Predicting the warfarin maintenance dose in elderly inpatients at treatment initiation: accuracy of dosing algorithms incorporating or not VKORC1/CYP2C9 genotypes

Optimal Strategies to Select Warfarin Dose for Thai Patients with Atrial Fibrillation

Background: Warfarin has been the mainstay treatment for the prevention of stroke and systemic thromboembolism in patients with atrial fibrillation (AF). The optimal starting dose of warfarin remains unclear. Objective: To investigate the most optimal dosing strategies for warfarin starting dose in Thai patients with AF. Material and Methods: We enrolled consecutive AF patients who were starting on warfarin and resulting in a stable INR of 2.0-3.0 at two consecutive time points. We measured the dose of warfarin at which INR achieved the target range. The optimal dosage was defined as the difference from the actual dose within 20%. We compared strategies of warfarin dosing, including warfarin dosing formula, 2.5 mg, 3 mg and 5 mg doses. The primary endpoints were the proportions of patients in optimal, underdosing, and overdosing categories. Results: Among 1207 patients visiting the Outpatient Clinic between October 2011 and September 2021, 531 patients were identified with AF and INR in the therapeutic range of 2.0-3.0 on at least two consecutive visits. The mean age of participants was 68 ± 11 years, and men accounted for 44.4% of the population. The warfarin dosing formula resulted in optimal dosing in 37% and overdosing in 24% of cases, whereas the 2.5 mg, 3 mg and 5 mg doses resulted in optimal dosing in 36%, 39%, and 11%, and overdosing in 33%, 44% and 88% of patients, respectively (p < 0.01). Conclusions: In Thai patients with AF, the optimal warfarin starting dose may be 2.5 mg, 3 mg or a simplified warfarin dosing formula, whereas the 5 mg dose should be avoided due to the high risk of overdosing.

Genetic polymorphisms and dosing of vitamin K antagonist in Indian patients after heart valve surgery

Purpose

Vitamin K antagonists (VKAs), such as warfarin and acenocoumarol, exert their anti-coagulant effect by inhibiting the subunit 1 of vitamin K epoxide reductase complex (VKORC1). CYP2C9 is a hepatic drug-metabolizing enzyme in the CYP450 superfamily and is the primary metabolizing enzyme of warfarin. Three single nucleotide polymorphisms, two in the CYP2C9 gene, namely CYP2C9*2 and CYP2C9*3, and one in the VKORC1 gene (c.- 1639G > A, rs9923231), have been identified to reduce VKA metabolism and enhance their anti-coagulation effect. The purpose of this study is to evaluate the prevalence of CYP2C9 and VKORC1 polymorphism in Indians receiving VKA-based anti-coagulation after valve surgery and to evaluate the usefulness of genetic information in managing VKA-based anti-coagulation.

Methods

In the current prospective observational study, 150 patients who underwent heart valve surgery and had stable INR were genotyped for VKORC1 (- 1639 G > A), CYP2C9*2, and CYP2C9*3. The VKA dosage was estimated from published algorithms and compared to the clinically stabilized dosage.

Results

Out of 150 patients, 101 (67.33%) were on warfarin and 49 (32.66%) were on acenocoumarol. Majority of the patients, the 83 in warfarin group and the 40 in acenocoumarol group, had a wild CYP2C9 diplotype. The rest had a mutant (CYP2C9*2 or CYP2C9*3) diplotype. Similarly, 67 patients in the warfarin group and 35 patients in the acenocoumarol group had wild type (G/G) of VKORC1 genotype. The rest had a mutant (G/A or A/A) VKORC1 genotype. In the warfarin group, based on the genotype, 51.5% of the patients were extensive or normal metabolizers, and 47.4% of the patients were intermediate metabolizers of VKAs. In the acenocoumarol group, 61.2% of the patients were extensive or normal metabolizers, and 38.8% of the patients were intermediate metabolizers. Individually, alleles of VKORC1 (- 1639 G > A), CYP2C9*2, and CYP2C9*3 had mean dosage reduction effect on VKA dosage, which co-related to the clinically stabilized dosages (P < 0.0001). Among the VKORC1 (- 1639 G > A) cohort, the reduction in warfarin mean weekly dosage was 13.48 mg as compared to the wild-type category (P < 0.0001) and similarly, the reduction in the mean weekly acenocoumarol dose was 6.07 mg (P < 0.03) as compared to the wild type after adjusting for age, gender, and body mass index.

