Review

Nature Reviews Drug Discovery 1, 37-44 (January 2002) | doi:10.1038/nrd705

The genetic basis of variability in drug responses

Dan M. Roden1 & Alfred L. George Jr1  About the authors

Top

It is almost axiomatic that patients vary widely in their beneficial responses to drug therapy, and serious and apparently unpredictable adverse drug reactions continue to be a major public health problem. Here, we discuss the concept that genetic variants might determine much of this variability in drug response, and propose an algorithm to enable further evaluation of the benefits and pitfalls of this enticing possibility.

Patients vary widely in their responses to drugs. This variability is seen, not only in beneficial responses, but also in adverse drug reactions, an increasingly recognized problem that extracts a huge toll in lives and in healthcare costs1, 2, 3, 4. Matching patients to the drugs that are most likely to be effective and least likely to cause harm is the main goal of modern therapeutics. Although a genetic contribution to variability in drug action has long been recognized, the sequencing of the human genome now offers a new opportunity for using genetic approaches to improve drug therapy5.

Historical development

The British physician Garrod, who coined the term “inborn error of metabolism” to refer to what we now recognize as monogenic diseases, such as ALKAPTONURIA, has been credited with first proposing a familial component to variability in drug action6. Garrod proposed that as drugs undergo biotransformation by specific pathways in a similar way to endogenous substrates, defects in such pathways could — as for inborn errors of metabolism — alter drug concentrations and therefore effect. The concept of familial clustering of unusual drug responses was strengthened during the 1940s with the observation of a high incidence of haemolysis on exposure to antimalarial drugs among individuals with glucose-6-phosphate dehydrogenase deficiency7. In the 1950s, Price-Evans and colleagues identified N-acetylation as a major route of isoniazid elimination8. Although individuals varied substantially in the extent to which a single dose of the drug was acetylated, variability between monozygotic twins was found to be small compared with that between dizygotic twins9, laying the groundwork for studies that have now defined the clinical consequences and genetic basis of the fast and slow acetylator PHENOTYPES. More generally, the past half-century has seen developments in the understanding of the molecular basis of drug disposition (Fig. 1) and drug action, and of the mechanisms that determine the observed variability in drug actions. Hence, the concept of a familial component in drug action initiated the field of 'pharmacogenetics', even before the discovery of DNA as the repository of genetic information. With the increased understanding of the molecular, cellular and genetic determinants of drug action has come the appreciation that variants in many genes might contribute to variability in drug action; the concept of using whole-genome information to predict drug action is one definition of the more recent term, 'pharmacogenomics'10, 11, 12.

Figure 1 | Determinants of drug delivery to target sites.
Figure 1 : Determinants of drug delivery to target sites. Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, or to obtain a text description, please contact npg@nature.com

The processes of absorption, distribution into tissues, metabolism and elimination determine the amount of drug and metabolites that are delivered to target sites. In this figure, blue arrows indicate processes that enhance drug delivery, whereas green ones show processes that decrease it. Although some of the processes shown here might occur passively, much of this drug handling is mediated by specific drug-uptake or drug-efflux transporters, as well as by drug-metabolizing enzyme complexes. An ingested drug enters enterocytes, from where it can undergo metabolism, efflux into the portal circulation, or efflux back into the gut lumen. Similarly, a drug delivered to hepatocytes can be metabolized and excreted into the bile, or returned to the systemic circulation, from where it can also be excreted, generally through biliary or renal routes. If the molecular target is not located within the circulation, further obstacles to a drug accessing its molecular targets might be encountered at plasma–tissue barriers, which could limit drug access to certain cell populations, such as the brain or testes. Some drugs must access intracellular molecular targets, in which case uptake into, and efflux out of, the target cell might be key determinants of drug delivery and hence drug action.


Pharmacokinetic variability

The term 'pharmacokinetic variability' refers to variability in the delivery of a drug or metabolite(s) to target molecules, and is traditionally subdivided into the processes of absorption, distribution, metabolism and elimination, or, collectively, drug disposition. Since Garrod's original postulate, and the description of a familial component in N-acetylation, DNA variants have been described that contribute to variability in specific pathways of drug disposition, with important clinical consequences. Examples include: N-acetylation13, 14; drug oxidation by CYP2D6 (Refs 15–21), CYP2C9 (Ref. 22) or CYP2C19 (Ref. 23); conjugation by thiopurine methyltransferase24, glucuronosyltransferases25 or sulphotransferases26; and cleavage by pseudocholinesterase27, 28(Table 1).


