Abstract
We present an hypothetical case to illustrate the potential use of pharmacogenetics in clinical practice. A 70-year-old patient of European descent is diagnosed with atrial fibrillation. The patient is 5’9, 200 pounds with a body mass index of 29.5 kg/m2. He has a history of chronic hypertension along with hyperlipidemia, and his daily medical regimen consists of hydrochlorothiazide, lisinopril, and atorvastatin. He quit smoking 20 years ago and is not currently on any other medications. The patient’s internist prescribes the anticoagulant warfarin to decrease the risk of stroke. The therapeutic window for warfarin is narrow and more challenging to use safely in patients of advanced age who have decreased capacity to metabolize drugs and an increased risk of falling that increases the risk of bleeding. These risks lead the patient’s internist to consider using pharmacogenetics to determine the warfarin dosage.
What is pharmacogenetic testing and how can it be helpful?
Pharmacogenetics can help physicians deliver individualized treatments based on how a person’s genes impact a drug’s effects and metabolism. Genetic polymorphisms in the genes encoding the drug target impact pharmacodynamics while genetic determinants of the drug’s metabolism or excretion such as the cytochrome P450 enzyme superfamily influence pharmacokinetics. In the United States, the 30 most commonly prescribed drugs with potential for pharmacogenetic utility account for over 15% of all prescriptions (1).
Pharmacogenetics can help prevent adverse events or improve drug efficacy by enabling the physician to optimize dosage or avoid a medication with adverse reactions and prescribe an alternative therapy. Additionally, drugs like warfarin have a narrow therapeutic range and genotype-based treatment can help minimize the risk of adverse events while maximizing time in the therapeutic window. Pharmacogenetic testing can also identify patients with variants who are responsive to certain drugs, such as cystic fibrosis patients with the G551D CFTR mutation who may benefit from Ivacaftor (2). Furthermore, pharmacogenetics is most impactful in populations in which the variant allele is found at high frequency. For instance, performing a genetic test for the HLA-B*15:02 variant to prevent associated Steven’s Johnson syndrome when taking carbamazepine is most beneficial in Southeast Asians because of the high HLA-B*15:02 allele frequency in the Southeast Asian population (2).
Which genes influence response to warfarin?
The warfarin-specific genes and their effect on determining warfarin dose are summarized in Table 1 along with specific recommendations about how much the warfarin dose should be adjusted based upon genotype.
Table 1.
Genetic polymorphisms affecting warfarin dosage
| Gene | Variant Allele | Allele Frequency by Ethnic Group* | Impact on Warfarin Dosing |
|---|---|---|---|
| VKORC1 | VKORC1 –1639G>A | East Asian 88% European Descent 41% African American 10% |
Decrease dose by 28% per variant allele† |
| CYP2C9 | CYP2C9*2 | East Asian 0.06% European Descent 13% African American 2.3% |
Decrease dose by 19% per variant allele† |
| CYP2C9*3 | East Asian 3.4% European Descent 7.1% African American 1.2% |
Decrease dose by 33% per variant allele† | |
| CYP2C9*5 | East Asian 0% European Descent 0% African American 1.3% |
Decrease dose by 15–30% per variant allele‡ | |
| CYP2C9*6 | East Asian 0% European Descent 0% African American 0.77% |
||
| CYP2C9*8 | East Asian 0% European Descent 0.14% African American 6.7% |
||
| CYP2C9*11 | East Asian 0% European Descent 0.17% African American 1.4% |
||
| rs12777823 A | African Americans ~25% | Decrease dose by 10–25%‡ | |
| CYP4F2 | CYP4F2 V433M | East Asian 22% European Descent 30% African American 7.7% |
Increase dose by 5–10%‡ |
All data on variant frequencies referenced from 3 (https://www.pharmgkb.org/page/pgxGeneRef))
Reference 4
Reference 5
Vitamin K epoxide reductase, encoded by the gene VKORC1, reduces vitamin K and allows vitamin K to activate coagulation factors II, VII, IX, and X. VKORC1 is the target of warfarin, and warfarin produces anticoagulation effects by inhibiting vitamin K epoxide reductase. The genetic polymorphism VKORC1 −1639G>A (guanine to adenine substitution at position −1639) is associated with increased sensitivity to warfarin, and patients with genotypes A/A or G/A require a lower dose of warfarin than G/G individuals (5). This is a pharmacodynamics genetic difference.