Conclusion

Single nucleotide polymorphism in the CYP2C9 gene and in the VKORC1 gene is present in nearly 40% of Indian patients. VKORC1 (- 1639 G > A), CYP2C9*2, and CYP2C9*3 genotypes have significant dosage-lowering effects on VKA-based anti-coagulation therapy. The trend in estimated dosages of VKAs co-related to that of observed the clinically stabilized dosage in the cohort. The pharmacogenomic calculators used in this study tend to overestimate the VKA dosages as compared to clinical dosage due to the limitations in the algorithms and in our study. A new algorithm based on a larger dataset capturing the vast genetic variability across the Indian population and relevant clinical factors could provide better results.

Vitamin K antagonists and emergencies

The recent emergence of 'non-VKA' oral anticoagulants may have led to some forgetting that vitamin K antagonists (VKA) are by far the most widely prescribed oral anticoagulants worldwide. Consequently, we decided to summarize the information available on them. This paper presents the problems facing emergency physicians confronted with patients on VKAs in 10 points, from pharmacological data to emergency management. Vitamin K antagonists remain preferable in many situations including in the elderly, in patients with extreme body weights, severe chronic kidney or liver disease or valvular heart disease, and in patients taking VKAs with well-controlled international normalized ratios (INRs). Given the way VKAs work, a stable anticoagulant state can only be achieved at the earliest 5 days after starting therapy. The induction phase of VKA treatment is associated with the highest risk of bleeding; validated algorithms based on INR values have to be followed. VKA asymptomatic overdoses and 'non-severe' hemorrhage are managed by omitting a dose or stopping treatment plus administering vitamin K depending on the INR. Major bleeding is managed using a VKA reversal strategy. A prothrombin complex concentrate infusion plus vitamin K is preferred to rapidly achieve an INR of up to 1.5 and maintain a normal coagulation profile. The INR must be measured 30 min after the infusion. Before an invasive procedure, if an INR of less than 1.5 (<1.3 in neurosurgery) is required, it can be achieved by combining prothrombin complex concentrate and vitamin K. A well-codified strategy is essential for managing patients requiring emergency invasive procedures or presenting bleeding complications.

Long-term outcomes of elderly patients with CYP2C9 and VKORC1 variants treated with vitamin K antagonists

Essentials The long-term effects of VKORC1 and CYP2C9 variants on clinical outcomes remains unclear. We followed 774 patients ≥65 years with venous thromboembolism for a median duration of 30 months. Patients with CYP2C9 variants are at increased risk of death and non-major bleeding. Patients with genetic variants have a slightly lower anticoagulation quality only.

SUMMARY

Background The long-term effect of polymorphisms of the vitamin K-epoxide reductase (VKORC1) and the cytochrome P450 enzyme gene (CYP2C9) on clinical outcomes remains unclear. Objectives We examined the association between CYP2C9/VKORC1 variants and long-term clinical outcomes in a prospective cohort study of elderly patients treated with vitamin K antagonists for venous thromboembolism (VTE). Methods We followed 774 consecutive patients aged ≥ 65 years with acute VTE from nine Swiss hospitals for a median duration of 30 months. The median duration of initial anticoagulant treatment was 9.4 months. The primary outcome was the time to any clinical event (i.e. the composite endpoint of overall mortality, major and non-major bleeding, and recurrent VTE. Results Overall, 604 (78%) patients had a CYP2C9 or VKORC1 variant. Three hundred and thirty-four patients (43.2%) had any clinical event, 119 (15.4%) died, 100 (12.9%) had major and 167 (21.6%) non-major bleeding, and 100 had (12.9%) recurrent VTE. After adjustment, CYP2C9 (but not VKORC1) variants were associated with any clinical event (hazard ratio [HR], 1.34; 95% confidence interval [CI], 1.08-1.66), death (HR, 1.74; 95% CI, 1.19-2.52) and clinically relevant non-major bleeding (sub-hazard ratio [SHR], 1.39; 95% CI, 1.02-1.89), but not with major bleeding (SHR, 1.03; 95% CI, 0.69-1.55) or recurrent VTE (SHR, 0.95; 95% CI, 0.62-1.44). Patients with genetic variants had a slightly lower anticoagulation quality. Conclusions CYP2C9 was associated with long-term overall mortality and non-major bleeding. Although genetic variants were associated with a slightly lower anticoagulation quality, there was no relationship between genetic variants and major bleeding or VTE recurrence.