Although familial aggregation of unusual responses to drugs has often been the first hint of the existence of clinically important variants in drug-metabolizing enzymes, modern genetic approaches have found several variants in single genes21 (see link to the Human Cytochrome P450 (CYP) Allele Nomenclature Committee). For example, over 70 variants in the CYP2D6 gene have been described, some of which lead to loss of function. Homozygotes, which comprise 7% of Caucasian and African-American populations, are rendered so-called 'poor metabolizers' on this basis. Such loss-of-function ALLELES are very uncommon among Asian populations, in which, however, alleles causing reduction of function have been described. At the other end of the catalytic spectrum are individuals with multiple functional copies of the gene, known as 'hyper-extensive metabolizers', who constitute up to 20% of some African populations. Variability in the frequency and, indeed, the types of allelic variant among ethnic populations is a common theme in contemporary genetics that could well underlie ethnic-specific beneficial and adverse drug responses. This is one of the challenges in contemporary pharmacogenomic analyses, as discussed further below.

As a general principle, the problem of DNA variants contributing to aberrant drug metabolism becomes most evident for drugs that have only a narrow margin between the dosages that are required for efficacy, and those that are associated with serious toxicity (such as is the case with cardiovascular or oncology drugs), as well as drugs that have only a single main pathway for elimination. Drugs whose biotransformation to inactive metabolites is CYP2D6 dependent (for example, some tricyclic antidepressants or beta-adrenergic blockers) cause side effects more often among poor metabolizers, and lack of efficacy among hyperextensive metabolizers. Conversely, drugs such as codeine, which undergoes CYP2D6-dependent biotransformation to form its more active metabolite (morphine), can have a lack of efficacy in poor metabolizers, and exaggerated effects among hyperextensive metabolizers. A minority of individuals with 'aberrant' metabolism makes up the subset that is generally identified in clinical investigation, and it is in this group that aberrant drug responses are most commonly seen early during drug therapy. A second, increasingly recognized, problem in the disposition of a drug that uses a single main pathway is the potential for drug–drug interactions. So, inhibition of CYP2D6 by co-administration of serotonin re-uptake inhibitors or tricyclic anti-depressant drugs29, or inhibition of CYP3A4 by co-administration of erythromycin or ketoconazole30, can cause adverse drug effects, which occur during chronic drug therapy and are therefore a risk in most subjects that have 'normal' metabolism.

Although the concept of genetic variants in the proteins that accomplish drug metabolism is a relatively mature one, several new areas in drug disposition are emerging. One is the increasing recognition that drug uptake into, and efflux from, intracellular sites is accomplished by specific drug transport molecules, and that these also exhibit pharmacologically important allelic variability31, 32, 33, 34. For example, the integrity of the blood–brain barrier is now known to arise, not just from tight junctions in the capillary endothelium in this region, but also from expression of the drug-efflux transporter P-glycoprotein on the luminal surface of these cells, which thereby limits the access of drugs to the brain35.

Another area of active enquiry is the transcriptional regulation of normal proteins, which can be highly variable because of allelic variants in regions of DNA that regulate expression25, 36. Variations in the function or expression of genes encoding factors, such as NUCLEAR ORPHAN RECEPTORS, that control the transcription of the genes encoding drug-metabolizing enzymes and transporters37, 38, 39, 40, 41, could also contribute to variable drug actions.

Ultimately, a goal of scientists who work in this area is to develop the ability to predict in vivo disposition in humans strictly from in vitro and animal data. An important step in this direction will be to catalogue the genetic variants in drug-metabolizing and drug-transport systems, evaluate their in vitro consequences, and relate these to clinical drug actions. As this field advances, it might become possible to predict individual concentration–time profiles for a specific drug, which are based on specific allelic variants that are present in an individual patient.

