The aim of this study was to evaluate the effect of the antiepileptic drug lacosamide on the pharmacokinetics and pharmacodynamics of the anticoagulant warfarin.
The aim of this study was to evaluate the effect of the antiepileptic drug lacosamide on the pharmacokinetics and pharmacodynamics of the anticoagulant warfarin.
In this open-label, two-treatment crossover study, 16 healthy adult male volunteers were randomized to receive a single 25-mg dose of warfarin alone in one period and lacosamide 200 mg twice daily on days 1–9 with a single 25 mg dose of warfarin coadministered on day 3 in the other period. There was a 2-week washout between treatments. Pharmacokinetic end points were area under the plasma concentration–time curve (AUC(0,last) and AUC(0,∞)) and maximum plasma concentration (Cmax) for S- and R-warfarin. Pharmacodynamic end points were area under the international normalized ratio (INR)–time curve (AUCINR), maximum INR (INRmax), maximum prothrombin time (PTmax) and area under the PT-time curve (AUCPT).
Following warfarin and lacosamide coadministration, Cmax and AUC of S- and R-warfarin, as well as peak value and AUC of PT and INR, were equivalent to those after warfarin alone. In particular, the AUC(0,∞) ratio (90% confidence interval) for coadministration of warfarin and lacosamide versus warfarin alone was 0.97 (0.94–1.00) for S-warfarin and 1.05 (1.02–1.09) for R-warfarin, and the AUCINR ratio was 1.04 (1.01–1.06). All participants completed the study.
Coadministration of lacosamide 400 mg/day did not alter the pharmacokinetics of warfarin 25 mg or the anticoagulation level. These results suggest that there is no need for dose adjustment of warfarin when coadministered with lacosamide.
The risk for drug–drug interactions among the elderly is of particular clinical concern given the high likelihood of concomitant medication use for comorbid conditions (Levy & Collins, 2007). Treatment of epilepsy in the elderly presents unique challenges in that the process of aging alone can influence drug absorption, distribution, metabolism and elimination, and pharmacokinetic and pharmacodynamic interactions (Gidal, 2006). Furthermore, many cases of both acute seizures and chronic epilepsy in the elderly are associated with stroke and other comorbid conditions for which patients are receiving pharmacologic treatment (Waterhouse & Towne, 2005). Because the incidence of new-onset epilepsy is highest among the elderly and most commonly due to stroke (Bergey, 2004), there is a greater likelihood that elderly patients requiring antiepileptic drug (AED) therapy will also require concomitant treatment with warfarin, a commonly prescribed oral anticoagulant.
Warfarin is used for the prevention of ischemic or recurrent ischemic events in patients with nonvalvular atrial fibrillation and heart valve prosthesis, or for the treatment of thromboembolism (Keeling et al., 2011). Although warfarin is the mainstay of treatment for reducing thromboembolic risk and its use has increased with the aging population (Huhtakangas et al., 2011), its pharmacokinetic and pharmacodynamic profiles, as well as its narrow therapeutic index, make it particularly susceptible to interactions with drugs and food. Clinical use requires monitoring of prothrombin time (PT), and individualized dosage adjustments are expected to maintain safe and effective anticoagulation (Keeling et al., 2011).
Warfarin inhibits vitamin K–dependent coagulation factors by inhibition of the vitamin K epoxide reductase enzyme complex subunit 1 (VKORC1), which is involved in the rate-limiting step of vitamin K recycling (D'Andrea et al., 2008; Johnson et al., 2011; Dean, 2012). The relationship between the dose of warfarin prescribed and the individual response is regulated by genetic and environmental factors that can influence the absorption of warfarin, its pharmacokinetics, and pharmacodynamics (D'Andrea et al., 2008; Johnson et al., 2011; Dean, 2012).
Warfarin is a racemic mixture of the R- and S-enantiomers. The S-enantiomer exhibits about 2–5 times more anticoagulant activity than the R-enantiomer in humans, but generally has a more rapid clearance (D'Andrea et al., 2008; Dean, 2012). S-warfarin is metabolized by hepatic isoenzyme cytochrome P450 (CYP)2C9, and R-warfarin is metabolized by CYP3A4, CYP1A2, and CYP2C19 (D'Andrea et al., 2008), thus introducing the potential for drug interactions. An illustrative example is the potent and selective inhibition of CYP2C9 by ataciguat, resulting in a threefold increase in S-warfarin plasma concentration but no impact on R-warfarin (Oberwittler et al., 2007). Even in the absence of concomitant medications, warfarin dosing is challenging due to individual variation in anticoagulation response caused by differences in age, diet, and genetic factors (D'Andrea et al., 2008). Prominently, polymorphisms in the gene encoding CYP2C9 as well as VKORC1 are associated with variable warfarin dose requirements (Aithal et al., 1999; D'Andrea et al., 2008).
