ABSTRACT
Title
Prospective validation of a genetic-based algorithm for warfarin dosing. An ad interim analysis
Authors
F. Groppa1, A. Rossetto1, V. Pengo3, C.F. Zambon2, G. De Rosa1, M. Pelloso2, P Fogar2, M. Gusella4, M. Plebani2, R. Padrini1.
1Dept. of Clinical and Experimental Medicine, University of Padova.
2Dept. of Laboratory Medicine, University Hospital of Padova.
3Dept. Cardiothoracic and Vascular Sciences, University of Padova.
4Lab. of Pharmacology and Molecular Biology, S. Luca Hospital, Trecenta, Rovigo.
1Dept. of Clinical and Experimental Medicine, University of Padova.
2Dept. of Laboratory Medicine, University Hospital of Padova.
3Dept. Cardiothoracic and Vascular Sciences, University of Padova.
4Lab. of Pharmacology and Molecular Biology, S. Luca Hospital, Trecenta, Rovigo.
Abstract
Warfarin, the most widely used oral anticoagulant, is administered as racemic mixture of S- and R-warfarin. The S-enanthiomer is much more potent than the R-enanthiomer and is metabolized by the polymorphic cytochrome CYP2C9 (slow variants: CYP2C9*2 and *3). Two other gene polymorphisms are associated with warfarin response: VKORC1-1639G>A (which modulates the expression vit.K-epoxide-reductase, molecular target of the drug), and CYP4F2*3 (which inactivates vitamin K in the liver). Our group has recently published an algorithm based on CYP2C9, VKORC1,CYP4F2 polymorphisms, body surface area and age to identify the individual warfarin maintenance dose (Zambon et al., 2011). In addition, the loading dose can be calculated by means of the VKORC1 genotype and body weight, using the PK-PD model proposed by Hamberg et al. (2010). The aim of the study was to compare the performance of this genetic method (GM) with that of the empirical method (EM) currently used in the Anticoagulation Clinic of the University Hospital of Padova (Pengo et al. 2001), which consists in administering a 5mg warfarin dose for 4 days followed by dose adjustment based on the International Normalized Ratio (INR) value. Patients who had to start warfarin treatment were randomized to follow the GM or the EM. Demographic data and blood samples for genotyping were collected in both experimental groups at the enrollment visit. Genetic analyses were performed within 2 days in the “genetic group” in order to calculate the individual loading and maintenance doses. All patients were seen 4 days after starting the therapy (day 1) and, next, after 3, 5, 8, 11, and 15 days. On each visit, the following data were measured or recorded: plasma concentrations of S- and R-warfarin (HPLC), INR and prescribed doses. At the moment of writing, 79 patients (of the 182 planned) completed the protocol and 8 dropped-out. The primary endpoint was the percentage of INR <2 or >3 in the 15-day observation period. The results were analyzed after classifying all patients in three groups according to their intrinsic sensitivity to warfarin, evaluated by means of the “genetic” maintenance dose (MD): high sensitivity group (HS) (MD < 25 mg/week); normal sensitivity group (NS) (MD 25-35 mg/week); low sensitivity group (LS) (MD > 35 mg/week).
The frequencies of INR values out of range (2-3) were as follows: a) HS group: GM 36.7%; EM 43.1% (n.s.); b) NS group: GM 47.4%; EM 38.5% (n.s.); c) LS group: GM 36.5%; EM 55.6% (p=0.0185). When we analyzed separately the frequency of INR <2 and >3, the results were: INR <2: a) HS group: GM 30.0%; EM 20.8% (n.s.); b) NS group: GM 31.7%; EM 32.3% (n.s.); c) LS group: GM 29.2%; EM 48.6% (p=0.0154); INR>3: a) HS group: GM 8.3%; EM 22.2% (p<.0335); b) NS group: GM 7.7%; EM 6.2% (n.s.); c) LS group: GM 7.3%; EM 6.9% (n.s). Collectively, these results indicate that the GM reduced the number out-of-range fluctuations of INR as compared with the EM, but only in the HS group (less INRs >3) and the LS group (less INRs <2). No therapeutic advantage emerged in the NS group. In addition, while in the “genetic group” the average daily dose was virtually the same at day 1 (HS: 3.1 ±0.7 mg; LS: 6.0 ±1.4 mg) and at day 15 (LS: 3.3 ±1.3 mg; HS: 6.2 ±2.8 mg) of treatment, in the “empirical group” the initial 5 mg dose had to decreased by -46% in the HS and increased by +24% in the LS group, on average. These preliminary results suggest that the use of a genetic-based algorithm to assess the loading and maintenance warfarin dose can reduce the risk of over-coagulation in the most responsive patients and of under-coagulation in the least responsive patients.
Zambon et al. (2011). Pharmacogenomics 12:15-25.
Hamberg et al. (2010). Clin Pharmacol Ther 87:727-34.
Pengo et al. (2001). Am J Cardiol 88: 1214-1216.
The frequencies of INR values out of range (2-3) were as follows: a) HS group: GM 36.7%; EM 43.1% (n.s.); b) NS group: GM 47.4%; EM 38.5% (n.s.); c) LS group: GM 36.5%; EM 55.6% (p=0.0185). When we analyzed separately the frequency of INR <2 and >3, the results were: INR <2: a) HS group: GM 30.0%; EM 20.8% (n.s.); b) NS group: GM 31.7%; EM 32.3% (n.s.); c) LS group: GM 29.2%; EM 48.6% (p=0.0154); INR>3: a) HS group: GM 8.3%; EM 22.2% (p<.0335); b) NS group: GM 7.7%; EM 6.2% (n.s.); c) LS group: GM 7.3%; EM 6.9% (n.s). Collectively, these results indicate that the GM reduced the number out-of-range fluctuations of INR as compared with the EM, but only in the HS group (less INRs >3) and the LS group (less INRs <2). No therapeutic advantage emerged in the NS group. In addition, while in the “genetic group” the average daily dose was virtually the same at day 1 (HS: 3.1 ±0.7 mg; LS: 6.0 ±1.4 mg) and at day 15 (LS: 3.3 ±1.3 mg; HS: 6.2 ±2.8 mg) of treatment, in the “empirical group” the initial 5 mg dose had to decreased by -46% in the HS and increased by +24% in the LS group, on average. These preliminary results suggest that the use of a genetic-based algorithm to assess the loading and maintenance warfarin dose can reduce the risk of over-coagulation in the most responsive patients and of under-coagulation in the least responsive patients.
Zambon et al. (2011). Pharmacogenomics 12:15-25.
Hamberg et al. (2010). Clin Pharmacol Ther 87:727-34.
Pengo et al. (2001). Am J Cardiol 88: 1214-1216.