Pharmacodynamic variability

Whereas the term pharmacokinetics describes the relationship between the drug dose and the resulting plasma and tissue drug concentrations, 'pharmacodynamics' refers to the relationship between the drug concentration and its effect. Individuals with identical plasma and tissue drug concentrations vary in their responses, indicating that pharmacodynamic mechanisms could contribute a second important component to variable drug actions (Fig. 2). Methods to evaluate such variability in response are less well standardized than those used to study drug disposition. In general, pharmacodynamic variability can arise from two distinct mechanisms. In the first mechanism, a drug exerts a variable effect because the specific molecular target on which it acts has some (often genetically determined) variability. For instance, the APOE GENOTYPE determines the extent of choline acetyltransferase expression, and has been linked to the response to therapy with tacrine, a choline acetyltransferase inhibitor that is used in the treatment of mild to moderate Alzheimer's disease42. A similar phenomenon is observed in patients who are exposed to drugs that prolong the QT INTERVAL on a surface electrocardiogram (an effect that, if exaggerated, can lead to fatal cardiac arrhythmias). Virtually all drugs that are known to cause this effect block one specific cardiac ion current, termed IKr, which is generated by expression of a potassium voltage-gated channel that is encoded by the human ether a go-go-related gene (HERG), also known as KCNH2, and an ancillary protein, the KCNE2 gene product (termed MiRP1). In a patient with exaggerated QT prolongation and arrhythmias that occurred in response to antibiotic therapy, a KCNE2 variant was identified that renders IKr threefold more sensitive to block by the culprit antibiotic43. This concept has been expanded to include variants in the transcriptional control of target molecules. Response to the 5-lipoxygenase inhibitor zileuton was markedly decreased in subjects that were HOMOZYGOUS for alleles that reduced 5-lipoxygenase expression44; that is, the effect was diminished in patients with decreased expression of the target molecule.

Figure 2 | Determinants of drug action at the target site.
Figure 2 : Determinants of drug action at the target site. Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, or to obtain a text description, please contact npg@nature.com

The drug is indicated as blue circles, and the metabolite as green squares. Delivery of drug and metabolite to, and removal from, target sites on the cell surface or within the cell are controlled by processes with activities that might be genetically determined (shown in Fig. 1). Variability in drug action can also arise through the pharmacodynamic mechanisms illustrated here and discussed further in the text: first, the drug might interact with several other targets; second, there might be variability in the function or expression level of the drug target; or third, other molecules might modulate the biological context within which the drug–target interaction takes place. DNA variants in elements that control each of these processes can lead to variable drug actions.


The second, more generic, form of pharmacodynamic variability is the variability of the broader biological context — a function of the expression of tens or hundreds of genes — in which the interaction between a drug and its target molecules takes place. As an example, administration of a drug with IKr blocking properties does not ordinarily perturb cardiac repolarization excessively, probably because of the presence of other repolarizing currents, notably one termed IKs. Individuals with subclinical loss-of-function MUTATIONS in genes encoding IKs,, who have normal cardiac repolarization at baseline, have been shown to develop marked repolarization abnormalities and arrhythmias when challenged with an IKr-blocking drug45, 46, 47, 48. Similarly, beta-blockers have been shown to be especially beneficial in a group of patients at high risk of heart failure who are homozygous for an intronic deletion in the angiotensin-converting enzyme (ACE) gene (the DD genotype49), which encodes a key enzyme in the renin–angiotensin system50, 51, even though beta-blockers do not act directly on the ACE gene itself51. More generally, each POLYMORPHISM that mediates the development or severity of a human disease can be viewed as a candidate for modulating the responses of drugs that are used to treat that disease.

Pharmacogenomics

These examples of individual DNA variants that might modulate response to drugs provide a conceptual framework for considering the way in which multiple DNA variants could influence the outcome of drug therapy. A general algorithm for identifying DNA variants that might modulate drug responses is presented in Fig. 3. The first step is to identify a phenotype of interest. This is ordinarily a clinically important end point — which can be a beneficial drug effect or a serious adverse effect — that has considerable inter-individual variability with no 'obvious' cause, and therefore might have a genetic underpinning. When the phenotype is a variable physiological trait or expression of disease, the next step is often to conduct epidemiological studies and/or studies in well-characterized kindreds to establish that variability in the trait under study is, indeed, familial. Studies in which the responses of twins that have been exposed to a drug are observed have been important for addressing this question9, 52. However, this approach is generally not applicable to pharmacogenetics, as, in drug therapy, large kindreds in which the response of each family member to drug exposure has been categorized are not generated. The second step is to accumulate patients (and their DNA) with the defined phenotype, along with control subjects who are ideally matched for all variables that are known or suspected to modulate the phenotype, such as ethnicity.

Figure 3 | An algorithm for evaluating the role of genetic factors in drug actions.
Figure 3 : An algorithm for evaluating the role of genetic factors in drug actions. Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, or to obtain a text description, please contact npg@nature.com

This flow chart, which is discussed further in the text, highlights the interactive and collaborative nature of the research that is required to accomplish the goal of defining the role of genetic factors in variable drug actions.


