Secondary Logo

Journal Logo

AAPA Members can view Full text articles for FREE. Not a Member? Join today!

Using pharmacogenetics in primary care

Murfin, Melissa PharmD, BCACP, PA-C

Journal of the American Academy of PAs: August 2019 - Volume 32 - Issue 8 - p 17–21
doi: 10.1097/
CME: Primary Care

ABSTRACT Pharmacogenetics offers a way to personalize medication prescribing for patients. Through the use of genetic tests that identify variations in enzymes important to drug metabolism, PAs can have patients' genetic information before prescribing a medication. This may reduce the risks of adverse reactions and lost treatment time when patients are given drugs to which they are unlikely to respond. Laboratory testing can identify common genetic variants that alter how the body metabolizes drugs. PAs with knowledge of these variants can choose medications that are more personalized and effective for each patient. Clinical pharmacogenetic guidelines are under development and will help providers identify which drugs are most likely to be affected by genetic variations so they can prescribe for patients based on their specific genetic phenotypes.

Melissa Murfin is chair and program director of the PA program at Elon (N.C.) University. The author has disclosed no potential conflicts of interest, financial or otherwise.

Earn Category I CME Credit by reading both CME articles in this issue, reviewing the post-test, then taking the online test at Successful completion is defined as a cumulative score of at least 70% correct. This material has been reviewed and is approved for 1 hour of clinical Category I (Preapproved) CME credit by the AAPA. The term of approval is for 1 year from the publication date of August 2019.



Box 1

Box 1

A 45-year-old man was prescribed paroxetine 4 weeks ago to treat generalized anxiety disorder. He has noticed some improvement in anxiety symptoms but expresses concern about adverse reactions he is experiencing, including insomnia, constipation, and dry mouth. He is not certain he wants to continue the medication due to the adverse reactions.

Clinicians are faced with situations like this whenever they prescribe medication. Will the drug work for the patient or will the patient be in that percentage who experience intolerable adverse reactions or fail treatment? The practice of drug selection has an element of trial and error: Clinicians choose medications based on evidence-based guidelines that have generally been developed from clinical trials in large populations. Although the technique may work well in most situations, some patients do not respond as expected. For example, about 25,000 patients are seen in the ED annually due to adverse reactions to selective serotonin reuptake inhibitors (SSRIs) such as paroxetine, and as many as 50% fail to respond to treatment with SSRIs.1

Enter precision, or personalized, medicine. Through simple laboratory testing, clinicians can identify patients with common genetic variations that can affect how they metabolize medications. Although prescribing based on genetic testing is common in oncology, it has not been applied to primary care thus far.

For patients who take frequently prescribed medications such as SSRIs and opioids, pharmacogenetic testing is available to identify common genetic variants that occur in specific racial populations. The FDA maintains a table of more than 120 drugs outside oncology that contain prescribing information instructions on potential genetic variations that could affect the medication.2 These variations can influence pharmacokinetic parameters such as metabolism or affect carrier proteins that influence how a drug is activated by or eliminated from the body, or how it binds to the appropriate receptor for its intended clinical effect.2

Box 2

Box 2

Pharmacogenetic testing that can identify some of these common variants offers clinicians potentially useful information about whether a medication will be inappropriate for a given patient. Clinicians can help a patient avoid adverse reactions or a medication that is less likely to work. This can improve prescribing in clinical practice, saving money and time for patients while offering better treatment outcomes.

Back to Top | Article Outline


The term pharmacogenetics often is used interchangeably with pharmacogenomics; however, there is a difference between the two. Pharmacogenomics is the broader term, and includes all the genes that code for medication-related activities in the body; the definition of pharmacogenetics is narrower, referring to activities coded by a single gene.3 Genetic testing can be ordered at many commercial laboratories to identify variations in genes responsible for determining the enzymes and processes that make up pharmacokinetic parameters in a patient. These polymorphisms occur at the gene base pair level. They are normal variants that promote genetic diversity, similar to the differences noted in human blood group types.4

Much of the focus of pharmacogenetics is on the cytochrome P (CYP) 450 enzymes responsible for drug metabolism. The CYP 450 family of isoenzymes is the main metabolic pathway for many drugs and a significant mechanism for drug interactions.5 These enzymes detoxify the body by converting substrates such as medications into inactive molecules that are easily eliminated. The 2C9 and 2C19 isoenzymes are important in the metabolism of anticoagulants.5 Several CYP 450 isoenzymes, such as 3A4 and 2D6, metabolize common drugs including SSRIs, calcium channel blockers, opioids, and most macrolides. Impairment of this process can lead to drug toxicity when medications such as clarithromycin and erythromycin, which are strong inhibitors of 3A4, are given with other medications also processed by the CYP 3A4 isoenzyme.5

