Two patients are admitted to different hospitals for treatment of deep vein thrombosis. One receives warfarin, with the dosage based on her age and weight. The other also is prescribed warfarin, but the treating physician determines the dosage with an algorithm that uses data about the patient’s genetic profile and activity of the cytochrome 2C9 enzyme and the VKORC1 enzyme (vitamin K epoxide reductase complex subunit 1); both enzymes metabolize warfarin.
The first patient suffers several bleeding episodes because her coagulation can’t be stabilized; she requires a transfusion and is hospitalized for 10 days. The second patient remains free from bleeding episodes and is discharged 3 days later.
Like most nurses, you’re probably aware that the usual dosage of a drug may turn out to be considerably too high or too low for a particular patient; a drug might prove to be ineffective with a certain patient; or a patient may suffer an
unexpected adverse drug effect. Chances are, you also know you can’t always predict whether a drug or a particular dosage will turn out to be harmful. Millions of adverse drug reactions occur each year in the United States. A 2002 study found that 10.3% of outpatients had been prescribed a drug with a “black box” warning, indicating the potential for a serious or even fatal adverse reaction.
Many drugs are considered effective only in a certain percentage of the population, estimated by one source at 25% to 60% and by another at 50% to 75% of patients. For the entire range of commonly prescribed drugs, the efficacy rate varies even more. Cyclo-oxygenase 2 (COX-2) inhibitors, such as celecoxib (Celebrex), are about 80% effective for the general population, while oncology drugs are only about 25% effective.
Unpredictability and lack of efficacy aren’t limited to new drugs, as the opening scenario illustrates. Warfarin, one of the most common drugs, has been used since the 1950s. Yet it’s still notoriously difficult to prescribe and often causes unpredictable effects. What’s more, individual dosing requirements for this drug vary greatly among patients.
Of course, many drugs work well for many people. Nonetheless, despite rigorous development and testing, some fail to produce the intended effects in certain individuals, and some can even cause harm.
Interplay between genetics and drugs
Why are drugs sometimes ineffective, harmful, or both? We know that age, gender, diet, liver and kidney status, the specific disease being treated (as well as its intensity and duration), concomitant drugs, and environmental factors can influence how well a drug works, what adverse effects are most likely to occur, and which patients a drug is likely to help.
But a growing body of evidence suggests genetics might be one of the most important factors in determining who will and who won’t benefit from a particular drug, how much to prescribe, and what adverse effects to expect. By some estimates, genetics may account for 20% to 95% of the variability in how the body handles drugs and what drug effects occur. Evidence supports a genetic predisposition for both expected and unexpected adverse reactions. One major study found incorrect dosing accounted for 42% of adverse reactions while genetic factors caused approximately 50%.
The science of pharmacogenetics focuses on identifying and predicting which drugs are likely to help and which are likely to harm a patient, as well as the proper dosage to prescribe. Researchers in this field study variations in deoxyribonucleic acid (DNA) and ribonucleic acid characteristics as they relate to drug response.
CYP450 enzyme system
To grasp the basics of pharmacogenetics, you need to understand drug metabolism—specifically the cytochrome P450 (CYP450) enzyme system. CYP450 enzymes are the most important drug-metabolizing enzymes, and the CYP450 enzyme system is the most important system affecting drug metabolism. Other genetic factors (not discussed in this article) also affect drug transport proteins, drug absorption, drug receptors, and drug excretion. (See CYP450 enzymes: What and where they are. by clicking the PDF icon above.)
Polymorphism and drug response
Wide individual variations exist in the expression and function of CYP450 enzymes. Research shows genetic variation, or polymorphism, in these enzymes is one of the most important causes of variable drug response. These variations result from single nucleotide polymorphisms (SNPs).
The nucleus of every cell contains 23 pairs of chromosomes, which are made up of DNA strands. In turn, DNA consists of nucleotides composed of bases (adenine, cytosine, guanine, and thymine). Normally, these bases occur in specific sequences.
The gene is the DNA segment responsible for directing the production of a specific protein. When the normal sequence of bases changes (that is, when SNP occurs), the gene’s function alters—and so does the function of the protein for which that gene is responsible. In this case, the protein is part of a CYP450 enzyme. A specific gene is associated with and encodes each CYP450 enzyme.