Identification of candidate genes

The third step in the algorithm is to identify genes, or sets of genes, in which polymorphisms might reasonably be expected to modulate the phenotype. Both candidate-gene and genome-wide approaches have been proposed. Genome-wide scans have been useful in POSITIONAL-CLONING STRATEGIES to find genes in Mendelian disorders, but the approach is much more difficult to adapt to pharmacogenetics, as kindreds are not generally available. Nevertheless, with advances in technology and in statistical approaches, it may be possible to identify a circumscribed set of polymorphisms, perhaps representing most of the HAPLOTYPES in the human genome, which could be used in genome-wide scans for important drug-response end points in large numbers of well-characterized patients and controls. The main advantage of such an approach is that genes or pathways that are unidentified at present may be implicated in mediating drug responses. Until such advances are made, however, it seems likely that pharmacogenetics will rely on a candidate-gene approach. Some candidate genes are relatively obvious. For example, association of an adverse drug response with elevated plasma concentrations would naturally focus attention on genes with products that mediate drug disposition — often a relatively circumscribed and tractable set. For pharmacodynamic issues, the most obvious candidate genes are those that encode the drug targets. However, as discussed above, genes with products that modulate the biological context in which the drug–target interactions occur should also be considered as candidates. Returning to the example of variable responses to anti-arrhythmic drugs, the list of candidates includes not only genes encoding ion channels, but also those encoding function-modifying subunits, elements of intracellular signalling pathways that modulate ion channel function, factors that control intracellular calcium (a key mediator of arrhythmias), and modulators of the extracellular matrix and cell-to-cell communications that establish re-entrant circuits, to list just a few of the possibilities. Hence, basic physiological considerations, driven by clinical research and molecular and cellular studies, including (where appropriate) model systems, such as mouse, yeast, or Caenorhabditis elegans, can all contribute to the definition of these biological contexts and so have a role in defining candidate sets. Indeed, the response of yeast to exogenous stimuli has been sufficiently well characterized that a 'compendium' of virtually all responses has been described53 and used to identify pathways modulated by drugs with mechanisms that were previously poorly understood. It would be desirable to be able to create a similar compendium of all possible responses of human cells to exogenous stimuli, but this vision remains futuristic.

Identification of polymorphisms

The next step in this algorithm is the identification of DNA variants in candidate genes or gene sets. The human genome probably includes 3–10 times 106 single nucleotide polymorphisms (SNPs)54 and, although considerable progress has been made in identifying these and assigning them to specific genes, the sheer numbers raise considerable questions as to how any sense can be made of statistical analyses of these numerous variants in relatively small (even thousands) numbers of patients (Box 1). One approach is to confine the analysis to polymorphisms that alter primary amino-acid sequences, known as NON-SYNONYMOUS CODING-REGION POLYMORPHISMS. As technologies advance, polymorphisms in known promoter regions, or in or near intron–exon boundaries, can readily be included in such a 'first-pass' analysis. However, these strategies make assumptions about our understanding of transcriptional and translational control mechanisms that are not well supported; the formal possibility exists that SYNONYMOUS CODING-REGION POLYMORPHISMS or distant intronic sequences might modulate the efficiency of transcription. The recognition that many SNPs are actually in LINKAGE DISEQUILIBRIUM, and that the number of haplotypes in the human genome is probably far less than the millions of SNPs, might provide a new tool with which to approach the problem of genomic variation as a contributor to disease and drug response55, 56. Rather than requiring analysis at each of the millions of polymorphic sites, it might be possible to perform such analyses with thousands, or, at most, tens of thousands, of haplotypes.

Once a list of candidate DNA variants is generated that might modulate the phenotype of interest, patients and controls need to be genotyped. The main problems here are choice of platform, the costs, and the issues of polymorphism versus haplotype that were discussed above. Furthermore, the biostatistical challenges in associating variant DNA sequences with pre-defined drug responses are considerable (Box 1). Finally, after a polymorphism or a set of polymorphisms that predict drug response is identified, it should be subjected to a prospective test in a new population to establish that the predictive value is reproducible, and cost–benefit issues should be analysed. In addition, the polymorphism information itself could be used to refine the definition of the phenotype and, perhaps, reinitiate the experiment.

Challenges in pharmacogenetic mapping

Some of the challenges that are involved in conducting this sort of experiment have been described: accumulation of well-characterized patient databases; identification and accumulation of appropriate control group(s); statistical and methodological approaches that are still in evolution; and cost. A fundamental conceptual issue that the field must grapple with is the question of whether it is worth finding very rare alleles that predict rare but serious adverse drug effects. Thiopurine-methyltransferase deficiency, which predicts BONE MARROW APLASIA during exposure to 6-mercaptopurine, a treatment for childhood leukaemia, is a good example57, 58. Although many variants that can cause this adverse drug effect have been identified, they are rare (approximately 1:300 individuals). In this instance, a clinically important but rare allelic variant was identified, not through a very large screen, but in targeted clinical research in affected subjects.