Toxicity is a major concern with changes in metabolic enzymes; however, the other side of the coin involves prodrugs such as clopidogrel. Prodrugs initially are inactive metabolites; they must be metabolized in the body to engage their mechanism of action.4,6 This process often is driven by CYP 450 enzymes. Drugs that are CYP 450 inhibitors can block this process, reducing the efficacy of a prodrug.6

Genetically, metabolic enzyme activity is coded by two alleles in each person, one inherited from the mother, one from the father. This pair of alleles composes the offspring's genotype. Each allele codes for function of the specific CYP 450 enzyme and can be nonfunctional or have reduced, normal, or increased function.7 The genotype creates the genetic phenotype (or observable expression of the genetic trait) of metabolic activity for each person.7

The phenotype expressed results in a number of possible combinations of metabolic enzyme activity that can occur in the patient, depending on which alleles were inherited. Phenotype is determined by an overall activity score given when the scores of each allele are added.7 Each allele has an activity score ranging from 0 to 1 depending on how the enzyme functions.7 The scores for individual allele function are added to produce an overall activity score for the patient's enzyme function. These activity scores range from 0 to 2; however, some patients with genetic variations that result in allele duplications may have a score that is greater than 2.7

Alleles also are designated by a different set of numbers, with 1 generally being the normal function (or natural form, also called wild-type) and other numbers delineating different levels of enzyme activity.8 Some CYP 450 enzymes such as 2C19 only have a few variations; others, such as the 2D6 enzyme that metabolizes SSRIs, tamoxifen, and opioids, have more than 100 different alleles that can code for enzyme activity.9 Pharmacogenetic laboratory results are reported with the patient's genotype, which identifies each allele with a star and a number.8 For example, in most genetic tests, a normal genotype will be *1/*1 with each *1 representing one of the two alleles the patient carries.8

Back to Top | Article Outline


Although genotypes are helpful information, the process of prescribing medication based on pharmacogenetics is largely based on the patient's phenotype. The phenotype is determined by the combined enzyme activity score predicted by each allele that the patient inherited. Depending on the enzyme combination, patients' metabolic activity can be described as one of four normal variants.1

  • Ultrarapid metabolizer. These patients can have two increased-function alleles, a normal allele and an increased-function allele, or duplicated alleles.1 In these patients, metabolism may proceed quickly and too efficiently, inactivating drugs before they have a chance to work.8 Ultrarapid metabolizers may actually have toxicity to prodrugs because the conversion process can create a higher concentration of the active metabolite.8 The number of people with this phenotypic variant differs according to enzyme: CYP 2D6 is found in up to 2% of patients and 2C19 can be found in as many as 30% of patients.1
  • Extensive metabolizers. This is the most common phenotype (two normal-function alleles or one normal- and one reduced-function allele). Patients are considered to have normal drug metabolism that is not impaired or more rapid than expected.6 Depending on the enzyme, up to 92% of patients will be extensive metabolizers.1
  • Intermediate metabolizers. About 45% of patients fall into this category, depending on the enzyme.1 Phenotypically, they exhibit lower metabolic efficiency although this may not always be clinically relevant. These patients have one reduced-function allele with a normal or a nonfunctional allele.
  • Poor metabolizers. Patients in this category may have two reduced-function alleles, two nonfunctional alleles, or one of each.1,9 Poor metabolizers are likely to exhibit clinically relevant changes in drug metabolism. For most drugs, poor metabolizers will be unable to effectively inactivate the medications, leading to possible toxicity as the drug builds up in their system.8 A prodrug would lead to the opposite problem: The lack of an efficient metabolic process could reduce the amount of active metabolite that is formed, rendering the drug ineffective.6 This may represent up to 15% of patients for some CYP 450 enzymes.1

These categories describe general activity of enzymes that are clinically important in drug metabolism. Though CYP 3A4 often is implicated in drug interactions, it is thought to be less clinically relevant in terms of pharmacogenetic concerns. Clinicians should be aware of clinical issues that occur in patients with metabolic variants in 2D6, 2C19, and 2C9. Evidence-based guidelines are available for a number of drugs to help prescribers make decisions about dosing or use of medications in these patients.10

