabolite E-3174 in urine was determined, and concomitant pharmacotherapy was analyzed. Comparing the data obtained with the selected warfarin doses allows assessment of the contribution of genetic and pharmacokinetic factors to maintaining the therapeutic range of the international normalized ratio (INR) and to the risk of hemorrhagic and thromboembolic complications.

INTRODUCTION

At present, it is difficult to imagine heart valve replacement in patients with acquired and congenital heart defects without concomitant anticoagulant therapy. Indirect anticoagulants (AKND), coumarins, were introduced into clinical practice in the middle of the twentieth century almost simultaneously with the era of formation of surgical treatment of heart defects. Vitamin K antagonists remain to this day the only and irreplaceable oral anticoagulants available for the long-term treatment of patients with implanted artificial heart valves (IKS). Phase II and III clinical trials of new direct (antithrombin III-independent, selective) thrombin inhibitors that could claim to be an alternative to AKND in the prevention of thrombotic complications in patients after IKS implantation have not been successful [125].

At the same time, the existing therapeutic window for anticoagulant therapy is rather narrow. This fact is particularly significant for patients with IKS, since on the one hand it is limited by the risk of thromboembolic complications, and on the other by bleeding, which account for up to 75% of all postoperative complications [71]. According to the latest data, this can be explained not only by the intake of vitamin K with food and diseases of the gastrointestinal tract (ZhKT), but also by the combined use of medicinal products (LS) of other groups and genetic predisposition, namely polymorphism of the CYP2C9 gene, which encodes the main biotransformation enzyme of AKND [17, 22, 24, 30, 58, 60, 91, 110]. At the same time, carriers of the allelic variants CYP2C9*2 and CYP2C9*3 show low activity of the CYP2C9 enzyme, which contributes to slower biotransformation of warfarin. The features of the pharmacokinetics and pharmacodynamics of AKND in patients with the "slow" metabolism genotype (carriers of allelic variants CYP2C9*2, CYP2C9*3) have been studied in many investigations. According to data obtained during warfarin use in carriers of allelic variants CYP2C9*2 and CYP2C9*3, the risk of excessive hypocoagulation (INR above 4) increases by 3-4 times, and the risk of bleeding by 2-3 times [122]. There is evidence that genotyping for CYP2C9 combined with phenotyping (determination of the activity of the CYP2C9 gene using the losartan test) can more accurately predict the maintenance dose of warfarin [1, 24, 79, 90, 130]. However, the influence of another genetic factor is not excluded, namely the polymorphism of vitamin K epoxide reductase (VKORC1), which also determines sensitivity to warfarin [18, 44]. A number of studies have been published on the influence of certain polymorphic markers of the VKORC1 gene on the risk of excessive hypocoagulation and bleeding during warfarin use [9, 26, 38, 72].

An atypical pharmacological response in the form of adverse drug reactions or ineffectiveness of LS may be determined in 50% of cases by the genetic characteristics of patients [14]. In this regard, the development of new comprehensive pharmacogenetic approaches to improve the safety of AKND therapy in patients after IKS implantation, taking into account polymorphisms of the CYP2C9 and VKORC1 genes, as well as drug-drug interactions, is currently relevant. It also seems relevant to supplement CYP2C9 genotyping of patients with phenotyping, that is, determination of CYP2C9 activity using the so-called losartan test [27].

Aim of the study: to evaluate the features of the influence of genetic factors (polymorphisms of the CYP2C9 and VKORC1 genes), the activity of the CYP2C9 biotransformation enzyme, and drug-drug interactions on the anticoagulant effect of warfarin in patients after IKS implantation in the early and late postoperative periods.

Objectives of the study:

1. To determine the frequency of occurrence of genotypes for polymorphic markers of the CYP2C9 and VKORC1 (G3673A) genes and their combinations involved in warfarin metabolism in cardiac surgery patients.

2. To compare the selected doses of warfarin in the early and late postoperative periods with polymorphisms of the CYP2C9 and VKORC1 (G3673A) genes.

3. To evaluate the influence of CYP2C9 gene polymorphism on CYP2C9 activity according to the losartan test in patients prior to IKS implantation.

