SEARCH

SEARCH BY CITATION

Keywords:

  • pharmacodynamics;
  • pharmacogenetics;
  • pharmacokinetics;
  • (R)-warfarin;
  • VKORC1;
  • warfarin

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Competing Interests
  8. Acknowledgments
  9. REFERENCES

WHAT IS ALREADY KNOWN ABOUT THIS SUBJECT

• The contribution of (S)-warfarin to the clinical effect of rac-warfarin is well understood. The extent to which (R)-warfarin contributes to the clinical effect of rac-warfarin is unclear.

WHAT THIS STUDY ADDS

• Using unequivocally pure (R)- and (S)-warfarin we have demonstrated that (R)-warfarin contributes to the hypoprothrombinaemic effect of single large doses of warfarin.

• The extent of the interaction is dependent on VKORC1 genotype.

AIMS 1) To determine the pharmacokinetics and pharmacodynamics of (R)- and (S)-warfarin given alone and in combination and 2) to determine whether the relative potency of (R)- and (S)-warfarin is dependent on VKORC1 genotype.

METHODS A three way crossover study was conducted in which 17 healthy male subjects stratified by VKORC1 1173 C>T genotype and all CYP2C9 1*/1* received (R)-warfarin 80 mg, (S)-warfarin 12.5 mg and rac-warfarin sodium 25 mg. Plasma (R)- and (S)-warfarin unbound and total concentrations and prothrombin time were determined at multiple time points to 168 h.

RESULTS Pharmacokinetic parameters for (R)- and (S)-warfarin were similar to the literature. (R)-warfarin 80 mg alone resulted in a mean AUCPT (0,168 h) of 3550 s h (95% CI 3220, 3880). Rac-warfarin sodium 25 mg containing (S)-warfarin 11.7 mg produced a greater effect on AUCPT (0,168 h) than (S)-warfarin 12.5 mg (mean difference 250 s.h, 95% CI 110, 380, P < 0.002) given alone. In a mixed effects model the ratio of response between (R)- and (S)-warfarin (AUCPT((R)-warfarin) : AUCPT((S)-warfarin)) was 1.21 fold higher (95% CI 1.05, 1.41, P < 0.02) in subjects of VKORC1 TT genotype compared with the CC genotype.

CONCLUSIONS (R)-warfarin has a clear PD effect and contributes to the hypoprothrombinaemic effect of rac-warfarin. VKORC1 genotype is a covariate of the relative R/S potency relationship. Prediction of drug interactions with warfarin needs to consider effects on (R)-warfarin PK and VKORC1 genotype.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Competing Interests
  8. Acknowledgments
  9. REFERENCES

Rac-warfarin has been the main drug used worldwide for the prophylaxis of thromboembolic disease. The clinical use of warfarin remains problematic due to the 10-fold range in dosage requirements between individuals [1–5], narrow therapeutic index, complex clinical pharmacology, multiple reported drug interactions and the potentially severe consequences of both under-anticoagulation and over-anticoagulation.

Multiple patient factors are known to influence the response to warfarin including age, weight, gender, smoking status, vitamin K intake and mutations in coagulation factors [1, 6]. Polymorphisms in both the CYP2C9 and the VKORC1 genes have been shown to have a large contribution to the interindividual variation in warfarin dose [3–5, 7]. Despite accounting for all of these genetic and non-genetic factors, only 60% of the variability in warfarin dose requirements in Caucasians can be explained, suggesting other factors need to be considered.

One of the major barriers to the wider clinical use of warfarin remains the real and perceived risk of drug–drug interactions. In a systematic review by Holbrook et al. [8], over 120 drugs had been reported to be implicated in clinically significant interactions with warfarin. Anthony et al. [9] assessed three drug information compendia and the warfarin sodium (Coumadin™) USA product label for information regarding substances that interact with warfarin and found 648 entries from the four sources, with only 50 being common to all. Given that the majority of interactions with warfarin are of a pharmacokinetic basis, it should be possible to use a mechanism-based approach to predict the frequency and intensity of a particular drug interacting when used with warfarin. However the majority of interactions have been detected through non-systematic approaches such as sporadic case reporting [8].

The pharmacokinetics and pharmacodynamics of (R)- and (S)-warfarin differ. Both enantiomers are fully bioavailable [6] and hence it is variability in clearance which determines variability in exposure. The clearance of (S)-warfarin is almost exclusively (>85%) by CYP2C9 which catalyzes the 6- and 7- hydroxylation reactions [10], with (S)-7-hydroxywarfarin being the predominant metabolite [11]. The clearance of (R)-warfarin is more complicated, with the formation of 6-, 7-, 8- and 10-hydroxylated metabolites and reduction to an alcohol all contributing [12]. CYP1A2 is the principal enzyme catalyzing 6-hydroxylation (71%) but is minor for 7- and 8-hydroxylation [10]. CYP2C19 is the principal enzyme catalyzing 8-hydroxylation but is minor for 6- and 7-hydroxylation and CYP3A4 is exclusively responsible for 10-hydroxylation [10].

Several decades ago O'Reilly et al. [13] administered single doses of 1.5 mg kg−1 of separate (R)- and (S)-warfarin to 10 healthy volunteers. They found that (R)-warfarin had a lower apparent total clearance than (S)-warfarin such that the area under the plasma concentration–time curve (AUC) of (R)-warfarin was 1.9-fold greater that the AUC of (S)-warfarin. However for the same dose, (S)-warfarin showed a 1.8-fold greater hypoprothrombinaemic effect than (R)-warfarin as measured by the area under the prothrombin–time curve (AUCPT). This confirmed that both enantiomers had activity, but based on the relative AUC of the enantiomers and the hypothrombinaemic response, the estimated eudismic potency ratio of (R)-warfarin was approximately 3.4 times less than that of (S)-warfarin. Breckenridge and colleagues [14] titrated warfarin to a target INR (International Normalized Ratio, the standard reporting of prothrombin time) in healthy subjects with (R)-warfarin alone and (S)-warfarin alone using a steady-state crossover design and found an in vivo dose potency difference in humans of 1.6. This suggested that the dose potency of R : S warfarin was approximately 1.6:1 whereas the eudismic potency was approximately 4:1. These findings showed that although (S)-warfarin is the dominant species for action, the contribution of (R)-warfarin may be important. In addition these studies did not report on unbound plasma concentrations.