Warfarin is highly bound to plasma protein (approximately 99%), and displacement has sometimes been ascribed as the cause of drug–drug interactions (Yoon et al., 2011), but this phenomenon is generally considered insignificant and confounded by CYP inhibition (Rolan, 1994; Benet & Hoener, 2002). Regardless of the underlying mechanism, drug–drug interactions with warfarin can dramatically alter a patient's anticoagulant response (Yoon et al., 2011). Clinically significant interactions between warfarin and enzyme-inducing AEDs such as carbamazepine and phenytoin or the enzyme inhibitor valproate have been well documented and demonstrate the need for anticoagulation monitoring, and when necessary, dosage adjustments of warfarin (Perucca, 2006; Yoon et al., 2011). Even so, phenytoin and carbamazepine are among the most frequently prescribed AEDs in elderly patients (Pugh et al., 2004; Gidal et al., 2009); therefore, the availability of additional AEDs that lack interactions with warfarin would be potentially beneficial to this patient population.
AEDs that are not enzyme inducers or inhibitors would be expected to have a lower potential for altering the pharmacokinetic or pharmacodynamic parameters of warfarin, allowing for more straightforward use with warfarin. A lack of clinically significant interactions between warfarin and levetiracetam (Ragueneau-Majlessi et al., 2001), oxcarbazepine (Kramer et al., 1992), and eslicarbazepine (Vaz-da-Silva et al., 2010) has been observed in healthy volunteers. Gabapentin and pregabalin have not been reported to interact with coumarin anticoagulants (Bockbrader et al., 2010). Lacosamide has not been shown to induce or inhibit CYP enzymes in preclinical studies or in clinical studies examining specific CYP substrates (UCB, 2011a; Cawello et al., 2012). Lacosamide also exhibits minimal protein binding (<15%; Cawello et al., 2013). The aim of this study was to evaluate potential pharmacokinetic and pharmacodynamic interactions associated with the coadministration of warfarin and lacosamide.
This was a phase I open-label, randomized, two-treatment crossover trial of warfarin (single 25-mg doses) alone and together with lacosamide under steady state (multiple 200-mg b.i.d. doses) conducted over a period of up to 31 days. The trial was conducted at a single site in The Netherlands in accordance with the International Conference on Harmonisation (ICH)-Good Clinical Practice (GCP), the Declaration of Helsinki, and the local laws of The Netherlands. The protocol, amendments, and subject informed consent were reviewed and approved by an Independent Ethics Committee (Stichting Beoordeling Ethiek Biomedisch Onderzoek, Assen, The Netherlands). Written informed consent was obtained from all participants.
Sixteen participants were randomized to receive a single 25-mg dose of warfarin alone in one period and lacosamide 200-mg twice daily on days 1–9, with a single 25-mg dose of warfarin coadministered on day 3 in the other period. There was a 2-week washout between treatments. All participants were confined to the clinic from the evening before the first dose of each treatment period to up to 72 h after the administration of warfarin, at which time an international normalized ratio (INR) ≤2 was required for departure. Safety and tolerability were monitored by clinical laboratory tests (hematology, coagulation, chemistry, and urinalysis), vital signs, and 12-lead electrocardiography (ECG).
All participants were healthy adult male patients (age 18–55 years) with PT, INR, activated partial thromboplastin time (aPTT), and plasma protein C or S activity (functional) all within the normal range. Those with a history of coagulation abnormalities, clinically relevant ECG findings, values above the upper limit of normal for liver function tests, known hypersensitivity to any drug, or intake of any medications within the last 2 weeks were excluded from the study. The participants were genotyped for the presence of common mutations of CYP2C9 and VKORC1.
In both treatment sequences, blood samples for the determination of S- and R-warfarin in plasma were collected prior to administration of the single warfarin dose and 0.5, 1, 2, 4, 8, 12, 24, 36, 48, 72, 96, 120, 144, and 168 h post dose. Blood samples for the determination of lacosamide concentration were collected prior to the morning doses. Samples were centrifuged at 1,600 g for 10 min at 4°C, and the supernatant was frozen at −20°C until analysis by SGS Life Science Services (Saint-Benoit Cedex, France).