4. Based on the results of the losartan test, to evaluate the relationship between the concentration of the active metabolite E-3174 and the selected dose of warfarin in the early and late postoperative periods.

5. To study the influence of drug-drug interactions on the maintenance dose of warfarin in the early postoperative period depending on polymorphisms of the CYP2C9 and VKORC1 (G3673A) genes.

Scientific novelty of the dissertation.

For the first time in Russian medicine, the influence of polymorphisms of the CYP2C9 and VKORC1 genes on the features of dosing and the anticoagulant effect of the indirect anticoagulant warfarin in patients with IKS in the early and late postoperative periods has been evaluated. For the first time in the world, in patients with IKS, a relationship has been revealed, according to the results of the losartan test, between the concentration of the active metabolite of losartan in urine and the selected dose of warfarin in the late postoperative period. For the first time in the Russian population, the influence of drug-drug interactions on the features of the warfarin dosing regimen depending on CYP2C9 and VKORC1 (G3673A) genotypes in patients with IKS in the early postoperative period has been evaluated.

Practical significance of the work

The necessity has been shown of determining carrier status of CYP2C9 and VKORC1 (G3673A) genotypes in all patients before IKS implantation who are scheduled to receive warfarin. In patients scheduled to receive warfarin, it is advisable to supplement determination of CYP2C9 genotype carrier status with determination of CYP2C9 activity (by the losartan test).

In the early postoperative period in cardiac surgery patients with CYP2C9*1/*1 and VKORC1-G3673A (GG/GA) genotypes, it is necessary to take into account the interaction of warfarin with antimicrobial agents (antibiotics, antifungal agents), amiodarone, prednisolone and other LS that can also affect biotransformation processes. After withdrawal of concomitant therapy affecting warfarin biotransformation, it is advisable in this category of patients to carry out stricter INR monitoring with subsequent dose adjustment of AKND.

Propositions put forward for defense

The warfarin dosing regimen in patients with IKS depends on CYP2C9 and VKORC1 (G3673A) genotypes in the late, but not in the early postoperative period.

A low concentration of losartan in urine (less than 2500 ng/ml according to the losartan test) before IKS implantation reveals low activity of CYP2C9*1/*1, which is associated with the selection of lower doses of warfarin in the late postoperative period.

Massive polypharmacotherapy in the early postoperative period affects the warfarin dosing regimen in patients with CYP2C9*1/*1 and GG/GA genotypes for the polymorphic marker G3673A of the VKORC1 gene.

Chapter 1.

LITERATURE REVIEW

1.1. Pharmacogenetic foundations determining adverse drug reactions of anticoagulant therapy

At present, the only effective method for correcting hemodynamic disorders in heart valve defects is IKS implantation [5]. Rehabilitation and preservation of the working capacity of patients after IKS implantation largely depend on the prevention of valve thrombosis and thromboembolic complications. Thromboembolic complications in patients after aortic valve (AK) replacement without anticoagulant therapy amount to approximately 4-12% per year; in patients with mitral valve (MK) replacement the figures are even higher [41]. The use of AKND, including warfarin, makes it possible to reduce the coagulation potential of blood due to the synthesis of inactive forms of factors II, VII, IX, X-PIVKA (protein induced vitamin K absence) and to "extinguish" the response of the hemostasis system to the presence of a foreign body in the active bloodstream [20] (Scheme 1.1.1).

CYP2C9

Inactive proenzyme (factors II, VII, IX, X, protein C, protein S)

WARFARIN

Oxidized form of vitamin K

Active enzyme (factors II, VII, IX, X; protein C, protein S)

PIVKA - proteins induced in vitamin K absence (factors II, VII, IX, X; protein C, protein S)

Scheme 1.1.1. Mechanism of action of warfarin - blockade of vitamin K-dependent factors. Vavilova T.V., 2010

However, the initiation of warfarin therapy remains largely empirical and its use requires careful patient monitoring. Bleeding during warfarin use occurs in 7.6-26.5% of cases per year, and serious hemorrhages at a rate of 0.3-4.2 per 100 patients annually [11, 13, 15, 31, 51, 120, 121]. In this case, bleeding most often develops in the first month of warfarin use. The mechanism of bleeding consists in an increase in the concentration of the oral anticoagulant in blood plasma and in increased sensitivity of the vitamin K epoxide reductase complex (VKOR). A decrease in vitamin K intake into the body enhances the anticoagulant effect of the drug, which in turn inhibits VKOR. Proteins formed during oral anticoagulant administration have low activity in calcium-dependent reactions occurring on the phospholipid surface, which is the anticoagulant effect, manifested as an increase in INR. An excessive increase in INR leads to bleeding.