CYP2C9 of the CYP superfamily is the hepatic enzyme responsible for the metabolism of warfarin or a pharmacokinetic genetic difference. Warfarin is a 50:50 racemic mixture of R-warfarin and S-warfarin enantiomers; S-warfarin is five times more active than R-warfarin. An asterisk after the CYP2C9 enzyme denotes the specific genetic variant or allele. CYP2C9*1 is the wild-type allele while CYP2C9*2 and CYP2C9*3 are variant alleles that are associated with slower warfarin metabolism (5). There is a ~30–40% decrease in the metabolism of S-warfarin in patients with the *2 allele and a ~80–90% decrease in metabolism in patients with the *3 allele (5). The *5, *6, *8, and *11 alleles are found at higher frequency in African Americans. Additionally, inclusion of the rs12777823 polymorphism in CYP2C cluster is also helpful to determine dosing in African Americans (5,6).
The CYP4F2 enzyme oxidizes vitamin K to the inactive form, and carriers of the CYP4F2 V433M polymorphism have decreased ability to oxidize vitamin K and have higher hepatic levels of activated vitamin K. Therefore, a higher dose of warfarin is required in V433M carriers (5).
What evidence is available to support the clinical utility of pharmacogenetics for warfarin and for other drugs?
Among drugs with potential for pharmacogenetic implementation, randomized controlled clinical trials provide the highest level of evidence to assess clinical utility, and these trials must be completed in each ethnic group for which testing is considered since the clinical utility can differ by ethnic group and the genetic variants assessed. For some drugs, there is evidence that the majority of individuals with a particular serious adverse event all carry a specific genotype, and on this basis pharmacogenetic testing is recommended to avoid adverse events (Table 2). For the majority of drugs with associations between genotype and drug dosing, there are only sparse data in the form of case reports to demonstrate the clinical utility of pharmacogenetic testing. Drugs like clopidogrel and tacrolimus have been tested through randomized controlled trials but the results of the benefit of pharmacogenetic testing are inconclusive (Table 2).
Table 2. The Clinical Pharmacogenetics Implementation Consortium (CPIC) level A gene/drug pairs.
The CPIC considers all level A drugs to have high or moderate evidence in favor of using available genetic information when prescribing treatment, and at least one moderate or strong clinical action is recommended in the CPIC guidelines. The common genetic variants that influence the response of each drug is included along with a description of the potential benefit of using pharmacogenetics to prescribe medications and the level of evidence to assess the clinical utility of pharmacogenetic testing.
| Drug | Genetic* variants | Potential Pharmacogenetic Benefit* | Level of Evidence to Assess Clinical Utility | Food and Drug Administration Labeling Requirement for Testing* |
|---|---|---|---|---|
| Antithrombic | ||||
| Clopidogrel | CYP2C19*1, *2, *3, *17 | Prevent cardiovascular events in patients with acute coronary syndromes undergoing percutaneous coronary intervention | Six randomized controlled trials† Conflicting evidence for pharmacogenetics. |
Not required |
| Warfarin | CYP2C9*1, *2, *3, *5, *6, *8, *11, rs12777823, VKORC1 –1639G>A, CYP4F2*3 | Achieve therapeutic INR faster, increase time within therapeutic INR, reduce adverse bleeding events | Three large randomized controlled trials‡ Conflicting evidence for pharmacogenetics. |
Not required |
| Narcotics | ||||
| Codeine | CYP2D6*1, *2, *3, *4, *5, *6, *9, *10, *41 | Identify patients with either high risk of drug toxicity (respiratory depression, nausea, vomiting, constipation) or ineffective response to codeine | No studies | Not required |
| Oxycodone (refer to codeine CPIC guideline) |
CYP2D6*1, *2, *3, *4, *5, *6 | Identify patients with either high risk of drug toxicity (respiratory depression, nausea, vomiting, constipation) or ineffective response to oxycodone | No studies | Not required |
| Tramadol (refer to codeine CPIC guideline) |
CYP2D6*1, *2, *3, *4, *5, *6, *10 | Prevent adverse events (respiratory depression, cardiotoxicity, or nausea) and maintain drug efficacy | No studies | Not required |
| Antidepressants | ||||
| Amitriptyline |
CYP2D6*1, *2, *3, *4, *5, *6, *9, *10, *41 CYP2C19*1, *2, *3, *17 |
Prevent side effects (anticholinergic, CNS, gastrointestinal, cardiovascular) and increase likelihood of drug efficacy | No studies | Not required |