Verification of five pharmacogenomics-based warfarin administration models

Objective:

This study aims to screen and validate five individual warfarin dosing models (four Asian model algorithms, namely, Ohno, Wen, Miao, Huang, and the algorithm of International Warfarin Pharmacogenetic Consortium, namely IWPC algorithm) with the aim of evaluating their accuracy, practicality, and safety.

Materials and Methods:

Patients’ CYP2C9*3 and VKORC1–1639G >A genes were genotyped, and patient-related information and steady warfarin doses were recorded. The difference between the predicted dose and actual maintenance dose of each model was compared.

Results:

The prediction accuracies of the Huang and Wen models were the highest. In terms of clinical practicality, the Huang model rated the highest for the low-dose group, whereas the Ohno and IWPC models rated the highest for the middle-dose group. The models tended to markedly overpredict the doses in the low-dose group, especially the IWPC model. The Miao model tended to severely underpredict the doses in the middle-dose group, whereas no model exhibited severe overprediction.

Conclusions:

Since none of the models ranked high for all the three criteria considered, the impact of various factors should be thoroughly considered before selecting the most appropriate model for the region's population.

A New Pharmacogenetic Algorithm to Predict the Most Appropriate Dosage of Acenocoumarol for Stable Anticoagulation in a Mixed Spanish Population

There is a strong association between genetic polymorphisms and the acenocoumarol dosage requirements. Genotyping the polymorphisms involved in the pharmacokinetics and pharmacodynamics of acenocoumarol before starting anticoagulant therapy would result in a better quality of life and a more efficient use of healthcare resources. The objective of this study is to develop a new algorithm that includes clinical and genetic variables to predict the most appropriate acenocoumarol dosage for stable anticoagulation in a wide range of patients. We recruited 685 patients from 2 Spanish hospitals and 1 primary healthcare center. We randomly chose 80% of the patients (n = 556), considering an equitable distribution of genotypes to form the generation cohort. The remaining 20% (n = 129) formed the validation cohort. Multiple linear regression was used to generate the algorithm using the acenocoumarol stable dosage as the dependent variable and the clinical and genotypic variables as the independent variables. The variables included in the algorithm were age, weight, amiodarone use, enzyme inducer status, international normalized ratio target range and the presence of CYP2C9*2 (rs1799853), CYP2C9*3 (rs1057910), VKORC1 (rs9923231) and CYP4F2 (rs2108622). The coefficient of determination (R2) explained by the algorithm was 52.8% in the generation cohort and 64% in the validation cohort. The following R2 values were evaluated by pathology: atrial fibrillation, 57.4%; valve replacement, 56.3%; and venous thromboembolic disease, 51.5%. When the patients were classified into 3 dosage groups according to the stable dosage (<11 mg/week, 11-21 mg/week, >21 mg/week), the percentage of correctly classified patients was higher in the intermediate group, whereas differences between pharmacogenetic and clinical algorithms increased in the extreme dosage groups. Our algorithm could improve acenocoumarol dosage selection for patients who will begin treatment with this drug, especially in extreme-dosage patients. The predictability of the pharmacogenetic algorithm did not vary significantly between diseases.

Methods for Predicting Warfarin Dose Requirements

BACKGROUND

The aim of this study was to compare the predictive performance of different warfarin dosing methods.

METHODS

Data from 46 patients who were initiating warfarin therapy were available for analysis. Nine recently published dosing tools including 8 dose prediction algorithms and a Bayesian forecasting method were compared with each other in terms of their ability to predict the actual maintenance dose. The dosing tools included 4 algorithms that were based on patient characteristics (2 clinical and 2 genotype-driven algorithms), 4 algorithms based on international normalized ratio (INR) response feedback and patient characteristics (2 clinical and 2 genotype-driven algorithms), and a Bayesian forecasting method. Comparisons were conducted using measures of bias (mean prediction error) and imprecision [root mean square error (RMSE)].

RESULTS

The 2 genotype-driven INR feedback algorithms by Horne et al and Lenzini et al produced more precise maintenance dose predictions (RMSE, 1.16 and 1.19 mg/d, respectively; P < 0.05) than the genotype-driven algorithms by Gage et al and Klein et al and the Bayesian method (RMSE, 1.60, 1.62, and 1.81 mg/d respectively). The dose predictions from clinical and genotype-driven algorithms by Gage et al, Klein et al, and Horne et al were all negatively biased. Only the INR feedback algorithms (clinical and genotype) by Lenzini et al produced unbiased dose predictions. The Bayesian method produced unbiased dose predictions overall (mean prediction error, +0.37 mg/d; 95% confidence interval, 0.89 to -0.15) but overpredicted doses in patients requiring >8 mg/d.