At variance with these findings, later studies in which the in vivo contribution of (R)-warfarin was sought from PK/PD [15] or population analyses [16] concluded that there was no significant contribution from (R)-warfarin and that the clinically significant biological effect was attributable to (S)-warfarin. Chan et al. [15] hypothesized that contamination of the allegedly pure (R)-warfarin by (S)-warfarin was a flaw in the previous studies [13, 14], despite the amount of contamination (8%) being insufficient to explain the results. These more recent studies have led to the widely held view that differences in the metabolism of (R)-warfarin, or drug interactions affecting (R)-warfarin metabolism, are unlikely to be clinically significant. However, of the long list of interactions with warfarin, many of the drugs are known to inhibit enzymes only responsible for the metabolism of (R)-warfarin. Hence it is plausible that these clinically significant interactions are mediated through the inhibition of (R)-warfarin metabolism.

The pivotal studies establishing the dose and concentration potency of warfarin enantiomers were completed several decades ago, prior to our current understanding of pharmacogenomics and ability to genotype, in small groups of subjects who were likely to be relatively homogeneous. On the basis of these studies, the potential significance of variability in (R)-warfarin metabolism both genetic and drug-related has been largely ignored or dismissed. This is despite numerous drugs (for example erythromycin, cimetidine, esomeprazole, fluvoxamine, quinolones, rosuvastatin [17]) being reported as potentially interacting with warfarin in the product information or case reports, where an interaction mediated via alteration in (R)-warfarin pharmacokinetics is more likely than alteration in (S)-warfarin pharmacokinetics. In addition, while it is clearly established that genotype of VKORC1 determines pharmacodynamic response to rac-warfarin, it is not known if there is an interaction between VKORC1 genotype and relative R : S potency. This would not necessarily be anticipated as the most clinically important VKORC1 genotypes result from the polymorphisms −1639G>A and 1173C>T occurring in the gene promoter and intron 1 of the gene respectively. These polymorphisms result in altered gene expression and are not known to affect the final protein structure [18–20]. There is strong linkage disequilibrium between these two polymorphisms [21].

These discrepant results may be partly attributable to heterogeneity in the clinical population with regard to the relative potencies of the enantiomers and heterogeneity in metabolic pathways. We aimed to examine the pharmacokinetics and pharmacodynamics of rac-warfarin and its R- and S-enantiomers, given individually, in subjects of different VKORC1 genotypes.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Competing Interests
  8. Acknowledgments
  9. REFERENCES

Participants

The intention was to recruit 18 healthy male participants between the ages of 18 and 45 years. All participants were genotyped for CYP2C9*1, *2 and *3 and only CYP2C9*1/*1 genotype subjects were included. By design there were to be equal distributions of homozygous wild-type, heterozygous and homozygous-variant for VKORC1 genotype (1173 C>T). Participants were healthy with normal ECG and calculated creatinine clearance by the Cockcroft–Gault method >80 ml min−1. Exclusion criteria were current illicit drug use, history of gastrointestinal bleeding, any regular medication use, ingestion of aspirin within 7 days of a study period or NSAIDs within five half-lives of any treatment period, alcohol consumption >two standard drinks a day on average and or more than four standard drinks in 1 day, current smoking, engaged in occupation or recreation which is at high risk of physical injury or trauma during the study period, BP>140/90 mmHg, abnormal fundoscopic examination, abnormal urinalysis, abnormal INR/APTT or platelet count at screening, history of bleeding disorders, clotting disorders or multiple miscarriage in first degree relatives or exclusively vegetarian diet.

Warfarin

To obtain pure (R)- and (S)-warfarin (C19H16O4, MW 308.33), rac-warfarin was synthesized using the method of Bush & Trager [22]. The racemic mixture was resolved by the method of West et al. [23] and checked for purity by melting point, optical rotation and h.p.l.c. [24] by the Chemistry Department of the University of Adelaide, Australia. (R)- and (S)-warfarin purity was confirmed as >99% pure with no evidence of cross enantiomer contamination. Pure (R)- and (S)-warfarin were prepared as a 50 ml suspension for dosing using the method of Sharley et al. [25]. For dosing rac-warfarin 25 mg of commercially available warfarin sodium (Coumadin™, Sigma Company Ltd, Clayton, Australia, C19H15NaO4, MW 330.3) tablets (5 × 5 mg) were crushed and prepared as a 50 ml suspension for administration.

Clinical study procedures

The study protocol was reviewed and approved by the Royal Adelaide Hospital Research Ethics Committee, Adelaide, Australia. Written informed consent for participation was obtained from all participants before enrolment in the study. Study participants were asked not to make significant changes in their diet for the duration of the study. We performed an open label, three way crossover study where participants on three separate occasions received 25 mg rac-warfarin sodium, 12.5 mg (S)-warfarin and 80 mg (R)-warfarin. A fixed dose design was chosen as the key comparisons were within subject. Each drug dose was at least 21 days apart. For each period, participants attended the study centre after an overnight fast from food for 10 h and water for 4 h. Subjects were then dosed with the warfarin suspension followed by a 2 h fast. Blood samples (n= 15 × 10 ml) were collected over 168 h for warfarin assay and prothrombin time at 0, 0.5, 1, 2, 4, 6, 8, 12, 24, 28, 72, 96, 120, 144 and 168 h. Subjects were directly questioned with regard to the occurrence of adverse events at these time points and at study closure. For the initial eight subjects enrolled in the protocol, dosing order was randomized. This was converted to open-label, fixed dose order (rac-warfarin 25 mg, (S)-warfarin 12.5 mg, (R)-warfarin 80 mg) when trace (R)-warfarin was detected in plasma following (R)-warfarin 80 mg despite at least a 21 day wash out period.