| Nortriptyline | CYP2D6*1, *2, *3, *4, *5, *6, *9, *10, *41 | Prevent side effects (anticholinergic, CNS, gastrointestinal, cardiovascular) and increase likelihood of drug efficacy | No studies | Not required |
| Fluvoxamine | CYP2D6*1, *2, *3, *4, *5, *6, *9, *10, *41 | Prevent side effects (CNS, gastrointestinal, sexual) | No studies | Not required |
| Citalopram | CYP2C19*1, *2, *3, *17 | Prevent adverse events (CNS, gastrointestinal, sexual, arrhythmias) and increase likelihood of drug efficacy | No studies | Not required |
| Escitalopram | CYP2C19*1, *2, *3, *17 | Prevent adverse events (CNS, gastrointestinal, sexual, arrhythmias) and increase likelihood of drug efficacy | No studies | Not required |
| Paroxetine | CYP2D6*1, *2, *3, *4, *5, *6, *9, *10, *41 | Prevent side effects (CNS, gastrointestinal, sexual) | No studies | Not required |
| Anticonvulsant | ||||
| Carbamazepine | HLA-B*15:02, HLA-A*31:01 | Reduce incidence of Stevens-Johnson syndrome, toxic epidermal necrolysis, maculopapular exanthema, and other cutaneous adverse reactions; greater benefit in Southeast Asian populations (HLA-B*15:02 Allele Frequencies: East Asian 6.9%, Oceanian 5.4%, South/Central Asian 4.6%) | Prospective screening study of HLA-B*15:02 in Taiwan§ Supportive evidence for pharmacogenetics. |
Testing required for HLA-B*15:02 |
| Phenytoin | HLA-B*15:02, CYP2C9*1, *2, *3 | Reduce risk of Steven Johnson Syndrome and toxic epidermal necrolysis in patients with HLA-B*15:02, avoid adverse events (sedation, ataxia, dizziness, nystagmus, nausea, cognitive impairment) in patients who are CYP2C9 poor metabolizers; high frequency of benefit in Southeast Asian populations (HLA-B*15:02 Allele Frequencies: East Asian 6.9%, Oceanian 5.4%, South/Central Asian 4.6%) | No studies | Not required |
| Oxcarbazepine | HLA-B*15:02 | Reduce risk of Stevens-Johnson syndrome and toxic epidermal necrolysis; high frequency of benefit in Southeast Asian populations (HLA-B*15:02 Allele Frequencies: East Asian 6.9%, Oceanian 5.4%, South/Central Asian 4.6%) | No studies | Not required |
| Antiemetic | ||||
| Ondasetron | CYP2D6*1, *2, *3, *4, *5, *6, *9, *10, *41 | Identify patients who do not respond effectively to Ondasetron | No studies | Not required |
| Tropisetron | CYP2D6*1, *2, *3, *4, *5, *6, *9, *10, *41 | Identify patients who do not respond effectively to Tropisetron | No studies | Not required |
| Anticancer/Immunosuppressive | ||||
| Capecitabine | DYPYD*2A, *13, rs67376798, rs75017182 | Reduce risk of drug toxicity (neutropenia, nausea, vomiting, severe diarrhea, stomatitis, mucositis, hand-foot syndrome) | No studies | Not required |
| Fluorouracil | DYPYD*2A, *13, rs67376798, rs75017182 | Reduce risk of drug toxicity (neutropenia, nausea, vomiting, severe diarrhea, stomatitis, mucositis, hand-foot syndrome) | No studies | Not required |
| Thioguanine | TPMT*1, *2, *3A, *3B, *3C, *4 | Reduce risk of acute myelosuppression and maintain drug efficacy | No studies | Not required |
| Azathioprine | TPMT*1, *2, *3A, *3B, *3C, *4 | Reduce risk of acute myelosuppression and maintain drug efficacy | Two randomized controlled trials|| Supportive evidence for pharmacogenetics. |
Not required |
| Mercaptopurine | TPMT*1, *2, *3A, *3B, *3C, *4 | Reduce risk of acute myelosuppression and maintain drug efficacy | One randomized controlled trial|| Supportive evidence for pharmacogenetics. | Not required |
| Tacrolimus | CYP3A5*1, *3, *6, *7 | Achieve target concentrations faster to reduce risk of graft rejection after transplantation and reduce the risk of drug toxicity (nephrotoxicity, hypertension, neurotoxicity, hyperglycemia) | Two randomized controlled trials¶ Conflicting evidence for pharmacogenetics |
Not required |
| Tamoxifen | CYP2D6*1, *2, *3, *4, *5, *6, *9, *10, *17, *41 | Optimize dose or identify patients who do not respond effectively to Tamoxifen | No studies | Not required |
| Irinontecan (CPIC guideline not yet available) | UGT1A1*1, *28 | Reduce risk for neutropenia, diarrhea, and asthenia | No studies | Not required |
| Anesthetic | ||||
| Isoflurane (CPIC guideline not yet available) | CACNA1S and RYR1 | Reduce risk of malignant hyperthermia | No studies | Not required |
| Desflurane (CPIC guideline not yet available) | CACNA1S and RYR1 | Reduce risk of malignant hyperthermia | No studies | Not required |
| Sevoflurane (CPIC guideline not yet available) | CACNA1S and RYR1 | Reduce risk of malignant hyperthermia | No studies | Not required |
| Antiviral | ||||
