Coumarin derivatives, such as warfarin, phenprocoumon and acenocoumarol, are used for long-term prevention of thromboembolic events. The management of oral anticoagulation with coumarin derivatives is complicated by a large variability in the dose-response relationship, which is partly determined by genetic constitution (Rost et al, 2004; Bodin et al, 2005; Sconce et al, 2005). Coumarins act by inhibiting the vitamin K epoxide reductase (VKOR), encoded for by the VKORC1 (VKOR complex, subunit 1) gene. This enzyme recycles vitamin K epoxide to the reduced form of vitamin K, an essential cofactor in the formation of the active clotting factors II, VII, IX, and X and the inhibitors protein C and S through γ-glutamyl carboxylation. While the most common VKORC1 genetic variants result in the need for lower doses of warfarin during long-term therapy (Rieder et al, 2005; Sconce et al, 2005), some genetic variants confer coumarin resistance (Bodin et al, 2005). International Normalized Ratio (INR) values in combination with a plasma concentration of the coumarin in use give a good indication of possible resistance (Harrington et al, 2008). The known VKORC1 sequence variants associated with coumarin resistance were recently summarised (Peoc’h et al, 2009).
We report three patients presented with confirmed (partial) coumarin resistance. Patient 1 was initially treated with acenocoumarol 12 mg/day, and subsequently with phenprocoumon, up to 9 mg/day. At this dose the phenprocoumon serum concentration, determined by non-stereospecific reversed phase high performance liquid chromatography and diode array detector detection, was 4·3 mg/l (therapeutic range 1–3 mg/l). The INR did not rise above 1·4. Patient 2 was treated with up to 9 mg phenprocoumon, which resulted in a serum phenprocoumon concentration of 7·6 mg/l, while INRs remained below 2·0. Patient 3 was initially treated with acenocoumarol 8 mg/day, subsequently with phenprocoumon 9 mg/day. Serum phenprocoumon concentration was 6·6 mg/l, while the INR was 1·3.
To investigate whether a genetic predisposition of coumarin resistance was present in these three patients the VKORC1 5′ UTR and coding sequence were analysed [AY587020 (Rieder et al, 2005) annotates the wild type VKORC1 genomic sequence]. Both Patients 1 and 3 were heterozygous for a previously described nucleotide variation g.1310T>C (= g.6621T>C in AY587020) in exon 2, leading to p.Trp59Arg (Wilms et al, 2008). Patient 2 was heterozygous for a new nucleotide variation, also a missense mutation g.155C>T in exon 1 leading to p.Ser52Leu. None of the other previously reported sequence variations associated with coumarin resistance were detected in the three patients. In addition, no other genetic alterations were found in the 5′UTR (positions g.−226–1 analysed), in exon 1 (positions g.1–173), in exon 2 (positions g.1309–1418) and exon 3 (positions g.3388–3596). Both Patients 1 and 2 were heterozygous for g.1173C>T, while Patient 3 carried wild type. Individuals carrying g.1173T allele in general require less phenprocoumon or acenocoumarol than individuals carrying g.1173C alleles (Bodin et al, 2005; Rieder et al, 2005; Sconce et al, 2005). Thus, the putative increased sensitivity to anticoagulants due to the presence of VKORC1*2 in Patients 1 and 2 was counteracted by other factor(s) that resulted in decreased sensitivity instead.
Genotyping for Cytochrome P450 2C9 variants showed the absence of CYP2C9*2 and CYP2C9*3 allelic variants in Patients 1 and 2. Patient 3 was heterozygous for CYP2C9*3. With wild type CYP2C9 activity, as in Patients 1 and 2, the elimination half-life of coumarin derivatives are expected to be in the normal range (Sconce et al, 2005; Schalekamp et al, 2006, 2007). Patient 3 was an intermediate metabolizer, resulting in a longer elimination half-life of the coumarin derivative. The three patients had no CYP2C9 inducing co-medication. With VKORC1 resistance, CYP2C9 had lost its contribution to anticoagulation.
If the two single nucleotide polymorphisms described were the cause of coumarin resistance, these polymorphisms do not occur in patients who normally respond to coumarins. We therefore analysed 100 anonymous control samples for the presence of g.155C>T and g.1310T>C. The samples were obtained from routine INR tests of patients receiving <6 mg acenocoumarol or <2 mg phenprocoumon per day with an INR between 2·0 and 4·0.
To detect g.155C>T a TaqMan assay was developed with forward primer 5′-GCG CTC TGC GAC GTG-3′ and reverse primer 5′-GTG CAC ACC TGG AGG AGA A-3′, and MGB probes VIC-AGC TGT TCG CGC GTC-MGB-BHQ and FAM-CAG CTG TTT GCG CGT C-MGB-BHQ to detect wild type and mutant allele, respectively. Due to the high CG content around g.1310T>C we were unable to develop a TaqMan assay for this nucleotide variation. We therefore developed an allele-specific polymerase chain reaction (PCR) to detect g.1310T>C. Primers to detect g.1310T>C were forward 5′-CCC ACC CCT CTG CCA GGC-3′, and reverse 5′-GGA GCC ACT CAC CTA ACA ATA GC-3′, yielding a product of 138 bp in the presence of g.1310T>C and no product in the absence of g.1310T>C. To monitor DNA content and inhibition of PCR, HGH was detected in the same reaction with forward primer 5′-GCC TTC CCA ACC ATT CCC TTA-3′, and reverse primer 5′-TCA CGG ATT TCT GTT GTG TTT C-3′, yielding a PCR product of 429 bp. Both g.155C>T and g.1310T>C were not detected in 100 individuals who responded normally to coumarin-derivates.
Although the data presented here are limited and the functionality of the p.Ser52Leu and p.Trp59Arg proteins was not assessed, we consider it likely that both sequence variations were responsible for the observed coumarin resistance. In support of this notion is the presence of p.Trp59Arg in three, seemingly unrelated phenprocoumon-resistant patients and the absence of both sequence variations in 100 patients who responded to coumarin treatment as expected. In addition, both the serine at position 52 and the tryptophan at position 59 of VKOR are conserved among the species human, rat, mouse, puffer fish and frog (Fig 1). This suggests that p.Trp59 and p.Ser52 are likely to be important for enzyme function.