R- and S-warfarin plasma concentrations were determined using a validated method involving solid phase extraction and high-performance liquid chromatography tendem mass spectrometry (HPLC/MS/MS) with a chiral column (Chiracel OD-RH; Daicel Corporation, Tokyo, Japan). Warfarin-d5 was used as internal standard. The lower limit of quantification (LLQ) was 0.002 μg/ml for each enantiomer using a 100-μl aliquot of plasma. The imprecision (% coefficient of variation) of quality control samples ranged from 3.5% to 4.5% for R-warfarin and from 3.0% to 3.3% for S-warfarin; inaccuracy (% bias) ranged from −1.5% to +7.8% and −4.3% to +6.1%, respectively. Plasma lacosamide was determined using a validated HPLC/MS/MS method with an LLQ of 0.005 μg/ml using a 50-μl aliquot of plasma; the imprecision of quality control samples was <6.0%, and inaccuracy ranged from −1.0% to +1.2%.
The following pharmacokinetic parameters were calculated for S- and R-warfarin using WinNonlin version 5.2 (Pharsight Corporation, Mountain View, CA, U.S.A.): area under the plasma concentration–time curve from time zero to infinity (AUC(0,∞)); AUC from time zero to the time of the last quantifiable concentration (AUC(0,last)); maximum plasma concentration (Cmax), time to reach Cmax (tmax), plasma half-life (t½), apparent plasma clearance (CL/F), and apparent volume of distribution (Vz/F).
The anticoagulation activity of warfarin in both treatment sequences was determined by monitoring PT in blood samples that were collected prior to administration of the warfarin single doses and 0.5, 1, 2, 4, 8, 12, 24, 36, 48, 60, 72, 96, 120, 144, and 168 h post dose. PT/INR was determined within 120 min of sample collection on an ACL-9000 analyzer using Hemosil RecombiPlasTin 2G for standardization (Instrumentation Laboratory, Breda, The Netherlands).
The pharmacodynamic parameters assessed following each warfarin administration were: PTmax, maximum observed PT time; INRmax, maximum observed international normalized ratio; AUCPT, area under the PT versus time curve computed from time 0 to the last sampling time (168 h); and, AUCINR, area under the INR versus time curve computed from time 0 to the last sampling time (168 h).
Statistical evaluation was performed using SAS version 9.1.3 (Cary, NC, U.S.A.). Safety analyses were performed for all participants in the Safety Set (SS), which included all participants who received at least one dose of study medication. Participants in the SS who completed the study without important protocol deviations affecting pharmacokinetic parameters and for whom reliable estimates for pharmacokinetic parameters could be calculated were included in the pharmacokinetic analysis. The pharmacodynamic analysis included all participants in the SS who completed the study without important protocol deviations affecting pharmacodynamic parameters and for whom reliable estimates for PTmax or INRmax and AUCPT or AUCINR could be calculated.
Following log-transformation, the pharmacokinetic parameters AUC(0,last), AUC(0,∞), and Cmax of S- and R-warfarin were analyzed separately using a mixed-effects model with sequence, period, and treatment as fixed effects and subject within sequence as a random effect. Point estimates and the corresponding 90% confidence intervals (CIs) for the difference between least squares means with and without coadministration of lacosamide were calculated. The point estimate and the 90% CI for the ratio of geometric means were obtained following back-transformation. Lack of treatment interaction was concluded if the 90% CIs of the least squares means ratio for AUC(0,last), AUC(0,∞), Cmax, AUCPT, PTmax, AUCINR, and INRmax fell within the acceptance range of 0.80–1.25. A similar linear mixed effects model was also fitted to each of the pharmacodynamic parameters, AUCPT, PTmax, AUCINR, and INRmax.
Under the assumption of a within-subject coefficient of variation (CV) of 10% and a true ratio of 1.0, a sample size of 12 participants was estimated to provide >90% power for the ratios of pharmacokinetic and pharmacodynamics parameters to be within the range of 0.80–1.25 (Malhotra et al., 2011). Sixteen participants were enrolled in order to account for drop outs.
All participants (N = 16) were healthy men with a mean age of 32.8 years, average weight of 78.5 kg, and body mass index of 23.7 kg/m2. Demographic and baseline characteristics were similar among volunteers in the two treatment sequences. All participants completed the study and were included in the pharmacokinetic and pharmacodynamic analyses.