There are considerable ethical issues involved in large-scale attempts to characterize the genetic basis of human physiology, pathology and response to drugs59, 60. A fundamental problem arises because information that is gained in the course of genetic studies could conceivably be used to harm individuals participating in the research, or individuals not participating, but whose genotype is under study. Such studies might, for instance, be made available to other entities, such as law-enforcement or insurance companies. Furthermore, after an individual provides consent for the use of his or her DNA and clinical information in a trial, the information that is generated could carry important implications (negative or positive) for others who share the genotypes, but who did not consent to the study, such as family members or members of the same ethnic group. For instance, characterizing drug responses in defined ethnic groups might carry with it a risk of stigmatization.

The need for collaboration

Variable responses to drug therapy are well recognized, and there are now many examples of individual DNA variants that mediate both pharmacokinetic and pharmacodynamic responses. The vision, raised by the sequencing of the human genome, of applying this information to more broadly understand the genetic basis of variability in drug responses is tantalizing, but faces considerable challenges. The successful implementation of this strategy will require intense collaboration among clinicians who see and accurately phenotype patients (and without whom the research could not be done), industry (where well-characterized databases of patients and their drug responses are in place), researchers in industry and in academic medicine who identify candidate genes and pathways, and scientists who are developing the technological advances that are required to both genotype large numbers of patients at several sites and analyse the flood of information that results. In the United States, another of the main partners is the National Institutes of Health, which has recently devoted considerable resources to large genomic-based projects, including the establishment of the Pharmacogenetics Research Network and Knowledge Base. The broad goal of this network is to further develop strategies for pharmacogenetic mapping and apply them to study variability in families of candidate genes, such as transporters or transferases, or in specific diseases, such as cancer, asthma, arrhythmias or hypertension. Although each constituency in this potential partnership has its own primary agenda, it is relatively clear that an individual entity will not be able to accomplish all the required steps without collaboration.

Future directions

One outcome of pharmacogenetic and pharmacogenomic research efforts might be better use of available drugs. For example, although beta-blockers are well recognized as effective therapies for patients with congestive heart failure, they are difficult to use in these patients and carry a relatively high incidence of adverse effects. The information that the benefit is largely confined to patients with the ACE DD genotype was obtained retrospectively, in a relatively small number of patients. Validation of this finding implies that genotyping could be used to identify patients who are especially likely to benefit from this cumbersome form of therapy and those in whom beta-blockade is unnecessary.

A second outcome might be validation or rejection of new drug targets at an early stage of development. A drug target with a functionally significant polymorphism might be rejected in favour of one that does not have such genetic variability. Similarly, as the molecular basis for unusual adverse drug effects, such as drug-induced arrhythmias or drug-associated hepatotoxicity, becomes better defined in pharmacogenetic studies, new screening algorithms could be developed to eliminate, at an early stage of development, drugs that are likely to be associated with these adverse effects.

It is also likely that, for many drugs, variability in response might reflect gene–environment interactions (with an emphasis on an environmental component), or large numbers of polymorphisms or haplotypes, each contributing only a very small percentage to the overall variability. In these cases, prediction of individual responses based on variation in such large numbers of genes would be extremely difficult to implement.

The approach of developing new drugs that meet major unmet medical needs and that can be used in all patients with a common disease (the 'blockbuster' strategy) has been highly successful in public health terms. ACE inhibitors and HMG-CoA reductase (3-hydroxy-3-methylglutaryl coenzyme A reductase) inhibitors (statins) are examples of agents that have provided benefits to numerous ungenotyped patients, albeit with some variability among defined subsets, and with serious toxicity in very small numbers of patients. ACE inhibitors reduce mortality in patients with congestive heart failure and those convalescing from acute myocardial infarction by 20–40% (depending on the subset)61, 62, but sometimes cause life-threatening angioedema, especially among African populations63, 64. Similarly, statins strikingly reduce the incidence of myocardial infarction among subjects at even moderate risk65, 66, but can occasionally cause RHABDOMYOLYSIS67. Pharmacogenetic studies designed to understand mechanisms underlying ACE-inhibitor-related angioedema or statin-related rhabdomyolysis might well make such agents safer, but incorporation of such studies during development might also have prevented these agents from reaching the public altogether. Ultimately, information that is gathered from pharmacogenetic studies might, in fact, have its greatest application, not so much in predicting responses to available agents, but in furthering our understanding of human physiology and pathology. This could lead to the development of new therapies68 for patients with diseases that are well characterized at the molecular and genetic levels, enabling treatment using drugs with disposition and effects that are well understood and highly predictable in an individual subject.

