Human genes discussed in the paper: CYP2C9, VKORC1.
Letters to the Editor
Genetic resistance to warfarin therapy masked by amiodarone in a 2-year-old girl with mitral valve replacement
Article first published online: 13 MAR 2013
© 2013 International Society on Thrombosis and Haemostasis
Journal of Thrombosis and Haemostasis
Volume 11, Issue 3, pages 555–572, March 2013
How to Cite
Genetic resistance to warfarin therapy masked by amiodarone in a 2-year-old girl with mitral valve replacement. J Thromb Haemost 2013; 11: 555–7., , , , .
- Issue published online: 13 MAR 2013
- Article first published online: 13 MAR 2013
- Accepted manuscript online: 15 DEC 2012 07:42AM EST
- Manuscript Accepted: 3 DEC 2012
- Manuscript Received: 21 AUG 2012
A 2.2-year-old Moroccan girl (11 kg; body surface area, 0.50 m2) underwent mitral valve replacement for congenital mitral stenosis. Two days later, she received amiodarone for atrial arrhythmia; the dose was 500 mg m−2 daily for 7 days, and then 250 mg m−2 daily (Fig. 1). Nine days after surgery, she was started on a standard warfarin regimen for children (0.2 mg kg−1) with an initial dose of 2 mg (day 1) . The target International Normalized Ratio (INR) was 3.3 (range, 2.5–4). The INR increase was slow over the first week of warfarin treatment, despite dose escalation from 2 mg on day 1 to 5 mg on day 6 (Fig. 1). After 20 days of warfarin treatment (day 20), keeping the INR within the target range required the high dose of 25 mg weekly (Fig. 1) . On day 112, amiodarone was withdrawn. The INR 7 days later was only 1.4 (Fig. 1). Because of infratherapeutic INRs, low molecular weight heparin was given (enoxaparin 1200 IU twice daily), and the warfarin dose was increased in increments of 1 mg daily with daily INR monitoring. Finally, a weekly warfarin dose of 51 mg was needed to achieve the target INR on day 137, i.e. two times the dose required when the patient received amiodarone. On day 142, a serum warfarin assay (HPLC with UV detection and diode array) showed a high level of 4.8 mg L−1 (reference range, 1.8–2.6 mg L−1), indicating good treatment adherence.
To investigate this partial pharmacodynamic resistance to warfarin , we analyzed the gene encoding vitamin K epoxide reductase complex subunit 1 (VKORC1), which is the pharmacologic target of vitamin K antagonists (VKAs). Exonic regions, intron–exon boundaries and fragments covering 2000 bp of the 5′-flanking VKORC1 region were sequenced as previously described [4, 5]. VKORC1 sequencing showed heterozygosity for c.106G>T (rs61742245), which is responsible for p.Asp36Tyr; in addition, g.−1185G>A and g.−679A>G were identified in the 5′-flanking region. The common VKORC1 (–1639 G/A) single-nucleotide polymorphism (SNP) was wild-type G/G; it is of note that the A allele is associated with decreased transcriptional activity of the VKORC1 promoter, which leads to a decrease in warfarin dosage requirements. In addition, using allelic discrimination, we genotyped the cytochrome P450 (CYP) 2C9 gene (CYP2C9) encoding the enzyme primarily responsible for the metabolism of the more active S-enantiomer of warfarin into inactive hydroxy compounds. The patient had no CYP2C9*2 or CYP2C9*3 variant alleles.
VKA therapy is challenging to manage, because of both a narrow therapeutic margin and considerable interindividual and intraindividual response variability. Demographic variables (age and body mass index), comorbidities, environmental factors, including interacting drugs and diet, and common genetic variations (VKORC1 [−1639G>A], CYP2C9*2 and CYP2C9*3) explain approximately 30–60% of warfarin response variability in adults . In children, this variability has been investigated in only a few studies, which have shown that height, target INR and VKORC1/CYP2C9 genotypes explain ~ 70% of the variability .
