Apolipoprotein E (APOE) and warfarin dosing in an Italian population

Abstract

Sirs, Despite large interindividual differences in dose requirement and a narrow therapeutic range, warfarin is still the most widely prescribed anticoagulant for thromboembolic therapy. However, the problems with dose titration, especially during the initial phase of treatment, lead to a high risk of either severe bleeding or the failure to prevent thromboembolism [1]. A working knowledge of warfarin interactive pathways and of the human genome will facilitate the determination of genetic factors that cause variable response to this drug. Warfarin acts through the inhibition of vitamin K reductase in the vitamin K cycle, thereby preventing the activation of clotting factors II, VII, IX and X and proteins C, S and Z [2, 3]. However, vitamin K1 can be reduced through another pathway, and a high intake of vitamin K1 may reverse the anticoagulant effect of warfarin [3–5]. Vitamin K1 is absorbed from the small intestine together with dietary fat, transported by chylomicrons in the blood and subsequently cleared by the liver through an apolipoprotein E (APOE) receptorspecific uptake [3, 6]. The major isoforms of apolipoprotein E are encoded by the three APOE alleles E2, E3 and E4 [7]. Individuals carrying the APOE*E4 allele have a more rapid uptake of chylomicron remnants following a meal than those that do not [8], and they also have lower serum concentrations of vitamin K1 following an overnight fast [6, 9, 10]. These findings suggest that carriers of APOE*E4 have a more rapid uptake of vitamin K1, leading to an enhanced availability of reduced vitamin K for the activation of clotting factors and proteins in the liver. Warfarin is administered as a racemate of Rand Senantiomers [11]. Cytochrome P450 2C9 (CYP2C9) is the main enzyme responsible for the metabolism of the more active S-warfarin [12], and genetic variability in CYP2C9 has a significant effect on warfarin dose requirement [13]. We have recently published a study on warfarin-treated Swedish patients showing that CYP2C9*1/*1 extensive metabolisers that are homozygous for APOE*E4 require higher warfarin doses than patients carrying other APOE alleles [14]. This study was repeated using 145 Italian patients, most of whom had participated in a previously published study on warfarin pharmacokinetics [15]. The local Ethics Committee (Comitato Etico dell’Azienda Ospedaliera di Padova) gave permission for the APOE genotyping. The warfarin patients had stable prothrombin time international normalised ratios (PT INR), and no interfering drugs were allowed during the study. Two APOE single nucleotide polymorphisms (SNPs) at nucleotide positions 334 and 472 were investigated using TaqMan (Applied Biosystems, Foster City, Calif.), and the ABI PRISM 7000 Sequence Detection System (Applied Biosystems) [14]. Twenty patients carried one APOE*E4 allele (E2/E4, n=1 and E3/E4, n=19), 105 were homozygous for E3 and 20 were heterozygous E2/E3. No one was homozygous for APOE*E4. There were no significant differences in age, gender, bodyweight or PT INR among the APOE genotype groups. We found that the weekly warfarin dose requirement did not vary between patients with different APOE genotypes (analysis of variance: p=0.5937; Fig. 1a). To test the contribution of other important factors, we combined the POE genotype with the CYP2C9 genotype, age, bodyweight and PT INR in a multiple regression model and found that CYP2C9 H. Kohnke AE M. G. Scordo AE M. Wadelius (&) Department of Medical Sciences, Clinical Pharmacology, University Hospital, Uppsala, Sweden E-mail: mia.wadelius@medsci.uu.se Tel.: +46-18-6114945