The developed correction schemes and criteria for adequacy of antithrombotic protection make it possible to find a balance between its effectiveness and safety. According to some data, the absence of a therapeutic effect from the use of certain LS is observed in 10-40% of cases [104, 115]. At the same time, in the United States alone, about 100,000 people die and more than 2 million are hospitalized annually due to adverse drug reactions (NLR) [65]. NLR can lead to disability or even death of patients; according to some data, they rank 4th-5th among causes of death [15].

An INR increase above 3.0 provides no additional advantage in the prevention of thromboembolic complications, while the risk of intracranial hemorrhage increases almost 2.5-fold, and an INR rise from 4.0 to 6.0 increases this risk almost 16-fold. According to the results of the ESPIRIT study, the risk of intracranial bleeding increases 1.37-fold for every 0.5 increase in INR above the therapeutic level. However, not all bleeding events can be associated with an elevated INR level. Thus, 30-40% of bleeding events, including major ones, occur within the therapeutic INR range. Various factors contribute to an INR increase above 4.0 during warfarin use; their consideration may influence the calculation of bleeding risk for a particular patient [12].

Factors modifying LS biotransformation [12, 25]:

• Body mass index. The higher the body mass index, the higher the required maintenance dose of oral anticoagulants.

• The presence of concomitant diseases (renal and hepatic impairment, heart failure, arterial hypertension, diabetes mellitus, malignant neoplasms, anemia) may increase sensitivity to oral anticoagulants.

• Concomitant use of warfarin and LS that suppress the hemostasis system (clopidogrel, acetylsalicylic acid, dipyridamole, ticlopidine, non-steroidal anti-inflammatory drugs) leads to an increased risk of bleeding. The use of warfarin in combination with drugs possessing pronounced CYP2C9 inhibitory activity (cimetidine, chloramphenicol, sulfonamides) may lead to suppression of warfarin biotransformation with a subsequent increase in its plasma concentration, which increases the danger of bleeding.

• Consumption of certain foods rich in vitamin K (green vegetables, grapefruit and cranberry juice, etc.) weakens the effect of warfarin.

• Bad habits. Alcohol intake may increase sensitivity to oral anticoagulants, thereby increasing the risk of bleeding. Smokers, as a rule, require a higher maintenance dose of oral anticoagulants.

• Human genetic characteristics - carrier status of certain allelic variants of the biotransformation enzyme gene.

Immediately after IKS implantation, sensitivity to warfarin increases many-fold. A reduction in the loading dose of warfarin and frequent laboratory monitoring in the early postoperative (p/o) period makes it possible to avoid the development of excessive hypocoagulation and the appearance of bleeding [77, 89]. The therapeutic hypocoagulation range with warfarin use depends on the location and type of prosthesis. According to the recommendations of the American Heart Association and the American College of Cardiology (2008), for a mechanical AK prosthesis, an INR of 2.0-3.0 must be maintained when bileaflet prostheses are used, and 2.5-3.5 for all other disc valves. For a mechanical MK prosthesis, an INR in the range of 2.5-3.5 must be maintained [39].

According to data from different authors, in 20-95% of cases the ineffectiveness or NLR of the organism is associated with the genetic characteristics of patients taking certain LS. These genetic factors, as a rule, are "realized" by polymorphic regions of genes encoding proteins involved in the pharmacokinetics or pharmacodynamics of oral anticoagulants. It is now known that these factors are the most clinically significant in the pharmacokinetic process of LS, the efficacy and safety of which depend primarily on the biotransformation system and LS transporters.