Polymorphism can determine how well a CYP450 enzyme metabolizes a drug—or whether it metabolizes it at all. Depending on the genetic makeup of the CYP450 enzyme system a person has inherited, the metabolic activity of a particular CYP450 enzyme can be categorized as poor, intermediate, extensive, or ultra-rapid. Correspondingly, some people are considered normal metabolizers of drugs, others are intermediate metabolizers, still others are poor metabolizers (experiencing no clinical effect from the drug), and some are ultra-rapid metabolizers (experiencing toxic drug effects).
Polymorphism and drug interactions
CYP450-enzyme polymorphism is crucial not just in terms of how it affects drug performance but in how drugs affect the enzymes. A particular drug may inhibit or induce (enhance) the metabolic activity of a certain CYP450 enzyme; if that enzyme is involved in metabolizing a second drug the patient takes, the second drug may not work. In other words, one drug’s effect on CYP450 enzymes may disrupt another drug’s metabolism. When enzymatic activity is inhibited, a drug’s concentration may rise to toxic levels, and more numerous and more serious adverse effects may occur. Conversely, when enzymatic activity is induced, drug concentration may decrease, severely limiting drug efficacy.
Not just an abstract science
Pharmacogenetics may sound abstract and theoretical, but the science is becoming well established. Clearly, genetics and drug performance are inseparable. For a small minority of drugs, pharmacogenetics already influences prescribing practices. (See Examples of personalized drug therapy by clicking the PDF icon above.)
The Food and Drug Administration (FDA) now requires drug manufacturers to include information about a drug’s effects on the CYP450 enzyme system and how genetic variants of this system influence the drug’s metabolism and adverse effects. Currently, labeling for more than 50 drugs includes pharmacogenetic information. The FDA also has issued guidelines for using genetic polymorphism information in clinical trials of drugs.
Genetically based dosing algorithms
Evidence from randomized controlled trials suggests that dosing algorithms based on genetic information can be valuable and may improve patient outcomes. Researchers predict drug prescribing will soon be based partly on genetic profiling. Given that differences in an SNP may produce a tenfold variation in blood drug concentrations, pharmacogenetics appears to have great potential to improve prescribing practices and help avoid adverse effects.
Yet, although we know that genes and pharmacology are inextricably linked, we don’t know how, when, and in what ways. More importantly, it’s not yet possible to use pharmacogenetic knowledge in a practical way. The FDA doesn’t require genetic testing for patients starting warfarin therapy. Pharmacogenetic algorithms and pharmacogenetic testing aren’t used in everyday clinical practice. And no randomized trials have proven unequivocally that this dosing approach has value. In a May 2009 memo, the Centers for Medicare & Medicaid Services stated it won’t pay for genetic testing to determine warfarin dosing for Medicare recipients because no evidence shows that pharmacogenomic testing to predict warfarin response improves health outcomes. (See Pharmacogenetic testing: Not ready for prime time by clicking the PDF icon above.)
Progressing toward personalized drug therapy
In the simplest sense, the goal of pharmacogenetics is to understand the effects of genetics on drug response. If this can be done, drug inefficacy and adverse effects could be predicted and avoided, and appropriate drugs could always be prescribed in the proper dosages.
Understanding the effects of genetic variations of CYP450 enzymes on drug metabolism is a vital part of this pursuit. Researchers have known for more than 40 years that inherited variations in drug metabolism can profoundly affect a patient’s drug response. Consider these examples:
- Impaired metabolism of the muscle relaxant succinylcholine can lead to profound adverse effects.
- Phenformin was developed in the 1970s as an option for treating noninsulin-dependent diabetes mellitus. But an unacceptably high percentage of patients (1 in 4,000) developed metabolic lactic acidosis, which has a mortality of 50% to 70%. In 1977, phenformin was banned in this country; research later showed a powerful genetic polymorphism caused a defect in its metabolism.