Top

Acknowledgements

Work in D.M.R.'s laboratory is supported in part by grants from the United States Public Health Service. D.M.R. holds the William Stokes Chair in Experimental Therapeutics, a gift from the Daiichi Corporation.

Top

References

  1. Kennedy, D. L., Johnson, J. M. & Nightingale, S. L. Monitoring of adverse drug events in hospitals. JAMA 266, 2878 (1991).

  2. Classen, D. C., Pestotnik, S. L., Evans, R. S., Lloyd, J. F. & Burke, J. P. Adverse drug events in hospitalized patients. Excess length of stay, extra costs, and attributable mortality. JAMA 277, 301–306 (1997).

  3. Dormann, H. et al. Incidence and costs of adverse drug reactions during hospitalisation: computerised monitoring versus stimulated spontaneous reporting. Drug Saf. 22, 161–168 (2000).

  4. Brewer, T. & Colditz, G. A. Postmarketing surveillance and adverse drug reactions: current perspectives and future needs. JAMA 281, 824–829 (1999).

  5. Phillips, K. A., Veenstra, D. L., Oren, E., Lee, J. K. & Sadee, W. Potential role of pharmacogenomics in reducing adverse drug reactions: a systematic review. JAMA 286, 2270–2279 (2001).

  6. Motulsky, A. G., Harper, P. S., Bobrow, M. & Scriver, C. in Pharmacogenetics 1st edn (ed. Weber, W. W.) 1–20 (Oxford Univ. Press, New York, 1997).

  7. Beutler, E., Dern, R. J. & Alving, A. S. The hemolytic effect of primaquine. VI. An in vitro test for sensitivity of erythrocytes to primaquine. J. Lab. Clin. Med. 45, 40–50 (1955).

  8. Price-Evans, D. A., Manley, F. A. & McKusick, V. A. Genetic control of isoniazid metabolism in man. Br. Med. J. 2, 485–491 (1960).

  9. Bönicke, R. & Lisboa, B. P. Uber die Erbbedingtheit der intraindividuellen Konstantz der Isoniazidausscheidung beim Menshen (Untersuchen an eineiggen und zweieiggen Zwilligen). Naturwissenschafften 44, 314 (1957).

  10. Evans, W. E. & Relling, M. V. Pharmacogenomics: translating functional genomics into rational therapeutics. Science 286, 487–491 (1999).

  11. Roses, A. D. Pharmacogenetics and the practice of medicine. Nature 405, 857–865 (2000).

  12. Meyer, U. A. Pharmacogenetics and adverse drug reactions. Lancet 356, 1667–1671 (2000).

  13. Woosley, R. L. et al. Effect of acetylator phenotype on the rate at which procainamide induces antinuclear antibodies and the lupus syndrome. N. Engl. J. Med. 298, 1157–1159 (1978).

  14. Grant, D. M., Blum, M. & Meyer, U. A. Polymorphisms of N-acetyltransferase genes. Xenobiotica 22, 1073–1081 (1992).

  15. Mahgoub, A., Idle, R. J., Dring, L. G., Lancaster, R. & Smith, R. L. Polymorphic hydroxylation of debrisoquine in man. Lancet 2, 584–586 (1977).
    Initial description of the clinical importance of the CYP2D6 polymorphism in drug disposition and drug action.

  16. Eichelbaum, M., Spannbrucker, N., Steincke, B. & Dengler, H. J. Defective N-oxidation of sparteine in man: a new pharmacogenetic defect. Eur. J. Clin. Pharmacol. 16, 183–187 (1979).

  17. Caraco, Y., Sheller, J. & Wood, A. J. Impact of ethnic origin and quinidine coadministration on codeine's disposition and pharmacodynamic effects. J. Pharmacol. Exp. Ther. 290, 413–422 (1999).

  18. Brosen, K., Zeugin, T. & Meyer, U. A. Role of P450IID6, the target of the sparteine–debrisoquin oxidation polymorphism, in the metabolism of imipramine. Clin. Pharmacol. Ther. 49, 609–617 (1991).

  19. Lee, J. T. et al. The role of genetically determined polymorphic drug metabolism in the beta-blockade produced by propafenone. N. Engl. J. Med. 322, 1764–1768 (1990).

  20. Lennard, M. S. et al. Oxidation phenotype — a major determinant of metoprolol metabolism and response. N. Engl. J. Med. 307, 1558–1560 (1982).

  21. Meyer, U. A. & Zanger, U. M. Molecular mechanisms of genetic polymorphisms of drug metabolism. Annu. Rev. Pharmacol. Toxicol. 37, 269–296 (1997).