With age and in patients taking many LS, the selection of a therapeutic dose of warfarin becomes even more complicated [85]. When several LS that are CYP2C9 inducers and inhibitors are combined, their biotransformation may change sharply and affect the concentration of LS in blood plasma (Table 1.1.1). A decrease in LS concentration in blood plasma is associated with ineffectiveness of the drug, whereas its increase leads to NLR, even to intoxication [63, 64, 116, 126]. The interaction of several LS becomes more complicated if one of them is a prodrug and undergoes biotransformation with the formation of several active metabolites. In this case, the safety and efficacy of the drugs is determined not only by the action of the parent drug, but also by the action of active metabolites.

Table 1.1.1. Interaction of indirect anticoagulants with other LS

Enhances the anticoagulant effect

• Amiodarone

• NSAIDs

• Digoxin

• Erythromycin

• Methyldopa

• Metronidazole

• Quinidine

• Cefamandole, moxalactam

• Levomycetin

• Omeprazole

• Chloramphenicol

Weakens the anticoagulant effect

• Diuretics

• Hormonal contraceptives

• Vitamin K

• Coenzyme Q10

• Mercaptopurine

• Ritonavir

• Rifampicin

• Cyclosporine

In the early p/o period in cardiac surgery patients taking warfarin, the concomitant use of LS must also be taken into account. For example, when amiodarone is prescribed simultaneously with oral anticoagulants, prolongation of prothrombin time and bleeding time due to CYP2C9 inhibition has been recorded. An increase in prothrombin time is detected 3-4 days after amiodarone prescription in patients who had previously received warfarin [55, 85, 99]. When amiodarone is prescribed at a dose of 400 mg/day, the warfarin dose must be reduced by 40%; at 300 mg/day by 35%; at 200 mg/day by 30%; and at 100 mg/day by 20% [21]. It is assumed that amiodarone can not only inhibit warfarin metabolism, but also reduce vitamin K absorption.

Lu Y. et al. (2008) analyzed data from 2434 patients receiving warfarin in combination with amiodarone, where in 43% (n = 1043) of cases INR was within the therapeutic range, in 34% (n = 820) INR was below the target range, and in 23% (n = 571) above the target range [82]. In total, in 102 patients (4%) the INR value was above 5. An INR rise above 5 with the combination of warfarin with amiodarone was observed during the first 12 weeks; thereafter, no peak INR increases were observed.

It has been proven that the metabolism of systemic antifungal agents, including voriconazole, proceeds via the cytochrome P450 system, one of whose isoenzymes is CYP2C9. Voriconazole can inhibit CYP2C9 activity and thereby increase the risk of bleeding [98]. According to some data, the concomitant use of warfarin with antibacterial and antiplatelet agents such as clopidogrel, aspirin, or non-steroidal anti-inflammatory drugs (NSAIDs) also leads to an enhancement of the anticoagulant effect of warfarin [83]. According to some authors, the prescription of corticosteroids leads to a reduction in the warfarin dose due to possible inhibition of warfarin metabolism via the cytochrome CYP3A system, which is partially involved in warfarin metabolism [47, 96]. At the same time, it must be remembered that diuretics can weaken the effect of warfarin. There is evidence of the influence of carbamazepine on the pharmacokinetics of warfarin in elderly patients with permanent atrial fibrillation (FP). In a study by Hermida et al. (2006), the prescription of carbamazepine in combination with warfarin led to an increase in the therapeutic dose of warfarin to an average of 9 mg/day compared with a group of patients not receiving carbamazepine who received warfarin at a dose of 3.86 mg/day [57]. Clinical cases of interaction between carbamazepine and indirect anticoagulants have also been described by other authors [35, 67, 95, 104, 108].

Thus, in the early postoperative period it is necessary to take into account the possible interaction of warfarin with LS that affect biotransformation processes. When prescribing or discontinuing drugs that affect warfarin metabolism, stricter INR monitoring must be carried out with subsequent dose adjustment of indirect anticoagulants.

1.2. Influence of cytochrome P-450 2C9 isoenzyme and the VKORC1 gene on the anticoagulant effect of warfarin

General information on the biotransformation system and LS transporters.