| Abacavir | HLA-B*57:01 | Reduce risk of hypersensitivity reactions | Randomized controlled trial with prospective screening for HLA-B*57:01|| Supportive evidence for pharmacogenetics. | Testing required |
| Atazanavir | UGT1A1*28, *37, rs887829 | Identify patients with a high risk of developing hyperbilirubinemia | No studies | Not required |
| Peginterferon alfa-2a | IFNL3 rs12979860 and rs8099917 | Predict drug response and predict eligibility for shorter durations of therapy when used in combination with protease inhibitors | No studies | Not required |
| Peginterferon alfa-2b | IFNL3 rs12979860 and rs8099917 | Predict drug response and eligibility for shorter durations of therapy used in combination with protease inhibitors | No studies | Not required |
| Ribavirin | IFNL3 rs12979860 and rs8099917 | Predict drug response and eligibility for shorter durations of therapy when used in combination with protease inhibitors | No studies | Not required |
| Antifungal | ||||
| Voriconazole | CYP2C19*1, *2, *3, *17 | Prevent side effects (hepatotoxicity, visual disturbances, visual hallucinations) or identify patients who do not respond effectively to Voriconazole | No studies | Not required |
| Antigout | ||||
| Allopurinol | HLA-B*58:01 | Reduce risk of hypersensitivity syndrome, Stevens-Johnson syndrome, and toxic epidermal necrolysis; high frequency of benefit in Han-Chinese and Thai populations (HLA-B*58:01 Allele Frequency: Taiwan Han Chinese 11%) | Prospective screening trial for HLA-B*58:01 in Taiwan§ Supportive evidence for pharmacogenetics. |
Not required |
| Rasburicase | G6PD I, II, III, IV | Reduce risk of acute hemolytic anemia; high frequency of benefit in Asia, Europe, Africa, and the Middle East | No studies | Testing required |
| Cystic Fibrosis | ||||
| Ivacaftor | G551D-CFTR (rs75527207) | Identify patients who could benefit from Ivacaftor treatment | Clinical trials provide supportive evidence for pharmacogenetics.§ | Testing required |
| Paralytic | ||||
| Succinylcholine (CPIC guideline not yet available) |
CACNA1S rs1800559 and rs772226819 | Reduce risk of malignant hyperthermia | No studies | Not required |
| Lipid-lowering | ||||
| Simvastatin | SLCO1B1*1, *5, *15, *17 | Reduce risk of myopathies and rhabdomyolysis | No studies | Not required |
There have been three large randomized controlled trials evaluating the clinical utility of genotype-guided warfarin therapy with somewhat conflicting results influenced by the patient population and genotypes tested in the studies. In the U.S Clarification of Optimal Anticoagulation through Genetics (COAG) trial (8), patients were treated with either a clinically-derived algorithm without pharmacogenetics or a genotype-based algorithm that incorporated the genotype data for CYP2C9*2, CYP2C9*3, and VKORC1 polymorphisms together with clinical factors. The COAG trial concluded that there was no significant difference in the primary outcome of percent of time in therapeutic range between the genotype-guided group (45.2%) and the clinically guided group (45.4%). No significance difference was observed in the rate of adverse bleeding events. The COAG trial consisted of 27% African Americans, and these patients spent significantly less time within the therapeutic range in the genotype-guided group than the clinically-guided group (35.2% vs 43.5%). COAG did not include all the relevant alleles for the African American population (CYP2C9 *5 *6 *8 *11, rs12777823), and that could partially account for the failure of the genotype-based algorithm to impact overall outcome.
The EU-PACT trial used the same genetic polymorphisms and the same primary outcome measures as the COAG trial (9). Overall, the EU-PACT trial demonstrated that the genotype-guided group was within the therapeutic INR range for 7% longer (67.4% vs 60.3%) than the control group that used an empiric dosing algorithm without incorporating any clinical factors. Like the COAG trial, no significance difference was observed in the occurrence of adverse bleeding events. However, the EU-PACT trial population was 98.5% of European ancestry, and the differences in ethnic composition of the two trials could in part explain the difference in the findings of the two trials. It is also possible that the success of the pharmacogenetic intervention in the EU-PACT trial was attributable to the other clinical factors used in the genetic dosing algorithm (age, height, weight, amiodarone use) that were not included in the dosing of the control group in the study.