Back to Top | Article Outline


Few guidelines incorporate pharmacogenetic testing recommendations. Some reasons for this include lack of data on prescribing outcomes with pharmacogenetic testing, lack of consensus opinion on pharmacogenetic testing recommendations, and lack of prescriber education on the use and interpretation of genetic testing.4

The Clinical Pharmacogenetic Implementation Consortium (CPIC) is a US-based group that is working to help clinicians by developing evidence-based pharmacogenetic guidelines for medication choice and dosing.10 The guidelines help clinicians interpret the results of pharmacogenetic testing and prescribe medications based on those results. The CPIC guidelines also define what is a clinically actionable phenotype that would alter prescribing compared with one that is just informative.10 Commonly used medications with clinically actionable guidelines include clopidogrel, SSRIs, tricyclic antidepressants, and opioids.

Back to Top | Article Outline


This drug provides one of the most dramatic examples of clinical issues that can occur with pharmacogenetic variants in drug metabolism. Patients who are poor metabolizers at 2C19 have a reduced ability to adequately metabolize this prodrug to its active form, leading to lower antiplatelet activity and the potential for stent restenosis after percutaneous coronary intervention (PCI).11 Up to 15% of patients of Asian descent have this phenotype (poor metabolization at 2C19), nearly three times as many as in white and black populations.11

The native Hawaiian population is 42% East Asian, 24% white, and 10% Pacific Islanders.12 Native Hawaiian patients treated with clopidogrel after PCI showed nearly double the risk of acute myocardial infarction compared with white patients in Hawaii.12 In 2014, the state's attorney general sued the drug manufacturers about their marketing of clopidogrel in the state, charging that the drug company was aware of this potential risk but marketed the drug without sharing that information with clinicians.12

This situation raises a liability question for clinicians when a pharmacogenetic test is available but is not ordered before a medication is prescribed. Testing for CYP 2C19 is now listed in the clopidogrel prescribing information black box warning, and clinicians are advised to use an alternative antiplatelet drug in patients who are poor metabolizers.13 CPIC guidelines recommend prasugrel or ticagrelor instead of clopidogrel for patients with acute coronary syndrome after PCI.11

Back to Top | Article Outline


Many psychiatric medications undergo metabolism through the CYP 450 pathway. SSRIs such as paroxetine and sertraline frequently are used to treat patients in primary care and behavioral medicine. This medication class may be heavily affected by pharmacogenetic variants in CYP 450. Pharmacogenetic testing before a patient starts an SSRI could help guide therapy to avoid adverse reactions. Also, because symptom improvement may take 4 or more weeks, pharmacogenetic testing could save patients from treatment with a medication that is ineffective for them.

SSRIs are primarily metabolized by two of the CYP enzymes, 2C19 and 2D6. Although 2C19 has only four to five clinically relevant polymorphic alleles, 2D6 is known to have more than 100.1 Up to 2% of patients are ultrarapid metabolizers at 2D6 and up to 30% at 2C19; 10% of patients are poor metabolizers at 2D6 and 15% at 2C19.1 Studies indicate an association between the ultrarapid metabolizer phenotype and suicide; however, whether this is related to medication treatment failure is unclear.14

Paroxetine is significantly metabolized by 2D6, leading to a potential for greater effect in a patient who has a 2D6 metabolic variation. Someone who is an ultrarapid metabolizer at 2D6 may find paroxetine ineffective because the drug is largely metabolized before it has a chance to work. On the other hand, a poor metabolizer would be more likely to see adverse reactions secondary to increased medication levels. CPIC guidelines recommend avoiding paroxetine in ultrarapid metabolizers and avoiding or starting with half the recommended dose in poor metabolizers.1

Citalopram and escitalopram are similar to paroxetine, but primarily metabolized by 2C19. For patients who are ultrarapid metabolizers, an alternate medication is recommended because of the potential for low drug levels; patients who are poor metabolizers should be offered another drug or started at half the recommended dose.1

Sertraline is somewhat different due to a lesser effect on metabolism at 2C19. Dosing recommendations for ultrarapid metabolizers suggest starting at the recommended dose of sertraline and changing the medication if the patient does not respond. A lower initiation dose is recommended in poor metabolizers because of the increased incidence of adverse reactions in this population.1

Tricyclic antidepressants such as nortriptyline and amitriptyline also are metabolized in part by 2D6; in addition, amitriptyline, imipramine, and doxepin are primarily metabolized by 2C19. Avoid tricyclic antidepressants in patients who are 2D6 and 2C19 ultrarapid metabolizers because of the potential ineffectiveness.9 Patients who are poor metabolizers are at greater risk for adverse reactions and overdose, and should avoid tricyclic antidepressants or be prescribed a 50% dose initially along with drug level monitoring.9 Patients treated for neuropathic pain with low doses of tricyclic antidepressants may be less likely to experience adverse reactions but should be closely monitored.9