CONCLUSIONS

Overall, warfarin dosing methods that included some measure of INR response (INR feedback algorithms and Bayesian methods) produced unbiased and more precise dose predictions. The Bayesian forecasting method produced positively biased dose predictions in patients who required doses >8 mg/d. Further research to assess differences in clinical endpoints when warfarin doses are predicted using Bayesian or INR-driven algorithms is warranted.

A review of a priori regression models for warfarin maintenance dose prediction

A number of a priori warfarin dosing algorithms, derived using linear regression methods, have been proposed. Although these dosing algorithms may have been validated using patients derived from the same centre, rarely have they been validated using a patient cohort recruited from another centre. In order to undertake external validation, two cohorts were utilised. One cohort formed by patients from a prospective trial and the second formed by patients in the control arm of the EU-PACT trial. Of these, 641 patients were identified as having attained stable dosing and formed the dataset used for validation. Predicted maintenance doses from six criterion fulfilling regression models were then compared to individual patient stable warfarin dose. Predictive ability was assessed with reference to several statistics including the R-square and mean absolute error. The six regression models explained different amounts of variability in the stable maintenance warfarin dose requirements of the patients in the two validation cohorts; adjusted R-squared values ranged from 24.2% to 68.6%. An overview of the summary statistics demonstrated that no one dosing algorithm could be considered optimal. The larger validation cohort from the prospective trial produced more consistent statistics across the six dosing algorithms. The study found that all the regression models performed worse in the validation cohort when compared to the derivation cohort. Further, there was little difference between regression models that contained pharmacogenetic coefficients and algorithms containing just non-pharmacogenetic coefficients. The inconsistency of results between the validation cohorts suggests that unaccounted population specific factors cause variability in dosing algorithm performance. Better methods for dosing that take into account inter- and intra-individual variability, at the initiation and maintenance phases of warfarin treatment, are needed.

Pharmacogenetics and individualizing drug treatment during pregnancy

Pharmacogenetics as a tool to aid clinicians implement individualized pharmacotherapy is utilized in some areas of medicine. Pharmacogenetics in pregnancy is still a developing field. However, there are several areas of obstetric therapeutics where data are emerging that give glimpses into future therapeutic possibilities. These include opioid pain management, antihypertensive therapy, antidepressant medications, preterm labor tocolytics, antenatal corticosteroids and drugs for nausea and vomiting of pregnancy, to name a few. More data are needed to populate the therapeutic models and to truly determine if pharmacogenetics will aid in individualizing pharmacotherapy in pregnancy. The objective of this review is to summarize current data and highlight research needs.

A model predicting fluindione dose requirement in elderly inpatients including genotypes, body weight, and amiodarone

Indandione VKAs have been widely used for decades, especially in Eastern Europe and France. Contrary to coumarin VKAs, the relative contribution of individual factors to the indandione-VKA response is poorly known. In the present multicentre study, we sought to develop and validate a model including genetic and non-genetic factors to predict the daily fluindione dose requirement in elderly patients in whom VKA dosing is challenging. We prospectively recorded clinical and therapeutic data in 230 Caucasian inpatients mean aged 85 ± 6 years, who had reached international normalized ratio stabilisation (range 2.0-3.0) on fluindione. In the derivation cohort (n=156), we analysed 13 polymorphisms in seven genes potentially involved in the pharmacological effect or vitamin-K cycle (VKORC1, CYP4F2, EPHX1) and fluindione metabolism/transport (CYP2C9, CYP2C19, CYP3A5, ABCB1). We built a regression model incorporating non-genetic and genetic data and evaluated the model performances in a separate cohort (n=74).Body-weight, amiodarone intake, VKORC1, CYP4F2, ABCB1 genotypes were retained in the final model, accounting for 31.5% of dose variability. None influence of CYP2C9 was observed. Our final model showed good performances: in 83.3% of the validation cohort patients, the dose was accurately predicted within 5 mg, i.e.the usual step used for adjusting fluindione dosage. In conclusion, in addition to body-weight and amiodarone-intake, pharmacogenetic factors (VKORC1, CYP4F2, ABCB1) related to the pharmacodynamic effect and transport of fluindione significantly influenced the dose requirement in elderly patients while CYP2C9 did not. Studies are required to know whether fluindione could be an alternative VKA in carriers of polymorphic CYP2C9 alleles, hypersensitive to coumarins.