The Genetic Informatics Trial (GIFT) was designed to study the incidence of warfarin-related bleeding and found a 3.9% reduction (10.8% vs 14.7% absolute difference) in the primary end point of major bleeding, INR of 4 or greater, venous thromboembolism, or death in the genotype-guided group compared to the clinically-guided group that based dosing only upon clinical factors (10). The genotype-guided group also experienced 3.4% absolute improvement in the percent of time within the therapeutic window, with the greatest improvement in patients who were genetically most sensitive to warfarin. Like the EU-PACT trial, the GIFT trial was composed of 91.0% of individuals of European ancestry.
Randomized controlled clinical trials provide the highest level of evidence to assess the utility of pharmacogenetic testing. Based upon the results of these studies, pharmacogenetic testing for warfarin can be helpful in determining the initial dose of warfarin for individuals of European ancestry. However, there is not yet available evidence to support pharmacogenetic testing for patients of African ancestry. Although pharmacogenetic testing helps to determine the initial dose of warfarin, INR must still be monitored with appropriate dose adjustments due to changes in diet and other factors that change over time.
Single-gene pharmacogenetic testing has clinical utility for a limited number of medications associated with high risk of adverse events, and the Food and Drug Administration labeling for a subset of these drugs specifies that pharmacogenetic testing is required (Table 2). Thiopurine methyltransferase (TPMT) genetic testing is used to screen for patients at risk for bone marrow toxicities, gastrointestinal symptoms, hepatotoxicity, pancreatitis, arthralgia, and rashes due to reduced ability to metabolize thiopurines (2, 11). The HLA allele (HLA-B*57:01) predicts hypersensitivity to the anti-HIV drug abacavir (2, 11). CFTR testing identifies individuals with the G551D CFTR mutation who should benefit from Ivacaftor (2). The majority of pharmacogenetic tests are currently performed in a targeted manner immediately prior to initiating treatment for a single specifically indicated medication. We provide key ‘take home points’ in Box 1 highlighting recommendations for warfarin and other drugs for which there is high quality evidence to support pharmacogenetic testing.
Box 1. Take Home Points.
Evidence is conflicting about the clinical utility of genotype-based warfarin therapy although there is supportive data to use pharmacogenetic testing in individuals of European ancestry. If genetic information is readily available, genotype-based warfarin therapy should be applied based on recommendations from the CPIC guidelines.
Warfarindosing.org provides a calculator where physicians can enter the patient’s genetic and clinical information to determine an appropriate warfarin dose.
Pharmacogenetic testing is not feasible for patients with an emergent need for warfarin due to long turnaround times for genetic testing.
Pharmacogenetics is not currently recommended when prescribing warfarin for individuals of African ancestry due to the lack of clinical studies for this population.
Evidence is available to support genotype-guided prescription for a limited number of drugs including abacavir, thiopurines, carbamazepine (Asian populations), allopurinol (Asian populations), and Ivacaftor but overall, there is still insufficient evidence to support genotype-guided therapy for most other medications with available pharmacogenetic tests.
In addition to high quality evidence, what infrastructure is necessary to support the use of pharmacogenetics to guide prescribing in clinical settings?
The clinical implementation of pharmacogenetics has been challenging, partly because of the need for infrastructure necessary for implementation. We provide a list of resources that can help physicians utilize pharmacogenetics in Box 2. The Food and Drug Administration has added genotype-based dosage recommendations to drug labels for some drugs. PharmGKB curates the knowledge to use pharmacogenetics when prescribing medications (3). PharmGKB and the Pharmacogenomics Research Network founded the Clinical Pharmacogenetics Implementation Consortium (CPIC) to advance the implementation of pharmacogenetics by publishing peer-reviewed, evidence-based practice guidelines for specific drugs. CPIC reviews the current state of research to assign each gene/drug pair a level of evidence linking genotype to phenotype (high, moderate, weak) and strength of recommendations for specific clinical actions (strong, moderate, optional). The genes/drugs are graded into A, B, or C with level A having the highest potential evidence for pharmacogenetic utility although clinical utility has not yet been demonstrated in all cases. CPIC has published guidelines for 15 genes and prescription recommendations for over 30 medications (2). The CPIC guidelines are currently limited by the lack of evidence showing improved clinical outcomes in patients with genotype-guided therapy for the majority of drugs, and therefore, the CPIC guidelines are not intended to justify clinical utility but to help guide the use of genotype information when the genotype is available. We summarize all level A gene/drug pairs in Table 2, providing their potential pharmacogenetic benefits and the current degree of evidence available to support clinical utility.