For S- and R-warfarin, similar mean plasma concentration–time profiles were observed following treatment with warfarin alone versus coadministration of lacosamide and warfarin (Fig. 1). Mean AUC(0,∞) and AUC(0,last) of S-warfarin decreased slightly (mean ratios of 0.969 and 0.971, respectively) following coadministration of lacosamide, as did Cmax (mean ratio of 0.984; Table 1). Mean AUC(0,∞) and AUC(0,last) for R-warfarin increased slightly (mean ratios of 1.052 and 1.034, respectively) following coadministration of lacosamide, whereas Cmax decreased slightly (mean ratio of 0.975; Table 1). Statistical comparison of lacosamide and warfarin coadministration versus warfarin alone for AUC(0,∞), AUC(0,last), and Cmax resulted in 90% CIs around the mean ratios being entirely contained within the acceptance range of 0.8–1.25 for all parameters for both S- and R-warfarin (Table 1). Within-subject variability, as assessed by the analysis of variance (ANOVA) residual coefficient of variation, was low, not exceeding around 5% for AUC(0,∞) of both warfarin enantiomers and 8.9% and 11.9% for Cmax of R- and S-warfarin, respectively (Table 1). Coadministration of lacosamide did not affect other pharmacokinetic parameters assessed for S- and R-warfarin (Table 1). In addition, lacosamide mean (standard deviation [SD]) trough plasma levels ranged from 5.6 (0.9) μg/ml on the third morning of lacosamide intake to 6.9 (1.3) μg/ml on the 10th morning, indicating that participants were adequately exposed to lacosamide (Cawello & Bonn, 2012) during the period of warfarin pharmacokinetic and pharmacodynamic measurements. No participant exhibited a CYP2C9 poor metabolizer genotype.
|Parameter||Warfarin + lacosamide n = 16||Warfarin alone n = 16||Treatment comparison warfarin + lacosamide/warfarin alone|
|Point estimate (ratio)||90% Confidence interval||ANOVA CV%a|
|AUC(0,∞) (μg·h/ml)||76.16 (19.8)||72.38 (18.1)||1.052||1.017, 1.088||5.4|
|AUC(0,last) (μg·h/ml)||67.78 (17.2)||65.53 (15.8)||1.034||1.006, 1.063||4.4|
|Cmax (μg/ml)||1.184 (13.8)||1.214 (12.7)||0.975||0.922, 1.031||8.9|
|tmax (h)||1.5 (0.5–4)||2.0 (0.5–4)|
|t1/2 (h)||53 (18)||49 (13)|
|CL/F (ml/min/kg)||0.035 (19.1)||0.037 (17.5)|
|Vz/F (L/kg)||0.16 (17.4)||0.16 (11.8)|
|AUC(0,∞) (μg·h/ml)||45.73 (35.0)||47.19 (33.7)||0.969||0.938, 1.001||5.2|
|AUC(0,last) (μg·h/ml)||43.14 (30.5)||44.43 (29.5)||0.971||0.943, 1.000||4.7|
|Cmax (μg/ml)||1.204 (16.7)||1.224 (15.2)||0.984||0.914, 1.059||11.9|
|tmax (h)||1.5 (0.5–4)||2.0 (0.5–4)|
|t1/2 (h)||39.39 (30.9)||40.67 (26.3)|
|CL/F (ml/min/kg)||0.058 (36.8)||0.056 (35.2)|
|Vz/F (L/kg)||0.20 (19.4)||0.20 (22.1)|
|AUCINR (h)||235.2 (15.3)||226.9 (17.7)||1.037||1.012, 1.062||3.9|
|INRmax||2.031 (22.8)||1.968 (27.2)||1.032||0.993, 1.073||6.2|
|AUCPT (s·h)||2,624 (15.1)||2,532 (17.5)||1.036||1.011, 1.061||3.9|
|PTmax (s)||22.58 (22.6)||21.88 (26.9)||1.032||0.994, 1.072||6.1|
Mean values of pharmacodynamic parameters (AUCPT, PTmax, AUCINR, and INRmax) were slightly higher after coadministration of lacosamide and warfarin compared with warfarin alone (mean ratios ranging from 1.032 to 1.037; Table 1). Mean INR values over time were similar for warfarin alone and warfarin coadministered with lacosamide (Fig. 2). Statistical comparison of lacosamide with warfarin versus warfarin alone for PT and INR resulted in treatment mean ratios of approximately 1.0 and corresponding 90% CIs being entirely contained within the limits of 0.8–1.25 for all pharmacodynamic parameters (Table 1). The residual coefficient of variation was 3.9% for AUCPT and AUCINR, and 6.2% for PTmax and INRmax (Table 1). A subgroup analysis without VKORC1 mutation carriers did not change the outcome.