Back to Top | Article Outline


These drugs also are processed by 2D6. The only guideline for clinically relevant opioid-related pharmacogenetics is for codeine, a prodrug that is converted by 2D6 to morphine, which provides the majority of the analgesic effect.15 This can be problematic for patients who are ultrarapid metabolizers, because the conversion process can lead to excess levels of morphine, toxicity, and respiratory depression. The FDA issued a safety announcement after several cases of respiratory issues in breastfeeding infants were noted. The mothers had been given prescriptions for codeine cough syrup and although they had no issues, some of their infants exhibited respiratory changes related to the infants' ultrarapid metabolism of the drug.16 The FDA now recommends against the use of codeine in breastfeeding women.16 These issues have not been seen with tramadol, but the potential is there because of its similar metabolic process.

Other opioids, such as hydrocodone, oxycodone, and methadone, also are prodrugs metabolized by 2D6. Morphine and hydromorphone are less likely to be a concern because they are metabolized by a glucuronidation pathway that is independent of CYP 450 pharmacogenetics. No guidelines exist on the clinical relevance of the ultrarapid or poor metabolizer phenotypes with these drugs. Although ultrarapid metabolizers may exhibit increased toxicity with these prodrugs, the poor metabolizers may actually note reduced efficacy because of lowered levels of the active analgesic metabolite.17 This may lead to an increase in dosing for adequate analgesia in these patients. Up to 34% of black patients may have either rapid or poor drug metabolism.18 Poor metabolizers also are more likely to experience postoperative pain and dissatisfaction with methadone treatment for heroin addiction.17,19

Back to Top | Article Outline

Other drugs

Additional guidelines address some medications that have pharmacogenetic concerns separate from CYP 450 metabolism. A variant in the SLCO1B1 gene creates the potential for increased uptake of simvastatin.20 Intermediate and poor metabolizers taking simvastatin have a 2.6 to 5.2 increased risk of myopathy compared with patients who are normal metabolizers.20

Allopurinol also has pharmacogenetic implications, with an increased risk of severe cutaneous reactions such as Stevens-Johnson syndrome in patients who carry the HLA-B58*01 gene.21 Although the overall risk of these reactions is low, patients who experience them have a high likelihood of carrying this gene. Significant mortality (25%) is associated with these skin reactions.21 As a result, allopurinol is contraindicated for patients who carry this gene.21

Back to Top | Article Outline


Despite the many advantages for patients and clinicians, some barriers to pharmacogenetic testing remain. One of the most difficult to overcome is the cost of testing. Direct costs vary according to the laboratory performing the test; estimates in the United States are from $33 to $710, which may not be covered by insurance.22 More studies of patient treatment based on pharmacogenetic data are needed to determine whether the cost of the testing is offset by the savings from avoiding adverse reactions and treatment failures.

Another barrier is incorporation into the electronic medical record (EMR), although some EMR systems suggest pharmacogenetic testing for specific medications. Some healthcare systems are moving in that direction, but the process takes time. One issue is where to store the results so that clinicians know where to find them. Because testing is only needed once in a lifetime, a patient who had a test several years ago for one drug may not have the results readily available when another medication is prescribed.

A final barrier is education. Few clinicians have had extensive training in pharmacogenetic testing or its interpretation, largely due to the newness of the field. Clinicians may not be aware that testing is available or may not feel comfortable ordering a test when they are unsure of the interpretation and benefit to the patient, or are concerned about cost. For a list of pharmacogenetic resources, see Table 1.



Back to Top | Article Outline


Pharmacogenetics has the potential to revolutionize medication treatment. The benefit to patients in terms of reducing adverse reactions and personalizing prescribing to avoid ineffective medications cannot be underestimated. Many drugs can be affected by genetic polymorphisms in metabolic processes, which can affect how well the drug works for the patient. Using pharmacogenetic testing has a significant potential to improve treatment, particularly in areas of behavioral medicine and pain management. Costs of testing are decreasing, with many tests marketed to consumers, though a clinician still must write an order for the test. EMR use and clinician education are current barriers that can be overcome with additional training and focusing on how best to define testing in an EMR.