Although the concept of personalized drug therapy is attractive, the literature indicates that reaching this goal is far from simple. Prescribing drugs based on genetic knowledge would require an understanding of the genetic basis of and influence on the disease or disorder being treated. It would entail development of tests that provide fast, accurate genetic information about drug metabolism—and this information would need to be easy to interpret and clinically useful. It also would require comprehensive knowledge of how genetics influences all aspects of pharmacokinetics—drug absorption, distribution, metabolism, and elimination.
Genetics also can influence the drug target in the body (pharmacodynamics). Drugs work by affecting the activity of enzymes or receptors. For instance, genetic polymorphisms exist for the beta1-adrenergic receptor gene ADRB1 that influence a patient’s response to beta blockers. This may explain why African-Americans, who have a lower frequency of the Arg389 allele that affects the body’s reaction to beta blockers, don’t respond to beta-blocker therapy as well as whites. Clearly, synthesizing pharmacodynamic, pharmacokinetic, and pharmacogenomic information and translating this knowledge into clinical practice pose a huge challenge.
The psychotropic drug example
A review of pharmacogenetic information on psychotropic drugs illustrates this difficulty. CYP450 enzymes (especially CYP2D6) are important metabolizers of virtually all selective serotonin reuptake inhibitors (SSRIs), which are used to treat major depressive disorder. These enzymes also are important metabolizers of many other drugs used to treat psychotic disorders, schizophrenia, and other types of depression. Polymorphism of these enzymes can affect the metabolism of such psychotropic drugs as aripiprazole, haloperidol, and risperidone.
But most data on the effects of CYP450 polymorphism and metabolism of these drugs come from animal studies, in vitro studies, single-dose pharmacokinetic studies, or drug-interaction studies. Studies of CYP2D6 polymorphism and extrapyramidal side effects and movement disorders (for instance, tardive dyskinesia and parkinsonism) linked to certain psychotropics have shown contradictory results. And although CYP450 enzymes metabolize SSRIs, a comprehensive review failed to find a strong link between genetic variants in these enzymes and differences in SSRI efficacy or tolerability. Tests used to detect these polymorphisms were sensitive and accurate, but the authors concluded the data weren’t clinically useful. Other researchers drew the same conclusion after reviewing the pharmacogenetic response to risperidone and other atypical antipsychotics. They speculated that drug-metabolizing enzyme activity may play only a minor role in the clinical response to these drugs.
One goal of pharmacogeneticists is to develop the ability to quickly and accurately determine a patient’s genotype, which would allow detection of all clinically important CYP450-enzyme polymorphisms. Right now, this isn’t possible.
The FDA-approved AmpliChip CYP450 test can detect many CYP2D6 and CYP2C19 polymorphisms. These two enzymes metabolize approximately 25% of all commonly used drugs, including many antidepressants, beta blockers, analgesics, anticonvulsants, benzodiazepines, and proton-pump inhibitors. The test can predict if a patient will be a poor, intermediate, extensive, or ultra-rapid metabolizer of CYP2D6 and CYP2C19. Although AmpliChip holds promise in helping to prevent adverse drug reactions, it can only predict—not confirm—a patient’s metabolizing status. And despite its 99% sensi-tivity and 100% specificity, some researchers note that studies of its effectiveness have been poorly designed and lack statistical power.
Pharmacogenetic drug testing
Some drugs have been relabeled (with FDA input) to include pharmacogenetic information, but this information isn’t specific enough to guide treatment recommendations. FDA-approved pharmacogenetic tests exist for only three drugs—voriconazole, atomoxetine, and irinotecan. And these tests predict increased plasma drug concentrations that may be linked to an increased incidence of adverse effects—nothing else.
Obviously, the most serious issues surrounding pharmacogenetic testing—clinical validity and utility—are still far from resolved. What’s more, other factors, such as concomitant medications, age, tobacco use, and diet, can affect drug metabolism, too. These variables haven’t been completely factored into pharmacogenetic testing for drug metabolism and SNPs. Although much work has been done, no studies have used pharmacogenetic information to adjust drug dosages.
In addition, quick and accurate testing that provides reliable, clinically valid and useful information is lacking. For example, the AmpliChip test can accurately identify breastfeeding mothers who are rapid codeine metabolizers. These women excrete larger amounts of morphine (the codeine metabolite) in breast milk, which can adversely affect breastfeeding infants. But evidence that this genetic variation can be used as a clinical tool to avoid adverse effects in infants is limited.