  22. Aithal, G. P., Day, C. P., Kesteven, P. J. & Daly, A. K. Association of polymorphisms in the cytochrome P450 CYP2C9 with warfarin dose requirement and risk of bleeding complications. Lancet 353, 717–719 (1999).

  23. Furuta, T. et al. Effect of genetic differences in omeprazole metabolism on cure rates for Helicobacter pylori infection and peptic ulcer. Ann. Intern. Med. 129, 1027–1030 (1998).

  24. Weinshilboum, R. Methyltransferase pharmacogenetics. Pharmacol. Ther. 43, 77–90 (1989).

  25. Ando, Y. et al. Polymorphisms of UDP-glucuronosyltransferase gene and irinotecan toxicity: a pharmacogenetic analysis. Cancer Res. 60, 6921–6926 (2000).

  26. Glatt, H. et al. Human cytosolic sulphotransferases: genetics, characteristics, toxicological aspects. Mutat. Res. 482, 27–40 (2001).

  27. Forbat, A., Lond, M. B., Lehmann, H. & Silk, E. Prolonged apnea following injection of succinylcholine. Lancet 2, 1067–1068 (1953).

  28. Whittaker, M. Genetic aspects of succinylcholine sensitivity. Anesthesiology 32, 143–150 (1970).

  29. Vandel, S. et al. Tricyclic antidepressant plasma levels after fluoxetine addition. Neuropsychobiology 25, 202–207 (1992).

  30. Woosley, R. L., Chen, Y., Freiman, J. P. & Gillis, R. A. Mechanism of the cardiotoxic actions of terfenadine. J. Am. Med. Assoc. 269, 1532–1536 (1993).
    Identification of the pharmacokinetic and ion-channel blocking mechanisms that underlie the potentially fatal cardiotoxicity that is associated with the widely used antihistamine terfenadine.

  31. Fromm, M. F., Kim, R. B., Stein, C. M., Wilkinson, G. R. & Roden, D. M. Inhibition of P-glycoprotein-mediated drug transport: A unifying mechanism to explain the interaction between digoxin and quinidine. Circulation 99, 552–557 (1999).

  32. Kim, R. B. et al. P-glycoprotein transporter limits oral absorption and brain entry of HIV protease inhibitors. J. Clin. Invest. 101, 289–294 (1998).
    HIV-protease inhibitors access the CNS in mice with disrupted P-glycoprotein, a potential explanation for the CNS as a viral 'sanctuary site' in protease-inhibitor therapy.

  33. Hoffmeyer, S. et al. Functional polymorphisms of the human multidrug-resistance gene: multiple sequence variations and correlation of one allele with P-glycoprotein expression and activity in vivo. Proc. Natl Acad. Sci. USA 97, 3473–3478 (2000).

  34. Kim, R. B. et al. Identification of functionally variant MDR1 alleles among European Americans and African Americans. Clin. Pharmacol. Ther. 70, 189–199 (2001).

  35. Schinkel, A. H. et al. Disruption of the mouse mdr1a P-glycoprotein gene leads to a deficiency in the blood–brain barrier and to increased sensitivity to drugs. Cell 77, 491–502 (1994).

  36. Kuehl, P. et al. Sequence diversity in CYP3A promoters and characterization of the genetic basis of polymorphic CYP3A5 expression. Nature Genet. 27, 383–391 (2001).

  37. Lehmann, J. M. et al. The human orphan nuclear receptor PXR is activated by compounds that regulate CYP3A4 gene expression and cause drug interactions. J. Clin. Invest. 102, 1016–1023 (1998).
    Regulation of expression of functionally normal genes as a mechanism that underlies drug action.

  38. Smirlis, D., Muangmoonchai, R., Edwards, M., Phillips, I. R. & Shephard, E. A. Orphan receptor promiscuity in the induction of cytochromes p450 by xenobiotics. J. Biol. Chem. 276, 12822–12826 (2001).

  39. Synold, T. W., Dussault, I. & Forman, B. M. The orphan nuclear receptor SXR coordinately regulates drug metabolism and efflux. Nature Med. 7, 584–590 (2001).

  40. Xie, W. et al. Reciprocal activation of xenobiotic response genes by nuclear receptors SXR/PXR and CAR. Genes Dev. 14, 3014–3023 (2000).

  41. Xie, W. et al. Humanized xenobiotic response in mice expressing nuclear receptor SXR. Nature 406, 435–439 (2000).

  42. Poirier, J. et al. Apolipoprotein E4 allele as a predictor of cholinergic deficits and treatment outcome in Alzheimer disease. Proc. Natl Acad. Sci. USA 92, 12260–12264 (1995).