A review of pharmacogenetics of adverse drug reactions in elderly people

Older adults are more susceptible to the prevalence of therapeutic failure and adverse drug reactions (ADRs). Recent advances in genomic research have shed light on the crucial role of genetic variants, mainly involving genes encoding drug-metabolizing enzymes, drug transporters and genes responsible for a compound's mechanism of action, in driving different treatment responses among individuals, in terms of therapeutic efficacy and safety. The interindividual variations of these genes may account for the differences observed in drug efficacy and the appearance of ADRs in elderly people. The advent of whole genome mapping techniques has allowed researchers to begin to characterize the genetic components underlying serious ADRs. The identification and validation of these genetic markers will enable the screening of patients at risk of serious ADRs and to establish personalized treatment regimens.The aim of this review was to provide an update on the recent developments in geriatric pharmacogenetics in clinical practice by reviewing the available evidence in the PubMed database to September 2012. A Pubmed search was performed (years 1999-2012) using the following two search strategies: ('pharmacogenomic' OR 'pharmacogenetic ') AND ('geriatric' or 'elderly ') AND 'adverse drug reactions'; [gene name] AND ('geriatric' or 'elderly ') AND 'adverse drug reactions', in which the gene names were those contained in the Table of Pharmacogenomic Biomarkers in Drug Labels published online by the US Food and Drug Administration ( http://www.fda.gov/drugs/scienceresearch/researchareas/pharmacogenetics/ucm083378.htm ). Reference lists of included original articles and relevant review articles were also screened. The search was limited to studies published in the English language.

Efficiency and effectiveness of the use of an acenocoumarol pharmacogenetic dosing algorithm versus usual care in patients with venous thromboembolic disease initiating oral anticoagulation: study protocol for a randomized controlled trial

Background

Hemorrhagic events are frequent in patients on treatment with antivitamin-K oral anticoagulants due to their narrow therapeutic margin. Studies performed with acenocoumarol have shown the relationship between demographic, clinical and genotypic variants and the response to these drugs. Once the influence of these genetic and clinical factors on the dose of acenocoumarol needed to maintain a stable international normalized ratio (INR) has been demonstrated, new strategies need to be developed to predict the appropriate doses of this drug. Several pharmacogenetic algorithms have been developed for warfarin, but only three have been developed for acenocoumarol. After the development of a pharmacogenetic algorithm, the obvious next step is to demonstrate its effectiveness and utility by means of a randomized controlled trial. The aim of this study is to evaluate the effectiveness and efficiency of an acenocoumarol dosing algorithm developed by our group which includes demographic, clinical and pharmacogenetic variables (VKORC1, CYP2C9, CYP4F2 and ApoE) in patients with venous thromboembolism (VTE).

Methods and design

This is a multicenter, single blind, randomized controlled clinical trial. The protocol has been approved by La Paz University Hospital Research Ethics Committee and by the Spanish Drug Agency. Two hundred and forty patients with VTE in which oral anticoagulant therapy is indicated will be included. Randomization (case/control 1:1) will be stratified by center. Acenocoumarol dose in the control group will be scheduled and adjusted following common clinical practice; in the experimental arm dosing will be following an individualized algorithm developed and validated by our group. Patients will be followed for three months. The main endpoints are: 1) Percentage of patients with INR within the therapeutic range on day seven after initiation of oral anticoagulant therapy; 2) Time from the start of oral anticoagulant treatment to achievement of a stable INR within the therapeutic range; 3) Number of INR determinations within the therapeutic range in the first six weeks of treatment.

Discussion

To date, there are no clinical trials comparing pharmacogenetic acenocoumarol dosing algorithm versus routine clinical practice in VTE. Implementation of this pharmacogenetic algorithm in the clinical practice routine could reduce side effects and improve patient safety.

Trial registration

Eudra CT. Identifier: 2009-016643-18.