Analysis

Venous blood samples for warfarin assay were centrifuged and plasma stored at −80°C. Total (R)- and (S)-warfarin concentrations were determined by liquid chromatography tandem mass spectrometry (LC-MS/MS). The full details of the assay are reported elsewhere [26]. Briefly, 50 µl plasma was acidified with 5% formic acid and extracted by liquid–liquid extraction with methyl tertbutyl ether (MTBE) using d6-warfarin (Cerilliant, Round Rock, TX, USA) as internal standard. The supernatant was evaporated and reconstituted in 100 µl 40% methanol. A 5 µl aliquot was injected onto a chiral 150 × 2.1 mm Chirobiotic V column (Astec, Supelco, USA) and eluted isocratically with 40% methanol/0.03% acetic acid. MS/MS detection was performed using negative mode electrospray ionization, monitoring the ion transitions m/z 307/161 for warfarin and 313/161 for d6-warfarin, on an AB SCIEX 4000 Q TRAP mass spectrometer (AB Sciex, Foster City, Ca, USA). The assay was validated according to FDA guidelines [27]. Standard curves of 20–2000 µg l−1 of (R)- and (S)-warfarin were made by quadratic regression, weighted 1/X and r2≥ 0.995. Inter- and intra-day precision and inaccuracy were determined for the lower limit of quantification (LLOQ) of 20 µg l−1 and for quality controls (QCs) at 50, 500 and 2000 µg l−1 (n= 6 at each level, each run). Interday coefficient of variation (CV) was <11% and bias was <13% (n= 6). Intraday CV was <4% and bias <8% (n= 42). Samples >2000 µg l−1 were diluted with blank plasma before analysis and samples <20 µg l−1 were considered below the limit of quantification. Both (R)- and (S)-warfarin concentrations were determined in all plasma samples regardless of dosing group in order to assess for contamination or evidence of chiral inversion.

Unbound warfarin concentrations were measured in triplicate on the 1 h samples by ultrafiltration [26]. In short, 500 µl plasma was ultrafiltered for 15 min at 32°C and 2000 g using Centrifree 30 K ultrafiltration devices (Millipore, Bedford, MA, USA). There was no evidence of non-specific binding to the devices. Ultrafiltrate samples (100 µl) were acidified with 5% formic acid and extracted by liquid–liquid extraction with MTBE using d6-warfarin as internal standard. The supernatant was evaporated and reconstituted in 100 µl 40% methanol before injection of 10 µl onto the chiral Chirobioitc V column and detected by MS/MS as for total concentrations. Standard curves were made in spiked ultrafiltrate by quadratic regression, weighted 1/X with r2≥ 0.998. The analytical range was 1–20 µg l−1. LLOQ was 1 µg l−1 and QCs were 2.5, 10 and 20 µg l−1. CV was <10% and bias was <9% (n= 6 at LLOQ and each QC level). Samples >20 µg l−1 were diluted with blank ultrafiltrate before LC/MS analysis. Outliers from triplicate analysis were excluded based on deviation from mean >20%.

Genomic DNA was extracted using a QIAamp DNA mini kit (Qiagen, Australia). The CYP2C9 variants *2 and *3 were classified by detection of 430C>T (R144C, rs1799853) and 1075A>C (I359L, rs1057910), respectively, using Taqman Drug Metabolism Genotyping Assays and Taqman universal PCR master mix with Amperase UNG (Applied Biosystems, Australia) as per manufacturer's instructions. The VKORC1 1173C>T variant (C6484T, rs 9934438) was detected using a previously published real-time PCR method with minor modifications [28]. Control samples for all assays were confirmed by direct DNA sequencing.

Prothrombin time (s) was determined by STAR Evolution Coagulation Analyzer, (Stago, Doncaster, Australia) (SA Pathology). The within run precision of this automated testing is a CV of <3%.

Pharmacokinetics

The pharmacokinetic parameters AUC(0,∞), tmax, Cmax, t1/2 for (R)- and (S)-warfarin were calculated by noncompartmental analysis using PKSolver [29] in Microsoft Excel. Apparent total oral clearance was calculated as dose/AUC(0,∞) for R- and S-warfarin. For rac-warfarin the dose of each enantiomer of warfarin administered was considered to be 11.67 mg after accounting for the molecular weight of the salt. AUC(0,∞) for the unbound fraction (AUC(0,∞ unbound)) for (R)- and (S)-warfarin was approximated by multiplying the fraction unbound (fu) at 1 h by AUC(0,∞). Unbound clearance (CLu) was subsequently estimated using AUC(0,∞unbound).

Pharmacodynamic response

The pharmacodynamic response was calculated as the AUC of prothrombin time from 0 to 168 h (AUCPT) calculated using the linear trapezoidal rule. The response was also assessed as the AUC of change in prothrombin time after subtraction of PT at time 0 (AUCPTADJ). In addition the ratio of response to (R)-warfarin 80 mg was compared with (S)-warfarin 12.5 mg with regard to VKORC1 genotype using a mixed model approach.

