Pharmacogenetic Testing to Predict Warfarin Response
Author: Kristine Reckenberg, PharmD, PGY1 Pharmacy Practice ResidentPreceptor: Justinne Guyton, PharmD, BCACP, PGY1 Pharmacy Practice Residency Director, St. Louis County Department of Public Health/St. Louis College of Pharmacy
Program Number: 2018-11-05Approval Dates: December 1, 2018 - June 1, 2019Approved Contact Hours: One (1) CE(s) per LIVE session.
Pharmacogenetics is a subtype of pharmacogenomics in which polymorphisms in genes that encode drug metabolizing enzymes, transporters, and/or targets can impact drug effects, leading to variability among individuals in response to a medication.1 Warfarin, a vitamin K antagonist (VKA), is an oral anticoagulant commonly used for prevention of stroke in atrial fibrillation and prevention and treatment of venous thromboembolism.1 It is also an agent that displays pharmacogenetic variations between individuals that impact both pharmacokinetics and pharmacodynamics.2 Established pharmacogenetic variables are caused by variations in the cytochrome P450 system and the warfarin targets.1 These variables have the potential for a large impact as warfarin is an agent with a narrow therapeutic index that is associated with serious adverse events, notably bleeding. These differences are most significant during the dose finding stage of warfarin initiation. Based on these factors, warfarin can be initiated via a pharmacogenetic dosing algorithm, dosed clinically (dosing based off of clinical factors – age, medications, comorbidities, bleed risk, and social history), or using a standard dose approach.3
Two guidelines provide recommendations regarding the use of warfarin pharmacogenetic testing prior to warfarin initiation. The Clinical Pharmacogenetics Implementation Consortium Guidelines (CPIC) has been updated with the results of newly published trials whereas the American College of Chest Physicians Guidelines (ACCP) have not.3-4 This continuing education article will focus on the factors that impact warfarin pharmacogenetics, current guideline recommendations, and recently published data regarding warfarin pharmacogenetics.
There are two predominant genes, that have been studied extensively, that contribute to the interpatient variability in warfarin dose requirements. These include cytochrome P450 2C9 (CYP450 2C9) and vitamin K epoxide reductase complex 1 (VKORC1).5 CYP450 4F2 also plays a role in variability, however, it has demonstrated a smaller part in dose requirements during warfarin initiation.5
CYP450 2C9 is associated with multiple different polymorphisms, however, the nonsynonymous single nucleotide polymorphisms (SNPs) that influence warfarin dose requirements the most are the *2 and *3 alleles.5-6 These SNPs lead to a reduction in the enzymatic activity of S-warfarin.5-6 The outcome is a decrease in warfarin requirements due to the slowed metabolism, as the S-enantiomer of warfarin has a greater anticoagulant effect as compared to the R-enantiomer. CYP450 2C9*2 is associated with only 70% activity as compared to the wild type allele, whereas CYP450 2C9*3 is associated with only 20% activity.6 These alleles are present most frequently in Europeans, and rarely in African Americans and Asians.5
VKORC1 is an enzyme that is responsible for the regeneration of reduced vitamin K, and is the target of warfarin therapy.5-6 The SNPs that influence this enzyme include 1173 C>T and 1639 G>A.6 These alleles result in decreased translation of mRNA into proteins; ultimately leading to a lower level of expression of the VKORC1 enzyme.5 This decrease in VKORC1 is associated with lower warfarin dose requirements accounting for 25 + 8% of the variance in warfarin dose requirements.5-6 VKORC1 1639 A/A has been found to decrease initial warfarin dose requirements by approximately 40%.14 Those patients that have the wild-type haplotype (GG) have a higher rate of metabolism as compared to those that are homozygous for the variant allele (AA).6 Those that are heterozygous for the variant allele have an intermediate rate of metabolism.6 The A allele is present most frequently in patients of Asian ethnicity, followed by Europeans, and rarely in African Americans.5
CYP450 4F2 is one of the CYP450 enzymes responsible for metabolism of vitamin K.5 Those patients with the CYP450 4F2 V433M SNP have decreased metabolism of vitamin K, resulting in higher warfarin dose requirements.5 Despite this finding, CYP450 4F2 V433M SNP only contributes to approximately one percent of variability in warfarin dose requirements.5 This polymorphism is present in Europeans and Asians, but rarely in African Americans.5
Table 1 located in the appendix provides ranges of expected warfarin total daily doses based on CYP450 2C9 and VKORC1 polymorphisms, according to the manufacturer.