Effect of genetic variants, especially CYP2C9 and VKORC1, on the pharmacology of warfarin

The genes encoding the cytochrome P450 2C9 enzyme (CYP2C9) and vitamin K-epoxide reductase complex unit 1 (VKORC1) are major determinants of anticoagulant response to warfarin. Together with patient demographics and clinical information, they account for approximately one-half of the warfarin dose variance in individuals of European descent. Recent prospective and randomized controlled trial data support pharmacogenetic guidance with their use in warfarin dose initiation and titration. Benefits from pharmacogenetics-guided warfarin dosing have been reported to extend beyond the period of initial dosing, with supportive data indicating benefits to at least 3 months. The genetic effects of VKORC1 and CYP2C9 in African and Asian populations are concordant with those in individuals of European ancestry; however, frequency distribution of allelic variants can vary considerably between major populations. Future randomized controlled trials in multiethnic settings using population-specific dosing algorithms will allow us to further ascertain the generalizability and cost-effectiveness of pharmacogenetics-guided warfarin therapy. Additional genome-wide association studies may help us to improve and refine dosing algorithms and potentially identify novel biological pathways.

An acenocoumarol dosing algorithm using clinical and pharmacogenetic data in Spanish patients with thromboembolic disease

Appropriate dosing of coumarins is difficult to establish, due to significant inter-individual variability in the dose required to obtain stable anticoagulation. Several genetic and other clinical factors have been associated with the coumarins dose, and some pharmacogenetic-guided dosing algorithms for warfarin and acenocoumarol have been developed for mixed populations. We recruited 147 patients with thromboembolic disease who were on stable doses and with an international normalized ratio (INR) between 2 and 3. We ascertained the influence of clinical and genetic variables on the stable acenocoumarol dose by multiple linear regression analysis in a derivation cohort (DC; n = 117) and developed an algorithm for dosing that included clinical factors (age, body mass index and concomitant drugs) and genetic variations of VKORC1, CYP2C9, CYP4F2 and APOE. For purposes of comparison, a model including only clinical data was created. The clinical factors explained 22% of the dose variability, which increased to 60.6% when pharmacogenetic information was included (p<0.001); CYP4F2 and APOE variants explained 4.9% of this variability. The mean absolute error of the predicted acenocoumarol dose (mg/week) obtained with the pharmacogenetic algorithm was 3.63 vs. 5.08 mg/week with the clinical algorithm (95% CI: 0.88 to 2.04). In the testing cohort (n = 30), clinical factors explained a mere 7% of the dose variability, compared to 39% explained by the pharmacogenetic algorithm. Considering a more clinically relevant parameter, the pharmacogenetic algorithm correctly predicted the real stable dose in 59.8% of the cases (DC) vs. only 37.6% predicted by the clinical algorithm (95% CI: 10 to 35). Therefore the number of patients needed to genotype to avoid one over- or under-dosing was estimated to be 5.

Pharmacogenetic warfarin dose refinements remain significantly influenced by genetic factors after one week of therapy

By guiding initial warfarin dose, pharmacogenetic (PGx) algorithms may improve the safety of warfarin initiation. However, once international normalised ratio (INR) response is known, the contribution of PGx to dose refinements is uncertain. This study sought to develop and validate clinical and PGx dosing algorithms for warfarin dose refinement on days 6-11 after therapy initiation. An international sample of 2,022 patients at 13 medical centres on three continents provided clinical, INR, and genetic data at treatment days 6-11 to predict therapeutic warfarin dose. Independent derivation and retrospective validation samples were composed by randomly dividing the population (80%/20%). Prior warfarin doses were weighted by their expected effect on S-warfarin concentrations using an exponential-decay pharmacokinetic model. The INR divided by that "effective" dose constituted a treatment response index . Treatment response index, age, amiodarone, body surface area, warfarin indication, and target INR were associated with dose in the derivation sample. A clinical algorithm based on these factors was remarkably accurate: in the retrospective validation cohort its R(2) was 61.2% and median absolute error (MAE) was 5.0 mg/week. Accuracy and safety was confirmed in a prospective cohort (N=43). CYP2C9 variants and VKORC1-1639 G→A were significant dose predictors in both the derivation and validation samples. In the retrospective validation cohort, the PGx algorithm had: R(2)= 69.1% (p<0.05 vs. clinical algorithm), MAE= 4.7 mg/week. In conclusion, a pharmacogenetic warfarin dose-refinement algorithm based on clinical, INR, and genetic factors can explain at least 69.1% of therapeutic warfarin dose variability after about one week of therapy.