Statistical analysis

Measurements from the same participants were compared by paired t-tests for normally distributed data. Comparisons between participants were made by unpaired t-tests or one way anova. For the pharmacodynamic response repeated measures anova was used to compare the AUCPT response. The data were analyzed using GraphPad Instat version 3.00 for Windows (GraphPad Software, San Diego, California USA, http://www.graphpad.com). P < 0.05 was considered significant. A mixed model approach accounting for treatment period, sequence and genotype was used to examine the interaction between VKORC1 genotype and the relative potency of (R)- and (S)-warfarin (SAS 9.2, SAS Institute Inc., Cary, North Carolina USA) as reflected in AUCPT((R)-warfarin) : AUCPT((S)-warfarin). Data are expressed as ratios with 95% confidence intervals.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Competing Interests
  8. Acknowledgments
  9. REFERENCES

Seventeen male participants of mean age 26 (range 17–41) years and mean weight 76 (range 55–105) kg were recruited into the study of whom six participants were VKORC1 CC (homozygous wild-type), six participants VKORC1 CT (heterozygotes) and five were VKORC1 TT (homozygous variant). Baseline demographic characteristics are summarized in Table 1. Subjects with CT genotype were older than those recruited with CC or TT genotype (P < 0.02). The weight of subjects with TT genotype was less than those recruited with the CC or CT genotype (P < 0.02). Adverse events reported during the trial were all minor and were not attributed to study drug. There were no serious adverse events and no instances of bleeding. There was no difference in tolerability of study drug on the three dosing occasions.

Table 1. Baseline demographic characteristics of participants stratified by VKORC1 genotype (1173C>T)
VKORC1 genotype n Age (years) (range) Weight (kg) (range)
  • *

    P < 0.02.

CC623 (17–26)82 (74–91)
CT633 (23–41)*81 (56–105)
TT523 (19–27)61 (55–70)*

Pharmacokinetics

There was no evidence of contamination or chiral inversion in the plasma sample analysis. The pharmacokinetic parameters from the three dose groups are summarized in Table 2. The exposure to (S)-warfarin 12.5 mg administered alone was greater than the exposure to (S)-warfarin 11.7 mg administered as rac-warfarin. The ratio of total AUC(S)-warfarin alone : total AUC(S)-warfarin racemic was 1.12 (95% CI 1.01, 1.24). When approximating unbound AUC(S)-warfarin alone : unbound AUC(S)-warfarin racemic the ratio was similar at 1.14 (95% CI 0.96, 1.32).

Table 2. Pharmacokinetic parameters of (R)-warfarin and (S)-warfarin administered alone and in racemic combination (n= 17). Note effective dose of racemic (R)- and (S)-warfarin 11.7 mg due to administration as a salt. Data are mean (95% CI) or (range)
  C max (95% CI) (µg ml−1) t max (range) (h) AUC(0,∞) (95% CI) (µg ml−1 h) t 1/2 (95% CI) (h) CL/F (95% CI) (ml h−1) % unbound warfarin (95% CI) AUC(0,∞unbound) (95% CI) (µg ml−1 h) CLunbound/F (95% CI) (l h−1)
(R)-warfarin 80 mg alone7.69 (6.78, 8.59)2.4 (0.5–12)504 (442, 569)51.1 (45.9, 56.4)168 (146, 189)0.98 (0.86, 1.1)4.8 (4.2, 5.5)18 (15, 21)
(R)-warfarin 11.7 mg racemic1.34 (1.17, 1.50)0.9 (0.5 −2)74.4 (64.9, 84.0)49.3 (44.8, 53.9)166 (145, 186)0.99 (0.88, 1.1)0.73 (0.62, 0.84)17 (15, 20)
(S)-warfarin 12.5 mg alone1.36 (1.20, 1.53)0.8 (0.5–2)44.5 (38.5, 50.6)33.1 (29.7, 36.5)299 (259, 340)0.72 (0.65, 0.79)0.32 (0.27, 0.37)43 (36, 50)
(S)-warfarin 11.7 mg racemic1.37 (1.20, 1.53)0.9 (0.5–2)40.1 (35.3, 44.8)34.1 (30.3, 38.0)306 (270, 343)0.72 (0.66, 0.78)0.29 (0.25, 0.32)43 (38, 48)

Total (S)-warfarin clearance was higher than total (R)-warfarin clearance (difference 131 ml h−1, 95% CI 91.0, 172, P < 0.0001). Similarly unbound (S)-warfarin clearance was higher than unbound (R)-warfarin clearance (difference 25 l h−1, 95% CI 16, 33, P < 0.0001). The mean difference between unbound (S)-warfarin clearance administered alone or as racemate was not significantly different (0.1 l h−1, 95% CI −6.0, 2.6). Similarly, the mean difference in unbound (R)-warfarin clearance when administered alone or as a racemate was not significantly different (0.5 l h−1, 95% CI −3.4, 4.4).

Pharmacokinetics by genotype for the pure (R)-warfarin and (S)-warfarin are summarized in Tables 3A and 3B. There was no significant difference in apparent oral clearance of total or unbound (R)-warfarin or (S)-warfarin between VKORC1 genotypes (P= 0.48 total (R)-warfarin, P= 0.82 total (S)-warfarin, P= 0.88 unbound (R)-warfarin, P= 0.11 unbound (S)-warfarin).

Table 3A. Pharmacokinetics of (R)-warfarin given as a single 80 mg oral dose according to VKORC1 genotype (n= 17). Data are mean (95% CI) or mean (range)
(R)-warfarin 80 mg alone C max (95% CI) (µg ml−1) t max (range) (h) AUC (95% CI) (µg ml−1 h) t 1/2 (95% CI) (h) CL/F (95% CI) (ml h−1) AUC(0,∞unbound) (95% CI) (µg ml−1 h) CLunbound/F (95% CI) (l h−1)
CC (n= 6)7.97 (6.09, 9.85)2.1 (0.5–8)462 (354, 570)53.0 (45, 61)181 (134, 229)4.4 (4.0, 4.9)18 (16, 20)
CT (n= 6)6.82 (5.17, 8.47)3.5 (0.5–1)542 (442, 643)56.1 (50, 63)151 (124, 179)5.0 (3.8, 6.2)17 (12, 21)
TT (n= 5)8.39 (6.06. 10.7)1.6 (0.5–6)509 (295, 723)42.9 (27, 59)171 (103, 155)5.1 (2.6, 7.6)19 (8, 30)
Table 3B. Pharmacokinetics of (S)-warfarin given as a single 12.5 mg oral dose according to VKORC1 genotype (n= 17). Data are mean (95% CI) or (range)
(S)-warfarin 12.5 mg alone C max (95% CI) (µg ml−1) t max (range) (h) AUC (95% CI) (µg ml−1 h) t 1/2 (95% CI) (h) CL/F (95% CI) (ml h−1) AUC(0,∞unbound) (95% CI) (µg ml−1 h) CLunbound/F (95% CI) (l h−1)
CC (n= 6)1.25 (0.86, 1.63)0.75 (0.5–2)43.9 (28.3, 59.4)34.8 (28.4, 41.2)312 (202, 423)0.32 (0.23, 0.40)42 (31, 53)
CT (n= 6)1.35 (1.13, 1.57)1 (0.5–2)43.3 (32.1, 54.6)30.6 (26.7, 34.4)301 (234, 369)0.27 (0.18, 0.36)51 (35, 67)
TT (n= 5)1.51 (1.01, 2.02)0.6 (0.5–1)46.8 (32.9, 60.7)34.2 (22.2, 46.1)281 (193, 368)0.39 (0.27, 0.51)34 (23, 45)