In clinical practice, a very small number of patients need unusually high VKA doses to achieve therapeutic anticoagulation. In some cases, despite enormous doses, complete resistance is observed, with failure to achieve therapeutic INR values. These excessive VKA dose requirements cannot be explained entirely by non-genetic factors and common VKORC1/CYP2C9 SNPs. Patients requiring such excessive doses should benefit from detailed VKORC1 genetic analysis ; rare VKORC1 mutations associated with VKA resistance have been reported in adults, but not in pediatric patients. According to our previously published dose prediction algorithm incorporating height, target INR, and classic VKORC1/CYP2C9 genotypes , the expected dose for our 2.2-year-old patient was calculated to be 25 mg weekly, i.e. two-fold lower than the observed dose (51 mg) after discontinuation of amiodarone.
Here, we report the first case of warfarin genetic resistance in a child. We identified three VKORC1 mutations: c.106G>T, leading to p.Asp36Tyr; and, in the 5′-flanking region, g.−1185G>A and g.−679A>G. The p.Asp36Tyr mutation has been reported in adults as a factor predisposing to warfarin resistance and a marker for high dose requirements that overrides the dose-reducing effect of CYP2C9*2, CYP2C9*3 and VKORC1 −1639A alleles [3, 4, 7]. Adults heterozygous for the p.Asp36Tyr mutation have been reported to require a mean weekly warfarin dose of ~ 70 mg in the largest published cohort, i.e. an approximately 2-fold dose increase, in the same order of magnitude as that observed in our child . The p.Asp36Tyr mutation is a common cause of VKA resistance in Jews of Ethiopian origin (allele frequency, 15%) and, to a lesser extent, in Ashkenazi Jews (4%), Sephardic Jews (0.6%) and Israelis . Few cases have been reported in Caucasians belonging to other ethnic groups or in African-Americans . The p.Asp36Tyr mutation affects the large loop located in the endoplasmic reticulum lumen between the first two transmembrane VKORC1 helices. Importantly, this loop is predicted to contain amino acids that are presumed to be located in the catalytic site . Furthermore, our patient, who is a Moroccan, carried two other mutations, g.−1185G>A and g.−679A>G, located in the VKORC1 5′-flanking region. In several other VKA-resistant patients referred to us and found to carry the p.Asp36Tyr mutation, we identified the same 5′-flanking region mutations, suggesting linkage disequilibrium between the g.−1185G>A and g.−679A>G mutations and the p.Asp36Tyr mutation ([3, 4] and personal data).
The genetic resistance to warfarin in our patient was initially partly masked by amiodarone therapy (125 mg daily). Amiodarone discontinuation led to major under-anticoagulation (INR, 1.4). Doubling of the warfarin dose (from 25 to 51 mg weekly) was needed to achieve stable anticoagulation. As amiodarone has a very long half-life (20–100 days), close INR monitoring is required several weeks after the interruption of the treatment. Amiodarone induces a broad range of adverse events, and is involved in complex drug interactions, especially with VKAs . Amiodarone strongly inhibits several CYPs, including CYP2C9, CYP1A2, and CYP3A4, thereby decreasing warfarin clearance and potentiating warfarin pharmacodynamic effects, resulting in increased anticoagulation . The warfarin–amiodarone interaction is considered to be clinically relevant, as it can lead to over-anticoagulation when amiodarone is added to warfarin, thus potentially increasing the risk of hemorrhagic complications [10, 11]. In adults receiving standard warfarin doses, a decrease of ~ 25% in the dose is required when amiodarone 100 mg dailyis started . Conversely, amiodarone withdrawal can lead to under-anticoagulation, and therefore to an increased risk of thrombosis.
In conclusion, we describe the first case of genetic VKA resistance in a child who had the p.Asp36Tyr mutation combined with g.−1185G>A and g.−679A>G mutations in the 5′-flanking region. VKA resistance was partially masked initially by concomitant amiodarone therapy. Both warfarin–drug interactions and genetic variations should be considered in the management of VKA therapy. Physicians should be aware that potentially interfering drugs can increase or decrease the INR when they are introduced or withdrawn in patients on warfarin therapy. Close INR monitoring is advisable in such situations. Pharmacodynamic VKA resistance resulting from genetic factors may be masked by concomitant drugs.
Disclosure of Conflict of Interests
The authors state that they have no conflict of interest.