So despite the large amount of pharmacogenetic information available, scientists aren’t sure how to use it. The FDA requires genetic testing only for four drugs—cetuximab, trastuzumab, maraviroc, and dasatinib—before therapy can begin. Such testing is recommended but not required for carbamazepine, valproic acid, and mercaptopurine. For other drugs, such as tretinoin and isoniazid, information about genetic variations and drug response is available, but we don’t know how to use it.
What types of tests could be used to make pharmacogenetic information applicable to clinical situations? And who should perform them? Only six FDA-approved pharmacogenetic tests are available to check for genetic variations, and they address just five drugs. Yet more than 1,300 tests that don’t require FDA approval are available for genetic testing. These tests are complex to perform, but laboratories aren’t required to demonstrate skill in this area. Also, few laboratories offer pharmacogenetic testing for clinicians, and turnaround time for results may be too long. Finally, although some tests may be covered by insurance plans, many are considered experimental and aren’t covered; few patients could afford to pay for them out of pocket.
As our knowledge of the interplay between genetics and drug response grows, nurses will need to be familiar with this information. The Consensus Panel on Genetic/Genomic Nursing Competencies recommended in its 2006 monograph that all nurses should understand the relationship of genetics to treatment selection and monitoring of treatment effectiveness, and should be able to identify patients who may benefit from genetic or genomic assessment. This means nurses will need to learn which patients may benefit from genetic testing. (See What to teach patients about genetic tests by clicking the PDF icon above.)
Making the ideal a reality
Ideally, drugs would produce only the desired therapeutic action, proper dosages could be calculated easily and precisely, and no adverse effects would occur. Such personalized drug therapy would avoid adverse effects or greatly decrease their incidence. Less time would be wasted finding the optimal dosage, and no time would be wasted giving drugs that would prove ineffective.
No one can predict when the science of pharmacogenetics will turn this ideal into reality. Some progress has been made, and for a handful of drugs, that ideal may soon be realized. But assessing the multitude of genetic variations that affect pharmacokinetics and pharmacodynamics is highly complex, and drug response involves additional factors that interact with a person’s genetic makeup.
Making clinically useful pharmacogenetic testing a reality will involve an enormous amount of time and research. The ideal of always being able to give the right drug in the right dosage to the right patient appears to be many years away.
Carey L. Pharmacogenetics: what PAs need to understand and why. JAAPA. 2009;22(7):61-62.
Consensus Panel on Genetic/Genomic Nursing Competencies. Essential nursing competencies and curricula guidelines for genetics and genomics. Silver Spring, MD: American Nurses Association; 2006. www.nursingsociety.org/Global
Action/Initiatives/Documents/geneticscompetency.pdf. Accessed March 21, 2011.
Crowley JJ, Sullivan PF, McLeod HL. Pharmacogenomic genome-wide association studies: lessons learned thus far. Pharmacogenomics. 2009;10:161-163.
Fleeman N, Dickson R. Providing patients with information on pharmacogenetic testing. Nurs Stand. 2009;23(21):46-48.
Haack S, Seeringer A, Thürmann PA, Becker T, Kirchheiner J. Sex-specific differences in side effects of psychotropic drugs: genes or gender? Pharmacogenomics. 2009;10(9):1511-1526.
Lynch T, Price A. The effect of cytochrome P450 metabolism on drug response, interactions, and adverse effects. Am Fam Physician. 2007;76:391-396.
Pestka EL, Hale, A, Johnson BL, Lee JL, Poppe KA. Cytochrome P450 testing for better psychiatric care. J Psychosoc Nurs Ment Health Serv. 2007;45:15-18.
van Schie RMF, Wadelius M, Kamali F, et al. Genotype-guided dosing of coumarin derivatives: the European pharmacogenetics of anticoagulant therapy (EU-PACT) trial design. Pharmacogenomics. 2009;10(1):1287-1295.
Dana Bartlett is a certified specialist in poison information at the Philadelphia Poison Control Center at Children’s Hospital in Philadelphia, Pennsylvania. The author and planners of this CNE activity have disclosed no relevant financial relationships with any commercial companies pertaining to this activity.