Trace (R)-warfarin concentrations were detected in four of the first eight subjects during the pure (S)-warfarin dosing period. These concentrations were all <0.06 µg ml−1 compared with Cmax (R)-warfarin 80 mg 7.69 µg ml−1 and Cmax (S)-warfarin 12.5 mg 1.36 µg ml−1. In subsequent subjects following the change to fixed dosing these residual concentrations of (R)-warfarin were no longer detected with administration of pure (S)-warfarin 12.5 mg.

Warfarin pharmacodynamics

All PTs had returned to the normal range (12–16 s) prior to beginning the next dosing period and there was no significant difference between baseline PT between dosing periods (P= 0.70). All three treatments produced a clear response–time profile (Figure 1, Table 4). The AUCPT of rac-warfarin sodium 25 mg was 2900 s h (95% CI 2700, 3100) compared with the AUCPT of (S)-warfarin 12.5 mg of 2650 s h (95% CI 2560, 2740), resulting in a mean difference of 250 s h (95% CI 110, 380, P < 0.002).

image

Figure 1. Mean (±SEM) prothrombin time (PT) response in healthy participants (n= 17) administered single dose (R)-warfarin 80 mg, (S)-warfarin 12.5 mg and rac-warfarin 25 mg. inline image, (R)-warfarin 80 mg; inline image, (S)-warfarin 12.5 mg; inline image, rac-warfarin 25 mg

Download figure to PowerPoint

Table 4. Prothromin time (PT) response by dose occasion for (R)-warfarin 80 mg, (S)-warfarin 12.5 mg and rac-warfarin 25 mg in 17 healthy subjects. Note effective dose of racemic (R)- and (S)-warfarin 11.7 mg due to administration as a salt
Dose AUCPT (s h) (range) AUCPTADJ (s h) (range)
  1. Data are mean ± SD (range). AUCPT is area under the PT–time curve from time 0 to 168 h. AUCPTADJ is area under the PT–time curve after adjustment for baseline PT from time 0 to 168 h.

R-warfarin3550 ± 638 (2500–5350)1120 ± 602 (198–2800)
S-warfarin2650 ± 178 (2400–3080)233 ± 154 (24–579)
Racemic warfarin2900 ± 382 (2430–4020)458 ± 369 (16–1570)

The pharmacodynamic response to (R)-warfarin 80 mg was greater than the response to either rac-warfarin (difference in AUCPT 649 s h, 95% CI 502, 795, P < 0.0001) or (S)-warfarin (difference in AUCPT 895 s h, 95% CI 623, 1170, P < 0.0001). After subtracting for PT at time 0, rac-warfarin 25 mg demonstrated a 2-fold greater pharmacodynamic response over 168 h than (S)-warfarin 12.5 mg alone (mean difference in AUCPTADJ 225 s h, 95% CI 78.0, 225, P < 0.01).

Mean response to (R)-warfarin 80 mg and (S)-warfarin 12.5 mg by VKORC1 genotype is shown in Figure 2. Analysis of the interaction between VKORC1 genotype and the relative potency of (R)- and (S)-warfarin on PT (AUCPT((R)-warfarin) : AUCPT((S)-warfarin)) was significant overall. Compared with the CC genotype, the ratio for TT genotype was 1.21-fold higher (95% CI 1.05, 1.41, P < 0.02). The ratio for the CT genotype was intermediate between those of CC and TT being 1.07-fold (95% CI 0.97, 1.18) higher than CC but 1.13-fold (95% CI 0.97, 1.32) lower than TT but these differences were not statistically significant (P= 0.11). There was no significant difference in mean response by sequence (P= 0.56) or period (P= 0.38).

image

Figure 2. Mean (±SEM) AUCPT response in participants (n= 17) administered single dose (R)-warfarin 80 mg and (S)-warfarin 12.5 mg by VKORC1 (1173C>T) genotype. inline image, (R)-warfarin 80 mg; inline image, (S)-warfarin 12.5 mg

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Competing Interests
  8. Acknowledgments
  9. REFERENCES

This is the first human study in which unequivocally pure (R)-warfarin demonstrated a therapeutic effect both alone and when administered with (S)-warfarin, replicating the findings of O'Reilly [13] and Breckenridge et al. [14] from three decades ago where the purity of warfarin was questioned.

The 80 mg dose of (R)-warfarin was intentionally chosen given doubt had previously been raised with regard to its potency [15, 16] and we wished to maximize our chances of achieving a pharmacodynamic response. (S)-warfarin 12.5 mg was given alone in order to ascertain the pharmacodynamic contribution of (R)-warfarin when rac-warfarin 25 mg was administered. The AUCPT of rac-warfarin sodium 25 mg was greater than the AUCPT of (S)-warfarin 12.5 mg confirming the contribution of (R)-warfarin to the pharmacodynamic response of rac-warfarin. The greatest pharmacodynamic response was to (R)-warfarin 80 mg, attributable to the large dose.