Box 2. Resources for the application of pharmacogenetics in clinical practice.
Food and Drug Administration table of pharmacogenetic biomarkers https://www.fda.gov/Drugs/ScienceResearch/ucm572698.htm
Provides a list of medications with pharmacogenetic biomarkers on the drug label. Some labels recommend specific actions to be taken based on the outcome of genetic testing.
PharmGKB https://www.pharmgkb.org/
Provides annotations of the CPIC guidelines and develops freely accessible resources to advance the implementation of pharmacogenetics in the clinic, such as summaries of genetic pathways and gene-drug information tables.
TPP Tables https://www.pharmgkb.org/page/tppTables
Tables that contains information on the genes currently implemented by TPP participating institutions and CDS information that can be accessed by providers.
Genetic Testing Registry https://www.ncbi.nlm.nih.gov/gtr/
Resource for ordering pharmacogenetic tests from specific labs.
CPIC Gene-Drug table https://cpicpgx.org/genes-drugs/
A list of the gene-drug pairs reviewed by the CPIC, including evaluations on the level of pharmacogenetic evidence and links to CPIC guidelines. The current table contains 127 unique genes and 223 unique drugs.
Warfarin dosing calculator http://www.warfarindosing.org/Source/Home.aspx
There is a need for real-time decision support in the electronic health record to facilitate implementation. Clinical decision support pop-ups to alert the physician of actionable genetic variants and provide dosage or alternative medication recommendations from guidelines will help providers make prescribing decisions. To facilitate the integration of pharmacogenetics into clinical practice, the Pharmacogenomics Research Network created the Translational Pharmacogenetics Program (TPP) with multigene array testing performed at baseline in individuals at multiple academic health care centers and incorporated results into the patient’s electronic health records for future drug prescribing decisions (13). As of June 2015, over 20,000 individuals have been tested through the Translational Pharmacogenetics Program with an average turnaround time of 14 days for preemptive testing. Nearly 100,000 test results have been added to patients’ electronic health records, and about one in four pharmacogenetic tests performed produced results that were actionable by the CPIC guidelines (14). Labs that ran the CYP2C19 test to assist with prescribing clopidogrel received an 85% reimbursement rate for outpatient claims (15). The Translational Pharmacogenetics Program has developed clinical decision support tables for the genes CYP2C19, SLCO1B1, TPMT, and CYP2D6.
The Genetic Testing Registry is a database of genetic tests available and the labs that perform them (https://www.ncbi.nlm.nih.gov/gtr/). There are currently 32 genes included in the pharmacogenetic tests listed in Genetic Testing Registry that can provide predictive information for 188 drug responses. The cost of pharmacogenetic testing ranges from $33 to $710 with a median value of $175 (16). From 2012 to 2013, genetic testing for the three CYP genes (CYP2D6, CYP2C19, CYP2C9) with the most significant influence on drug metabolism was billed to Medicare at an average cost of $260 per individual genetic test (17). Turnaround time for genetic testing ranges from 1–2 days to 4–6 weeks, and this is problematic in clinical situations in which patients require immediate therapy. Payers remain cautious about the economic benefit of universal testing of adults to inform possible prescriptions at an unknown future time until there is demonstrated clinical utility through randomized clinical trials. Although pharmacogenetic variants have been demonstrated to be an independent risk factor for hospitalization in older adults (≥65), it remains to be determined what average age of adults would be most cost effective and beneficial for universal testing (17, 18).
If testing is done, how should the results guide warfarin prescription?
The International Warfarin Pharmacogenetics Consortium (IWPC) has developed a consensus model for warfarin dosing. The IWPC and Gage et al. have published algorithms that incorporate clinical and genetic factors to calculate warfarin dosing (4, 19). The two algorithms are comparable and contain similar variables such as VKORC1 −1639G>A, CYP2C9*3, CYP2C9*2, height, weight, age, target INR, amiodarone, smoker status, and race. Both algorithms were created before significant evidence for African-specific polymorphisms was discovered. The collaboration between the creators of the two algorithms lead to the creation of http://www.warfarindosing.org, a website where physicians can enter a patient’s variables into the Gage and IWPC algorithms to predict the dose of medication to administer. More recent cohort studies showed that accounting for CYP2C9*5, *6, *8, *11 and rs12777823 leads to improvements in both the IWPC and Gage et al. algorithms (6,20). The CPIC guideline has incorporated recommendations for CYP2C9*5, *6, *8, *11 and rs12777823 polymorphisms.