The PA profession could lead this charge into personalized medicine. Genomic competencies for PAs were updated in 2016 and include a focus on the incorporation of genetic testing into clinical practice.23 By developing greater knowledge in this area, PAs could be well positioned to become experts, improving patient outcomes and reducing cost of treatment.

Back to Top | Article Outline


1. Hicks JK, Bishop JR, Sangkuhl K, et al Clinical Pharmacogenetics Implementation Consortium (CPIC) guideline for CYP2D6 and CYP2C19 genotypes and dosing of selective serotonin reuptake inhibitors. Clin Pharmacol Ther. 2015;98(2):127–134.
2. US Food & Drug Administration. Table of pharmacogenomic biomarkers in drug labeling. Accessed April 2, 2019.
4. Weinshilboum RM, Wang L. Pharmacogenomics: precision medicine and drug response. Mayo Clin Proc. 2017;92(11):1711–1722.
5. Indiana University School of Medicine. Drug interactions Flockhart table. Accessed April 2, 2019.
6. Lynch T, Price A. The effect of cytochrome P450 metabolism on drug response, interactions, and adverse effects. Am Fam Physician. 2007;76(3):391–396.
7. Wu AH. Drug metabolizing enzyme activities versus genetic variances for drug of clinical pharmacogenomic relevance. Clin Proteomics. 2011;8(1):12.
8. Chang KL, Weitzel K, Schmidt S. Pharmacogenetics: using genetic information to guide drug therapy. Am Fam Physician. 2015;92(7):588–594.
9. Hicks JK, Sangkuhl K, Swen JJ, et al Clinical Pharmacogenetics Implementation Consortium guideline (CPIC) for CYP2D6 and CYP2C19 genotypes and dosing of tricyclic antidepressants: 2016 update. Clin Pharmacol Ther. 2017;102(1):37–44.
10. Clinical Pharmacogenetics Implementation Consortium. Genes-drugs. Accessed April 2, 2019.
11. Scott SA, Sangkuhl K, Stein CM, et al Clinical Pharmacogenetics Implementation Consortium guidelines for CYP2C19 genotype and clopidogrel therapy: 2013 update. Clin Pharmacol Ther. 2013;94(3):317–323.
12. Wu AH, White MJ, Oh S, Burchard E. The Hawaii clopidogrel lawsuit: the possible effect on clinical laboratory testing. Accessed April 2, 2019.
13. Apotex Corp. Clopidogrel prescribing information. Revised October 2018.
14. Eap CB. Personalized prescribing: a new medical model for clinical implementation of psychotropic drugs. Dialogues Clin Neurosci. 2016;18(3):313–322.
15. Crews KR, Gaedigk A, Dunnenberger HM, et al Clinical Pharmacogenetics Implementation Consortium guidelines for cytochrome P450 2D6 genotype and codeine therapy: 2014 update. Clin Pharmacol Ther. 2014;95(4):376–382.
16. US Food and Drug Administration. Use of codeine and tramadol products in breastfeeding women—questions and answers. April 2017. Accessed April 1, 2019.
17. Agarwal D, Udoji MA, Trescot A. Genetic testing for opioid pain management: a primer. Pain Ther. 2017;6(1):93–105.
18. Smith HS. Opioid metabolism. Mayo Clin Proc. 2009;84(7):613–624.
19. Pérez de los Cobos J, Siñol N, Trujols J, et al Association of CYP2D6 ultrarapid metabolizer genotype with deficient patient satisfaction regarding methadone maintenance treatment. Drug Alcohol Depend. 2007;89(2–3):190–194.
20. Ramsey LB, Johnson SG, Caudle KE, et al The Clinical Pharmacogenetics Implementation Consortium Guideline for SLCO1B1 and simvastatin-induced myopathy: 2014 update. Clin Pharmacol Ther. 2014;96(4):423–428.
21. Hershfield MS, Callaghan JT, Tassaneeyakul W, et al Clinical Pharmacogenetics Implementation Consortium guidelines for human leukocyte antigen-B genotype and allopurinol dosing. Clin Pharmacol Ther. 2013;93(2):153–158.
22. Verbelen M, Weale ME, Lewis CM. Cost-effectiveness of pharmacogenetic-guided treatment: are we there yet. Pharmacogenomics J. 2017;17(5):395–402.
23. Goldgar C, Michaud E, Park N, Jenkins J. Physician assistant genomic competencies. J Physician Assist Educ. 2016;27(3):110–116.

pharmacogenetics; pharmacogenomics; personalized medicine; drug metabolism; phenotype; CYP 450

Copyright © 2019 American Academy of Physician Assistants