Our findings are at variance with previous population pharmacokinetic studies such as those of Hamberg et al. [16] which have failed to detect concentrations of (R)-warfarin as significant covariates of the PT response. This discrepancy is likely to be due to the different methodology. We have used a carefully controlled crossover experiment in which (R)-warfarin was administered at two different doses which enabled a clear estimate of its effect. The analysis of Hamberg et al. came from populations where patients on potentially interacting concomitant medications were excluded, reducing the sensitivity to detect the effects of elevated concentrations of (R)-warfarin due to metabolic inhibitory concomitant medication. It is noted to be difficult in population analysis to detect the effect of a covariate in the presence of a related covariate with much larger effect. In the study of Kerbusch et al. [30], in which population PK/PD was used to estimate the in vivo potency of a metabolite, a situation analogous to that in the current study, it was noted that precision was highly dependent on a specific dataset in which the parent to metabolite ratios were remarkably different from the other studies, enabling signal detection. Hence the inability to detect (R)-warfarin as a significant covariate in population analysis should not be taken as evidence against the clear demonstration of effect in this study.

A potential criticism of the applicability of our study to the clinical use of warfarin is that it used single dose data rather than steady-state. However, well-accepted PK/PD models of warfarin [31] have incorporated single dose pharmacokinetics. Additionally in the population PK/PD study of Hamberg et al. [16], slightly more than half of the data came from single dose administration which was combined with data following multiple dose administration, confirming that both types of data are compatible and informative. Furthermore the agreement between single dose and multiple dose approaches with regard to the relative in vivo potency of (R)- and (S)-warfarin by O'Reilly [13] and Breckenridge et al. [14], respectively, suggests that our single dose design is appropriate to investigate relative clinical potencies.

A fixed dose design was selected as opposed to a body weight adjusted regimen as our principal comparison with regard to the relatively potency of (R)- and (S)-warfarin were within subject comparisons. Although the intention was to dose 25 mg rac-warfarin and 12.5 mg (S)-warfarin in order to achieve identical exposure to (S)-warfarin, different formulations of warfarin were used by necessity as we had limited quantities of pure (R)- and (S)-warfarin. The effective dose of (R)-warfarin and (S)-warfarin administered with racemic dosing from the commercial tablet was 11.7 mg. This resulted in approximately 10% less exposure to (S)-warfarin as determined by total AUC when it was administered in the racemate compared with the administration of 12.5 mg pure (S)-warfarin which is entirely accounted for by the use of warfarin sodium for racemic dosing. Some subjects showed delayed tmax when administered (R)-warfarin 80 mg, however drug exposure was dose proportional as reflected by the identical apparent oral clearance for both (R)-warfarin and (S)-warfarin when given alone or together. Despite the use of the two formulations, the pharmacokinetics of the enantiomers when given alone or together in this study were consistent with those reported in the literature [31]. Additionally, the apparent oral clearance estimates for each enantiomer were almost identical from each formulation, suggesting they were bioequivalent.

Pure (R)-warfarin administered alone showed a robust pharmacodynamic response. In addition, (R)-warfarin made a clear contribution to the activity of the racemate as exhibited by greater pharmacodynamic response to the racemate compared with the (S)-warfarin component given alone. The hypoprothrombinaemic effect of (R)-warfarin in rac-warfarin may in fact be underestimated in this study, given the exposure to (S)-warfarin in the racemic dosing was less than with (S)-warfarin alone and yet a clearly greater pharmacodynamic response resulted. This is contrary to the conventional wisdom that (R)-warfarin makes little contribution to the clinical effect of rac-warfarin.

For both enantiomers, VKORC1 genotype was a determinant of response, with participants of TT genotype being both more sensitive to (S)-warfarin and relatively more sensitive to (R)-warfarin. Given that the effect size was modest and the sample size was small this needs to be interpreted with caution, but is potentially important as it suggests that patients with TT genotype might show disproportionately more anticoagulant response following co-administration of an inhibitor of (R)-warfarin metabolism. A mechanism by which VKORC1 genotype alters the eudismic potency of the warfarin enantiomers is not readily apparent as the 1173 C>T mutation is in a non-coding part of the gene and hence is not expected to result in structural modification of the protein product, although may affect protein stability. Alternatively the mutation may be a marker for other genetic changes which affect the drug–protein interaction.

VKORC1 genotype had no apparent impact on either pharmacokinetic parameters including clearance, despite previous speculation that there may be an interaction between VKORC1 genotype and clearance of warfarin enantiomers [32]. It has recently been demonstrated that polymorphisms in CYP2C19 and CYP3A4 do impact on (R)-warfarin clearance but we did not measure these in this study [33]. However any mutation reducing (R)-warfarin clearance without affecting (S)-warfarin clearance will increase the R : S dose potency ratio in those individuals.

Given the large number of drug interactions with warfarin, many of which are not readily explained by alterations in (S)-warfarin clearance (CYP2C9), alterations in vitamin K production or pharmacodynamic factors this supports the contention that these interactions may be mediated through changes in (R)-warfarin clearance and thus enhanced pharmacodynamic effect. There is evidence to suggest that (R)-warfarin drug interactions are clinically significant, although there has been inconsistency when this has been formally examined. Given that typical warfarin interaction studies have small sample sizes, often 12 participants, this might be an explanation for case reports of clinically significant interactions occurring when prospective studies in volunteers have not shown a significant overall effect. For example, there are case reports of patients stable on warfarin who developed clinical overanticoagulation with co-administration of cimetidine [34], an inhibitor of CYP1A2 and CYP3A4. This was studied by O'Reilly [35] who examined the anticoagulant effect of 1.5 mg kg−1rac-warfarin in combination with 1200 mg cimetidine and found it augmented both the hypoprothrombinaemic effects and blood concentrations of (R)-warfarin, a clinically significant result. Later, Toon et al. [36] found that in 12 healthy male volunteers, 800 mg cimetidine caused an approximately 20% reduction in the clearance of (R)-warfarin with no effect on (S)-warfarin and no effect on INR. Toon et al. concluded that this interaction would not have clinical consequences. The relative contribution of (R)-warfarin to drug–drug interactions may be dependent on VKORC1 genotype and hence drug–drug interaction studies where (R)-warfarin is a possible suspect may be falsely reassuring unless VKORC1 genotype has been taken into account.