In most cases, effective and safe LS therapy depends on the concentration of the drug at the target molecule, which in turn depends on the processes of absorption, distribution and elimination. LS are eliminated from the body by biotransformation, most often via the liver and/or intestinal mucosa, and/or by excretion of LS through the kidneys with urine and/or through the liver with bile [24]. The greatest amount of the cytochrome P450 isoenzyme (CYP2C9) is located in the endoplasmic reticulum of hepatocytes of the liver, where phase I of biotransformation takes place, while phase II of LS biotransformation takes place with the participation of such enzymes as N-acetyl- and glucuronosyltransferase [14].

The cytochrome P-450 isoenzyme is one of the main enzymes that carry out the biotransformation of all known chemical compounds. At present, 53 genes and 24 pseudogenes of cytochrome P-450 located in different loci of different chromosomes have been identified [103]. At present, more than 1000 isoforms of cytochrome P-450 are known, which are divided into families, and these in turn into subfamilies. The isoenzymes CYP2C8, CYP2C9, CYP2C18, CYP2C19 belong to one of these subfamilies. CYP2C9 is the main enzyme of AKND (warfarin) metabolism [24].

Warfarin is presented as a racemic mixture of R- and S-warfarin stereoisomers, whose metabolism is carried out by different cytochrome P-450 isoenzymes. It should be noted that the main anticoagulant function is performed by S-warfarin, which is metabolized through the CYP2C9 genotype, whereas the metabolism of R-warfarin proceeds via other cytochrome isoenzymes [3]. The target for R- and S-warfarin is not only the K-epoxide reductase complex, but also vitamin K quinone reductase, which promotes the reduction of vitamin K quinone to vitamin K hydroquinone (Scheme 1.2.1).

R-WARFARIN

S-WARFARIN

6-hydroxyl 7-hydroxyl 10-hydroxyl

CYP2C9 7-hydroxywarfarin S-6-hydroxywarfarin

R-warfarin S-warfarin

Vitamin K

Reduced vitamin K

Vitamin K epoxide

Precursors of factors II, VII, IX, X of blood coagulation + proteins C and S

Vitamin K hydroquinone

Activated factors II, VII, IX, X of blood coagulation + proteins C and S

Ca2+

Binding of vitamin K-dependent factors to phospholipids

Scheme 1.2.1. Metabolism and anticoagulant effect of warfarin [Baidak D.V., 2007]

The cytochrome P-450 2C9 isoenzyme is localized on chromosome 10, locus 10q24.1-24.3, and consists of 55 thousand base pairs. CYP2C9 metabolizes more than 20% of all LS, as well as numerous endogenous compounds [2]. There are more than 20 allelic variants for this gene, but the most common of them is CYP2C9*1, the so-called "wild-type".

In individuals with the "wild-type" allelic variant, the biotransformation of coumarins, including warfarin, proceeds at normal intensity and the therapeutic INR range is achieved with the prescription of standard doses of anticoagulants [25]. The allelic variants of the "slow" type CYP2C9*2 and CYP2C9*3 of the CYP2C9 gene are also frequently encountered. These genotypes carry the amino acid substitutions CGT→TGT Arg144Cys and ATT→CTT Ile359Leu and possess reduced rates of metabolism of oral anticoagulants and other CYP2C9 drug substrates [23, 75, 117].

The allelic variant CYP2C9*4 has been detected only in the Mongoloid race; moreover, this allele is quite rare, and NLR with the use of LS metabolized via CYP2C9 occur more often in carriers of this allele [32]. Not long ago, "slow" alleles CYP2C9*5 and CYP2C9*6 were found in the Negroid race, with a frequency of occurrence of no more than 1% [25]. A high frequency of the allelic variant CYP2C9*8 has been found among African Americans. Genotyping of this allele in this racial group can predict the warfarin dose. In the Caucasian population, the following allelic variants have been identified: CYP2C9*14 (R125H), CYP2C9*15 (S162X), CYP2C9*16 (T299A), CYP2C9*17 (P382S), CYP2C9*18 (D397A), and CYP2C9*19 (Q454H); their influence on LS metabolism is under study [37, 46, 54]. The frequency of occurrence of CYP2C9 slow-metabolism genotypes in different ethnic groups is presented in Table 1.2.1.

Obviously, studies devoted to the influence of genetic aspects on anticoagulant therapy in different ethnic groups are highly relevant. Undoubtedly, such studies are also necessary in such a multinational country as Russia.