Current Guideline Recommendations
There are two main sources that provide recommendations regarding warfarin pharmacogenetic testing for initiation of warfarin therapy. These include guidelines from the Clinical Pharmacogenetics Implementation Consortium (CPIC) and the American College of Chest Physicians (ACCP).3-4
The CPIC Guidelines, prepared by an international consortium of volunteers, released an update in 2017 following release of results of the Genetics-Informatics Trial (GIFT).3 The CPIC recommends that in patients of non-African ancestry, warfarin dosing should be calculated using a published pharmacogenetic algorithm including genotype information for VKORC1 1639 C>A, CYP450 2C9*2 and *3.3 If this genetic information is not available, warfarin should be initiated based on clinical factors; this has a strong recommendation rating.3 In patients of African ancestry it is important to know the additional genotypes of CYP2C9*5, *6, *8, and *11.3 If available, the warfarin dose should be calculated using a validated pharmacogenetic algorithm including the variables of VKORC1 1639 C>A, CYP2C9*2 and *3.3 If the patient carries CYP2C9*5, *6, *8, or *11 alleles, the dose should be decreased by 15-30%; with a moderate recommendation rating by the CPIC.3 If this genetic information is unavailable, warfarin should be dosed clinically.3
The ACCP Guidelines published in 2012 recommend against the routine use of pharmacogenetic testing for guiding doses of vitamin K antagonists with a grade of 1B.4 This is due to the limited availability of four small randomized controlled trials, availability of a single systematic review concluding lack of evidence to support using pharmacogenetic testing to guide therapy, and multiple cost effectiveness analyses indicating lack of cost-effectiveness at the time of guideline update.4
Summary of Evidence
Before pharmacogenetic dosing of warfarin can be considered as a potential for adoption in routine clinical practice, there must be sufficient evidence provided by large randomized controlled trials. To date most randomized controlled trials have had small sample sizes with the most recent study being the largest to date. Here the three major landmark clinical trials and one systematic review exploring pharmacogenetic-guided warfarin dosing versus clinically-guided warfarin dosing will be reviewed.
The European Pharmacogenetics of Anticoagulant Therapy (EU-PACT) trial by Verhoef and colleagues was published first in 2013.8 This study combined data from two single-blind, randomized trials for the initiation of VKAs, acenocoumarol or phenprocoumon, in the treatment of 548 patients with atrial fibrillation or venous thromboembolism. VKAs were initiated with either a genotype-guided algorithm or clinically-guided algorithm that included clinical variables and genotyping for CYP2C9 and VKORC1, or a dosing algorithm that included only clinical variables for the first 5-7 days. After the first 5-7 days, the patients were treated based on international normalized ration (INR) and local clinical practice with intended follow-up for 12 weeks.8 The primary outcome, percent of time in therapeutic INR, was observed in 61.6% of patients in the genotype-guided group and 60.2% in the clinically-guided group (p = 0.52) during the first 12 weeks.8 However, the percentage of time in therapeutic range after the first four weeks of treatment was 52.8% in the genotype-guided group versus 47.5% in the clinically-guided group (p = 0.02). There were no differences in bleeding event rates or thromboembolic events.8