This study has established the basics of the PK/PD relationship of the individual enantiomers of warfarin in man. From these data it will be possible, using PK/PD modelling combined with in vitro drug metabolism data to simulate the likely frequency and extent of pharmacokinetic drug–drug interaction with warfarin in a population. Another important finding from this study is that in addition to its known effect on warfarin dose requirements, VKORC1 genotype may be a determinant of the relative contribution of (R)-warfarin to the clinical effect of warfarin and thus to drug–drug interactions. Hence such modelling should include VKORC1 genotype as a potential covariate, in addition to genotypes for relevant drug metabolizing polymorphisms. (R)-warfarin has anticoagulant activity and cannot be dismissed as having no relevance.

Competing Interests

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Competing Interests
  8. Acknowledgments
  9. REFERENCES

All authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declare this work has been funded internally by the Discipline of Pharmacology, University of Adelaide. No support from any external organization has been received for the submitted work. There are no financial relationships with any organizations that might have an interest in the submitted work in the previous 3 years. There are no other relationships or activities that could appear to have influenced the submitted work.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Competing Interests
  8. Acknowledgments
  9. REFERENCES

We would like to thank Dr Peter Valente (University of Adelaide) for synthesizing the warfarin enantiomers, Peter Slobodian, Nick Sharley, Valerie Sim and Melissa Teo (Royal Adelaide Hospital Pharmacy) for purification, testing and formulation of the warfarin enantiomers suitable for human use, Dr Dan Barratt (Discipline of Clinical Pharmacology, University of Adelaide) for performing DNA extraction, Victoria Tsagareli (Pain and Anaesthesia Research Clinic, University of Adelaide) for clinical study coordination, Michelle Lorimer (University of Adelaide) for performing the statistical modelling of the effect of VKORC1 genotype on enantiomer potency ratio and Prof Mats O Karlsson (Uppsala University) for helpful discussions.