Table 1.2.1. Prevalence of "slow" CYP2C9 metabolizers in different populations [Sychev D.A., 2006]

White population of the USA — 0.06%

African Americans — 0.05%

Europeans — 1-3%

Chinese — 0.026%

Studies investigating the influence of carrier status of allelic variants CYP2C9*2 and CYP2C9*3 on warfarin anticoagulant therapy have shown that the risk of bleeding and hypocoagulation (INR above 4) in this category of patients increases several-fold [4, 60, 107, 122, 127]. At the same time, many studies have revealed an increase in S-warfarin plasma concentration and a decrease in warfarin clearance due to its slow metabolism [24, 58, 91]. Low selected warfarin doses and rapid achievement of the therapeutic INR range in patients carrying the "slow" allelic variants (CYP2C9*2 and CYP2C9*3) have been demonstrated in many studies. In 72% of cases, low therapeutic warfarin doses (less than 26.25 mg/week) were selected in this category of patients in studies conducted by Scordo M.G. et al. (2002) [111]. Peyvandi F. et al. (2004) found that in 65-66% of homozygous and heterozygous carriers of CYP2C9*2 and CYP2C9*3 allelic variants, on day 4 after starting warfarin the INR approached 3 [106]. Whereas in only 33% of homozygotes with the CYP2C9*1/*1 genotype, a similar INR value is achieved 4 days after warfarin treatment.

The influence of CYP2C9*2 and CYP2C9*3 allelic variants on anticoagulant activity has been studied not only in foreign [81, 94, 113] but also in domestic studies [4, 20, 22, 24]. Sirotkinа O.V. et al. (2004) and Mishin I.V. et al. (2007) obtained similar data in patients after IKS implantation - carriers of CYP2C9*2 and CYP2C9*3 allelic variants reached the therapeutic hypocoagulation range faster, which required a reduction in the warfarin dose [19, 22]. In the study of Bulytova Yu.M. et al. (2009), episodes of excessive hypocoagulation developed significantly more often (up to 85% of cases) in patients carrying CYP2C9*2 and CYP2C9*3 allelic variants with FP compared with patients carrying the "wild-type" [4]. The influence of CYP2C9*2 and CYP2C9*3 allelic variants on the warfarin dosing regimen in patients with permanent FP was demonstrated in the work of Mikheeva Yu.A. et al. (2008) [19]. In patients carrying the "wild-type" (CYP2C9*1/*1), the mean therapeutic warfarin dose was 6.1 mg/day, whereas in patients with CYP2C9*2 and CYP2C9*3 allelic variants the therapeutic warfarin dose was significantly lower - 3.6 and 3.1 mg/day.

VKORC1 is a target molecule for oral anticoagulants (including warfarin) located in the endoplasmic reticulum of hepatocytes [103]. Polymorphism of the VKORC1 gene is also currently associated with an inadequate response of the organism to the administration of oral anticoagulants. A mutation of the gene encoding subunit 1 of the VKORC1 complex is associated with coumarin resistance, and a congenital deficiency of vitamin K-dependent factors is associated with bleeding. More than 10 single-nucleotide substitutions in the VKORC1 gene are currently known.

A systematic review and meta-analysis have shown that in patients with VKORC1 gene polymorphism, hypocoagulation and INR values above the therapeutic range are noted more often, requiring careful INR monitoring when prescribing warfarin. Changes in the warfarin dosing regimen depending on the ethnic background of patients have also been noted [127].

The warfarin dosing regimen depending on VKORC (C1173T) gene polymorphism was studied in 147 patients included in a study conducted by D'Andréa et al. (2005) [44]. The mean therapeutic warfarin dose in carriers of the CC genotype for the C1173T polymorphic marker was 6.2 mg/day, in carriers of the CT genotype it was 4.8 mg/day. The presence of the TT genotype in patients was associated with lower warfarin doses - 3.5 mg/day. Similar results were obtained in studies by Hermida H.P. et al. (2005), conducted during patient use of acenocoumarol and phenprocoumon, and Baidak D.V. et al. (2007) in patients taking warfarin and acenocoumarol [3, 100].