Warfarin is a frequently prescribed oral anticoagulant; however, its use is complicated by a narrow therapeutic range, sensitivity to genetic polymorphisms, susceptibility to metabolic changes with increasing age, and propensity for drug–drug interactions with enzyme-inducing or inhibiting drugs (D'Andrea et al., 2008). The availability of AEDs suitable for use in patients receiving warfarin is particularly relevant when considering that enzyme-inducing AEDs (e.g., phenytoin and carbamazepine) are among the most frequently prescribed AEDs in elderly patients (Pugh et al., 2004; Gidal et al., 2009).
In the present study, no effects on the pharmacokinetic or pharmacodynamic profiles of warfarin were detected when warfarin was added to the highest recommended dose of lacosamide (400 mg/day). Lacosamide plasma trough concentrations were similar to those previously reported at the same daily dose (Cawello & Bonn, 2012), and the anticoagulation level reached after the warfarin 25-mg single dose was similar to that reported in other studies using warfarin single dose or titration (Ragueneau-Majlessi et al., 2001; Oberwittler et al., 2007; Vaz-da-Silva et al., 2010; Frey et al., 2011; Malhotra et al., 2011; Yin et al., 2011). The residual variability of the pharmacokinetic and pharmacodynamic parameters was very low, resulting in narrow CIs around the treatment mean ratios (Table 1). Maximal induction of hepatic enzymes generally takes several weeks of repeated dosing of the inducer, although the onset of induction may be evident much sooner when the half-life of the inducer is shorter than the enzyme turnover (Brodie et al., 2013). Warfarin was administered on the third day of lacosamide intake; the AUCs of R- and S-warfarin were essentially unchanged relatively to the warfarin alone period and even 3% lower for S-warfarin, suggesting no evidence of enzyme induction. Given that lacosamide is not known to be an enzyme-inducing or inhibiting AED (Bialer et al., 2007; Cawello et al., 2010, 2012; UCB, 2011a,b; Cawello & Bonn, 2012), the results summarized here are consistent with its known properties.
Because many patients with epilepsy may attempt several AEDs before achieving seizure control and/or may require multiple AEDs (Kwan & Brodie, 2006), the availability of additional treatment options that do not impact warfarin response is important to the overall effective management of epilepsy, particularly for an elderly population. Demographic data indicate that new-onset epilepsy is more prevalent in the elderly than in any other age group (Bergey, 2004; Bergey et al., 2006), suggesting that the coadministration of warfarin and the AED lacosamide will likely occur in elderly patients in whom other comorbidities, polypharmacy, and age-related metabolic changes can be expected (Patsalos & Perucca, 2003; Stephen, 2003; Gidal, 2006). Although participants in this study were healthy male volunteers, none of whom were elderly or taking additional medications, the results suggest that lacosamide/warfarin coadministration would not require warfarin dose adjustment in clinical practice.
Because patients generally receive long-term warfarin therapy for thromboprophylaxis, a potential limitation of this study is the use of steady-state conditions for lacosamide but not for warfarin. However, the design of this study was typical of warfarin drug–drug interaction studies and was not intended to address long-term treatment effects of concomitant administration of lacosamide and warfarin, but to provide a sensitive measure of a potential interaction. This design was deemed appropriate because the pharmacokinetics of warfarin is linear and allows for extrapolation while being more sensitive to potential pharmacokinetic changes than steady-state conditions, as pointed out previously (Oberwittler et al., 2007; Frey et al., 2011). The single dose of warfarin evaluated in this study was sufficient to increase INR from 1 to 2. In conclusion, this study did not reveal any significant interaction between lacosamide and warfarin, indicating that the pharmacokinetic profile and anticoagulant properties of warfarin will not be affected when coadministered with lacosamide. Administration of a single 25-mg dose of warfarin during maintenance treatment with lacosamide was well tolerated. These results suggest that warfarin dose adjustment is not required during concomitant administration with lacosamide.
This study was funded by UCB Pharma. Merrilee Johnstone, Ph.D., Jennifer Hepker, Ph.D., and Kristen Andersen, Ph.D. from Prescott Medical Communications Group (Chicago, IL, U.S.A.) provided writing assistance. Azita Tofighy, Ph.D., employee of UCB, coordinated the publication process.
Armel Stockis, Willi Cawello, Thomas Kumke, and Klaus Eckhardt are employees of UCB Pharma. Jan Jaap van Lier is an employee of PRA-EDS, which is the CRO that conducted the study with payment from UCB; the outcome of the study did not influence the salary of Jan Jaap van Lier. The authors confirm they have read the Journal's position on issues involved in ethical publication and affirm this report is consistent with those guidelines.