REFERENCES

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Competing Interests
  8. Acknowledgments
  9. REFERENCES
  • 1
    Sconce EA, Khan TI, Wynne HA, Avery P, Monkhouse L, King BP, Wood P, Kesteven P, Daly AK, Kamali F. The impact of CYP2C9 and VKORC1 genetic polymorphism and patient characteristics upon warfarin dose requirements: proposal for a new dosing regimen. Blood 2005; 106: 232933.
  • 2
    Wadelius M, Chen LY, Downes K, Ghori J, Hunt S, Eriksson N, Wallerman O, Melhus H, Wadelius C, Bentley D, Deloukas P. Common VKORC1 and GGCX polymorphisms associated with warfarin dose. Pharmacogenomics J 2005; 5: 26270.
  • 3
    Millican EA, Lenzini PA, Milligan PE, Grosso L, Eby C, Deych E, Grice G, Clohisy JC, Barrack RL, Burnett RS, Voora D, Gatchel S, Tiemeier A, Gage BF. Genetic-based dosing in orthopedic patients beginning warfarin therapy. Blood 2007; 110: 15115.
  • 4
    Zhu Y, Shennan M, Reynolds KK, Johnson NA, Herrnberger MR, Valdes JR, Linder MW. Estimation of warfarin maintenance dose based on VKORC1 (-1639 G>A) and CYP2C9 genotypes. Clin Chem 2007; 53: 1199205.
  • 5
    Aquilante CL, Langaee TY, Lopez LM, Yarandi HN, Tromberg JS, Mohuczy D, Gaston KL, Waddell CD, Chirico MJ, Johnson JA. Influence of coagulation factor, vitamin K epoxide reductase complex subunit 1, and cytochrome P450 2C9 gene polymorphisms on warfarin dose requirements. Clin Pharmacol Ther 2006; 79: 291302.
  • 6
    Majerus PWeaDMT. Blood coagulation and anticoagulation, thrombolytic, and antiplatelet drugs. In: Goodman and Gilman's The Pharmacological Basis of Therapeutics, 11th edn. ed. Brunton LL. New York: McGraw-Hill, 2006; 1477.
  • 7
    Yin T, Miyata T. Warfarin dose and the pharmacogenomics of CYP2C9 and VKORC1 – rationale and perspectives. Thromb Res 2007; 120: 110.
  • 8
    Holbrook AM, Pereira JA, Labiris R, McDonald H, Douketis JD, Crowther M, Wells PS. Systematic overview of warfarin and its drug and food interactions. Arch Intern Med 2005; 165: 1095106.
  • 9
    Anthony M, Romero K, Malone DC, Hines LE, Higgins L, Woosley RL. Warfarin interactions with substances listed in drug information compendia and in the FDA-approved label for warfarin sodium. Clin Pharmacol Ther 2009; 86: 4259.
  • 10
    Kunze KL, Trager WF, Warfarin-fluconazole III. A rational approach to management of a metabolically based drug interaction. Drug Metab Dispos 1996; 24: 42935.
  • 11
    Rettie AE, Eddy AC, Heimark LD, Gibaldi M, Trager WF. Characteristics of warfarin hydroxylation catalyzed by human liver microsomes. Drug Metab Dispos 1989; 17: 26570.
  • 12
    Black DJ, Kunze KL, Wienkers LC, Gidal BE, Seaton TL, McDonnell ND, Evans JS, Bauwens JE, Trager WF, Warfarin-fluconazole II. A metabolically based drug interaction: in vivo studies. Drug Metab Dispos 1996; 24: 4228.
  • 13
    O'Reilly RA. Studies on the optical enantiomorphs of warfarin in man. Clin Pharmacol Ther 1974; 16: 34854.
  • 14
    Breckenridge A, Orme M, Wesseling H, Lewis RJ, Gibbons R. Pharmacokinetics and pharmacodynamics of the enantiomers of warfarin in man. Clin Pharmacol Ther 1974; 15: 42430.
  • 15
    Chan E, McLachlan A, O'Reilly R, Rowland M. Stereochemical aspects of warfarin drug interactions: use of a combined pharmacokinetic-pharmacodynamic model. Clin Pharmacol Ther 1994; 56: 28694.
  • 16
    Hamberg AK, Dahl ML, Barban M, Scordo MG, Wadelius M, Pengo V, Padrini R, Jonsson EN. A PK-PD model for predicting the impact of age, CYP2C9, and VKORC1 genotype on individualization of warfarin therapy. Clin Pharmacol Ther 2007; 81: 52938.
  • 17
    eMIMs. Warfarin Product Information. eMIMs Version 5.01.0093. CMP Medica Australia, 2008.
  • 18
    D'Andrea G, D'Ambrosio RL, Di Perna P, Chetta M, Santacroce R, Brancaccio V, Grandone E, Margaglione M. A polymorphism in the VKORC1 gene is associated with an interindividual variability in the dose-anticoagulant effect of warfarin. Blood 2005; 105: 6459.
  • 19
    Rieder MJ, Reiner AP, Gage BF, Nickerson DA, Eby CS, McLeod HL, Blough DK, Thummel KE, Veenstra DL, Rettie AE. Effect of VKORC1 haplotypes on transcriptional regulation and warfarin dose. N Engl J Med 2005; 352: 228593.
  • 20
    Wang D, Chen H, Momary KM, Cavallari LH, Johnson JA, Sadee W. Regulatory polymorphism in vitamin K epoxide reductase complex subunit 1 (VKORC1) affects gene expression and warfarin dose requirement. Blood 2008; 112: 101321.
  • 21
    Limdi NA, Wadelius M, Cavallari L, Eriksson N, Crawford DC, Lee MT, Chen CH, Motsinger-Reif A, Sagreiya H, Liu N, Wu AH, Gage BF, Jorgensen A, Pirmohamed M, Shin JG, Suarez-Kurtz G, Kimmel SE, Johnson JA, Klein TE, Wagner MJ. Warfarin pharmacogenetics: a single VKORC1 polymorphism is predictive of dose across 3 racial groups. Blood 2010; 115: 382734.
  • 22
    Bush E, Trager WF. High-yield synthesis of warfarin and its phenolic metabolites: new compounds. J Pharm Sci 1983; 72: 8301.
  • 23
    West BD, Preis S, Schroeder CH, Link KP. Studies on the 4-hydroxycoumarins. XVII.1a the resolution and absolute configuration of warfarin1b. J Am Chem Soc 1961; 83: 267679.
  • 24
    de Vries JX, Schmitz-Kummer E. Direct column liquid chromatographic enantiomer separation of the coumarin anticoagulants phenprocoumon, warfarin, acenocoumarol and metabolites on an [alpha]1-acid glycoprotein chiral stationary phase. J Chromatogr A 1993; 644: 31520.
  • 25
    Sharley NA, Yu AMC, Williams DB. Stability of mixtures formulated from warfarin tablets or powder. J Pharm Pract Res 2007; 37: 957.
  • 26
    Jensen BP, Chin PK, Begg EJ. Quantification of total and free concentrations of R- and S-warfarin in human plasma by ultrafiltration and LC-MS/MS. Anal Bioanal Chem 2011; 401: 218793.
  • 27
    US Department of Health and Human Services. US FDA Center for Drug Evaluation and Research M, USA. 2001). Guidance for industry: bioanalytical method validation.
  • 28
    Li T, Lange LA, Li X, Susswein L, Bryant B, Malone R, Lange EM, Huang TY, Stafford DW, Evans JP. Polymorphisms in the VKORC1 gene are strongly associated with warfarin dosage requirements in patients receiving anticoagulation. J Med Genet 2006; 43: 7404.
  • 29
    Zhang Y, Huo M, Zhou J, Xie S. PKSolver: an add-in program for pharmacokinetic and pharmacodynamic data analysis in Microsoft Excel. Comput Methods Programs Biomed 2010; 99: 30614.
  • 30
    Kerbusch T, Milligan PA, Karlsson MO. Assessment of the relative in vivo potency of the hydroxylated metabolite of darifenacin in its ability to decrease salivary flow using pooled population pharmacokinetic-pharmacodynamic data. Br J Clin Pharmacol 2004; 57: 17080.
  • 31
    Holford NH. Clinical pharmacokinetics and pharmacodynamics of warfarin. Understanding the dose-effect relationship. Clin Pharmacokinet 1986; 11: 483504.
  • 32
    Osman A, Enstrom C, Lindahl TL. Plasma S/R ratio of warfarin co-varies with VKORC1 haplotype. Blood Coagul Fibrinolysis 2007; 18: 2936.
  • 33
    Lane S, Al-Zubiedi S, Hatch E, Matthews I, Jorgensen AL, Deloukas P, Daly AK, Park BK, Aarons L, Ogungbenro K, Kamali F, Hughes D, Pirmohamed M. The population pharmacokinetics of R- and S-warfarin: effect of genetic and clinical factors. Br J Clin Pharmacol 2012; 73: 6676.
  • 34
    Serlin MJ, Sibeon RG, Mossman S. Cimetidine: interaction with oral anticoagulants in man. Lancet 1979; 2: 31719.
  • 35
    O'Reilly RA. Comparative interaction of cimetidine and ranitidine with racemic warfarin in man. Arch Intern Med 1984; 144: 98991.
  • 36
    Toon S, Hopkins KJ, Garstang FM, Rowland M. Comparative effects of ranitidine and cimetidine on the pharmacokinetics and pharmacodynamics of warfarin in man. Eur J Clin Pharmacol 1987; 32: 16572.