The CPIC 2017 warfarin dosing guideline can be accessed at https://cpicpgx.org/content/guideline/publication/warfarin/2017/28198005.pdf. The guideline has strong recommendations for using either the Gage et al. or IWPC algorithm for warfarin dosing when genotype information for VKORC1, CYP2C9 *2 and *3 polymorphisms are available for patients of non-African ancestry, with the best evidence for patients of Europeans and East Asian descent (5). The online calculator warfarindosing.org allows physicians to estimate warfarin dosages for patients who are either beginning therapy or have taken warfarin for no more than five days. The website also includes dose refinement calculations based on existing INR. In addition to the factors in the original Gage et al. or IWPC algorithm, the calculator also accounts for GGCX rs11676382, a variant in gamma-glutamyl carboxylase that leads to a decrease in warfarin dosage. Furthermore, with the accumulation of evidence in support of African-specific alleles, warfarindosing.org has included CYP2C9*5 and *6 to the list of alleles that physicians can enter.
Should the patient described in this case undergo pharmacogenetic testing before initiation of warfarin?
The outcomes of the randomized controlled trials suggest that genetically-guided warfarin dosing may be most beneficial to those who already have genotype data available before receiving the initial dose of warfarin. The algorithms for genotype-based warfarin dosing are designed to help predict the initial treatment response for individuals who have yet to received more than five doses of warfarin. Due to the variable turnaround time for test results previously discussed, pharmacogenetic testing is best suited for patients who do not need to begin warfarin therapy emergently or who already have genotypes available and deposited in the electronic health record.
Because our patient does not have pharmacogenetic data available, it is not feasible to postpone initiation of warfarin treatment and initial dosing should be based upon available clinical variables alone. However, if genetic results were readily available and clinical decision support popups showed that the patient carries the AG genotype for VKORC1 and the *3/*3 CYP2C9 variant, then the patient’s internist should refer to the CPIC 2017 guideline, which strongly recommends the use of genotype-based warfarin dosing for VKORC1 and CYP2C9 variants in patients of non-African ancestry. After entering the genotype information and the relevant clinical variables (70-yrs- old, male, European descent, 5’9, 200, non-smoker, atrial fibrillation, atorvastatin, target INR 2.5) into the warfarindosing.org calculator, the Gage algorithm estimates a therapeutic dose of 2.1 mg/day and the IWPC estimates a dose of 1.5 mg/day. A mini-loading dose of 4.6 mg is also recommended to help the patient reach the therapeutic INR faster. The site recommends recalculating the patient’s INR after 3 warfarin doses to obtain an improved estimate.
Summary
Evidence demonstrating the clinical utility of pharmacogenetic testing is available for a small number of medications. Food and Drug Administration labels for a small number of drugs requires pharmacogenetic testing to avoid genetically associated adverse events or less often to identify molecular disease subtypes for which a medication has clinical utility. Randomized controlled clinical trial data are conflicting but support the clinical utility of pharmacogenetic testing for warfarin for individuals of European ancestry. The cost of genetic testing is decreasing, but the routine use of preemptive pharmacogenetic testing has yet to be widely adopted due to insufficient evidence to support routine use and insufficient infrastructure to support prescriber decision support for these complicated prescription algorithms.
Pharmacogenetics is currently routinely used for a small number of medications and in some cases for specific ethnic groups. Although pharmacogenetics is not currently routinely used by most practitioners, if there is evidence of clinical utility in the future, pharmacogenetic testing panels costs may increasingly be covered by Medicare and insurance. With payment for pharmacogenetic testing and as pharmacogenetic decision support is built into electronic health records, there may be increased utilization of pharmacogenetic information for certain medications in the future.