The Clarification of Optimal Anticoagulation Through Genetics (COAG) trial by Kimmel and colleagues was published in 2013.9 This was a multicenter, double blind trial comparing a warfarin dosing algorithm including genotype-guided therapy versus clinically-guided therapy during the first 5 days of warfarin therapy in 1015 patients.9 The patients received follow-up through the first 4 weeks of therapy.9 The primary outcome (percent of time in therapeutic INR from day 4 or 5 through day 28 of therapy) was observed in 45.2% in the genotype-guided group and 45.4% in the clinically guided group, with an adjusted mean difference of -0.2 (95% CI -3.4 to 3.1, p = 0.91).9 Statistically significant findings were reported with subgroups; black patients had a lower mean percentage of time in the therapeutic range in the genotype-guided group as compared to the clinically-guided group.9 There were no significant differences with regard to INR > 4, major bleeding, or thromboembolism.9
The most recent warfarin pharmacogenetic study is the Genetic Informatics Trial (GIFT) published by Gage and colleagues in 2017.10 This was a multicenter randomized clinical trial in patients initiating warfarin at the time of elective hip or knee arthroplasty. A total of 1597 patients were randomized to receive genotype-guided dosing of warfarin during the first 11 days or clinically-guided dosing of warfarin with follow-up for 90 days.10 The primary outcome (a composite of major bleeding within 30 days, INR of 4 or greater within 30 days, death within 30 days, and symptomatic or asymptomatic venous thromboembolism within 60 days of arthroplasty) was observed in 10.8% of patients in the genotype-guided group versus 14.7% in the clinically-guided group with an absolute difference of 3.9% (95% CI 0.7-7.2%, p = 0.02).10
Lastly, a meta-analysis was published in 2014 by Stergiopoulos and colleagues, which included nine randomized controlled trials and 2812 patients, and compared genotype-guided initial dosing of warfarin and its analogues to clinical dosing protocols.11 In this study, the standardized difference in means of the percent of time that the INR was therapeutic was 0.14 (95% CI -0.10-0.39, p = 0.25).11 There was not a significant difference found for risk ratio for an INR greater than 4, major bleeding, or thromboembolic events.11
With regards to pharmacoeconomic studies, varying results have been presented. Three studies found pharmacogenetic-guided warfarin dosing to be cost effective, whereas four studies were inconclusive, and five studies found pharmacogenetic-guided warfarin dosing to not be cost effective.12 Furthermore, coverage of pharmacogenetic testing for warfarin initiation varies based on insurance company. The Centers for Medicare and Medicaid Services (CMS) released an update regarding their organization’s coverage of this testing. As of January 25th, 2018, CMS does not believe that current evidence supports pharmacogenetic testing for CYP2C9 or VKORC1 for warfarin initiation due to limited studies indicating a benefit in health outcomes.12 Testing will not be covered under the Social Security Act, but may be covered under the Coverage with Evidence Development (CED) section of the Social Security Act if specific requirements are meant.12 According to CMS, these include patients who are “candidates for anticoagulation with warfarin who: have not been previously tested for CYP2C9 or VKORC1 alleles; and have received fewer than 5 days of warfarin in the anticoagulation regimen for which the testing is ordered; and are enrolled in a prospective, randomized, controlled clinical study when that study meets the standards specified in the decision memorandum.”12
Pharmacogenomic studies have indicated that warfarin pharmacogenetics can play a role in warfarin dose requirements through polymorphisms in CYP450 2C9 and VKORC1. A number of randomized controlled trials, each with their own limitations have presented variable results with regards to the benefit of pharmacogenetic versus clinically guided dosing of warfarin. In addition, cost-effectiveness studies have also produced opposing results. Therefore, clinicians should be aware of the impact pharmacogenetic testing results can have on warfarin dose adjustments and use this information, should it be available. Pharmacists should be aware of advances in pharmacogenetic dose adjustments of medications to serve as a resource for both other healthcare providers and patients.
Submit for CE