Acknowledgements
Not Applicable
References
- 1.Dunnenberger HM, Crews KR, Hoffman JM, Caudle KE, Broeckel U, Howard SC, et al. Preemptive clinical pharmacogenetics implementation: current programs in five US medical centers. Annu Rev Pharmacol Toxicol. 2015;55:89–106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Clinical Pharmacogenetics Implementation Consortium [Internet]. Pharm GKB and PGRN; c2009–2018; Guidelines; [Accessed 2018 Jul 18]; Available from: https://cpicpgx.org/guidelines/
- 3.Whirl-Carrillo M, McDonagh EM, Hebert JM, Gong L, Sangkuhl K, Thorn CF, et al. Pharmacogenomics knowledge for personalized medicine. Clin Pharmacol Ther. 2012;92(4):414–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Gage BF, Eby C, Johnson JA, Deych E, Rieder MJ, Ridker PM, et al. Use of pharmacogenetic and clinical factors to predict the therapeutic dose of warfarin. Clin Pharmacol Ther. 2008;84(3):326–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Johnson JA, Caudle KE, Gong L, Whirl-Carrillo M, Stein CM, Scott SA, et al. Clinical Pharmacogenetics Implementation Consortium (CPIC) Guideline for Pharmacogenetics-Guided Warfarin Dosing: 2017 Update. Clin Pharmacol Ther. 2017;102(3):397–404. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Perera MA, Cavallari LH, Limdi NA, Gamazon ER, Konkashbaev A, Daneshjou R, et al. Genetic variants associated with warfarin dose in African-American individuals: a genome-wide association study. Lancet. 2013;382(9894):790–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Kheiri B, Osman M, Abdalla A, Haykal T, Pandrangi PV, Chahine A, et al. CYP2C19 pharmacogenetics versus standard of care dosing for selecting antiplatelet therapy in patients with coronary artery disease: A meta-analysis of randomized clinical trials. Catheter Cardiovasc Interv. 2018. [DOI] [PubMed] [Google Scholar]
- 8.Kimmel SE, French B, Kasner SE, Johnson JA, Anderson JL, Gage BF, et al. A pharmacogenetic versus a clinical algorithm for warfarin dosing. N Engl J Med. 2013;369(24):2283–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Pirmohamed M, Burnside G, Eriksson N, Jorgensen AL, Toh CH, Nicholson T, et al. A randomized trial of genotype-guided dosing of warfarin. N Engl J Med. 2013;369(24):2294–303. [DOI] [PubMed] [Google Scholar]
- 10.Gage BF, Bass AR, Lin H, Woller SC, Stevens SM, Al-Hammadi N, et al. Effect of Genotype-Guided Warfarin Dosing on Clinical Events and Anticoagulation Control Among Patients Undergoing Hip or Knee Arthroplasty: The GIFT Randomized Clinical Trial. JAMA. 2017;318(12):1115–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Goulding R, Dawes D, Price M, Wilkie S, Dawes M. Genotype-guided drug prescribing: a systematic review and meta-analysis of randomized control trials. Br J Clin Pharmacol. 2015;80(4):868–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Shuker N, Bouamar R, van Schaik RH, Clahsen-van Groningen MC, Damman J, Baan C, et al. A Randomized Controlled Trial Comparing the Efficacy of Cyp3a5 Genotype- Based With Body-Weight-Based Tacrolimus Dosing After Living Donor Kidney Transplantation. Am J Transplant. 2016;16(7):2085–96. [DOI] [PubMed] [Google Scholar]
- 13.Shuldiner AR, Relling MV, Peterson JF, Hicks JK, Freimuth RR, Sadee W, et al. The Pharmacogenomics Research Network Translational Pharmacogenetics Program: overcoming challenges of real-world implementation. Clin Pharmacol Ther. 2013;94(2):207–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Luzum JA, Pakyz RE, Elsey AR, Haidar CE, Peterson JF, Whirl-Carrillo M, et al. The Pharmacogenomics Research Network Translational Pharmacogenetics Program: Outcomes and Metrics of Pharmacogenetic Implementations Across Diverse Healthcare Systems. Clin Pharmacol Ther. 2017;102(3):502–10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Weitzel KW, Elsey AR, Langaee TY, Burkley B, Nessl DR, Obeng AO, et al. Clinical pharmacogenetics implementation: approaches, successes, and challenges. Am J Med Genet C Semin Med Genet. 2014;166C(1):56–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Verbelen M, Weale ME, Lewis CM. Cost-effectiveness of pharmacogenetic-guided treatment: are we there yet? Pharmacogenomics J. 2017;17(5):395–402. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Keeling NJ, Rosenthal MM, West-Strum D, Patel AS, Haidar CE, Hoffman JM. Preemptive pharmacogenetic testing: exploring the knowledge and perspectives of US payers. Genet Med. 2017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Finkelstein J, Friedman C, Hripcsak G, Cabrera M. Pharmacogenetic polymorphism as an independent risk factor for frequent hospitalizations in older adults with polypharmacy: a pilot study. Pharmgenomics Pers Med. 2016;9:107–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.International Warfarin Pharmacogenetics C, Klein TE, Altman RB, Eriksson N, Gage BF, Kimmel SE, et al. Estimation of the warfarin dose with clinical and pharmacogenetic data. N Engl J Med. 2009;360(8):753–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Drozda K, Wong S, Patel SR, Bress AP, Nutescu EA, Kittles RA, et al. Poor warfarin dose prediction with pharmacogenetic algorithms that exclude genotypes important for African Americans. Pharmacogenet Genomics. 2015;25(2):73–81 [DOI] [PMC free article] [PubMed] [Google Scholar]