  43. Sesti, F. et al. A common polymorphism associated with antibiotic-induced cardiac arrhythmia. Proc. Natl Acad. Sci. USA 97, 10613–10618 (2000).

  44. Drazen, J. M. et al. Pharmacogenetic association between ALOX5 promoter genotype and the response to anti-asthma treatment. Nature Genet. 22, 168–170 (1999).
    Shows that a promoter polymorphism regulates the expression of a drug target, and therefore drug effect.

  45. Donger, C. et al. KVLQT1 C-terminal missense mutation causes a forme fruste long-QT syndrome. Circulation 96, 2778–2781 (1997).

  46. Napolitano, C. et al. Evidence for a cardiac ion channel mutation underlying drug-induced QT prolongation and life-threatening arrhythmias. J. Cardiovasc. Electrophysiol. 11, 691–696 (2000).

  47. Yang, P. et al. Frequency of ion channel mutations and polymorphisms in a large population of patients with drug-associated Long QT Syndrome. Pacing Clin. Electrophys. 24, 579 (2001).

  48. Wei, J. et al. KCNE1 polymorphism confers risk of drug-induced long QT syndrome by altering kinetic properties of IKs potassium channels. Circulation 100, 1–495 (1999).

  49. Rigat, B. et al. An insertion/deletion polymorphism in the angiotensin I-converting enzyme gene accounting for half the variance of serum enzyme levels. J. Clin. Invest. 86, 1343–1346 (1990).

  50. Ueda, S., Meredith, P. A., Morton, J. J., Connell, J. M. & Elliott, H. L. ACE (I/D) genotype as a predictor of the magnitude and duration of the response to an ACE inhibitor drug (enalaprilat) in humans. Circulation 98, 2148–2153 (1998).

  51. McNamara, D. M. et al. Pharmacogenetic interactions between beta-blocker therapy and the angiotensin-converting enzyme deletion polymorphism in patients with congestive heart failure. Circulation 103, 1644–1648 (2001).
    Variability in response to beta-blocker therapy in heart failure due to a polymorphism in the ACE gene, the activity of which determines the biological context in which beta-blockers act.

  52. Vesell, E. S. & Page, J. G. Genetic control of drug levels in man: antipyrine. Science 161, 72–73 (1968).

  53. Hughes, T. R. et al. Functional discovery via a compendium of expression profiles. Cell 102, 109–126 (2000).

  54. Sachidanandam, R. et al. A map of human genome sequence variation containing 1.42 million single nucleotide polymorphisms. Nature 409, 928–933 (2001).

  55. Stephens, J. C. et al. Haplotype variation and linkage disequilibrium in 313 human genes. Science 293, 489–493 (2001).

  56. Judson, R. & Stephens, J. C. Notes from the SNP vs haplotype front. Pharmacogenomics 2, 7–10 (2001).

  57. Otterness, D. M., Szumlanski, C. L., Wood, T. C. & Weinshilboum, R. M. Human thiopurine methyltransferase pharmacogenetics. Kindred with a terminal exon splice junction mutation that results in loss of activity. J. Clin. Invest. 101, 1036–1044 (1998).

  58. McLeod, H. L., Krynetski, E. Y., Relling, M. V. & Evans, W. E. Genetic polymorphism of thiopurine methyltransferase and its clinical relevance for childhood acute lymphoblastic leukemia. Leukemia 14, 567–572 (2000).
    The activity of thiopurine methyltransferase determines both the efficacy and the severe toxicity of drug therapy for the treatment of acute lymphoblastic leukaemia and, in some centres, genotyping to guide dosage is now routine; probably the first example of pre-prescription genotyping in clinical practice.

  59. Rothstein, M. A. & Epps, P. G. Ethical and legal implications of pharmacogenomics. Nature Rev. Genet. 2, 228–231 (2001).

  60. Robertson, J. A. Consent and privacy in pharmacogenetic testing. Nature Genet. 28, 207–209 (2001).
    Discusses the substantial ethical issues involved in genetic and, especially, in pharmacogenetic studies.

  61. The SOLVD Investigators. Effect of enalapril on survival in patients with reduced left ventricular ejection fractions and congestive heart failure. N. Engl. J. Med. 325, 293–302 (1991).

  62. Pfeffer, M. A. et al. Effect of captopril on mortality and morbidity in patients with left ventricular dysfunction after myocardial infarction. Results of the survival and ventricular enlargement trial. N. Engl. J. Med. 327, 669–677 (1992).

  63. Brown, N. J., Ray, W. A., Snowden, M. & Griffin, M. R. Black Americans have an increased rate of angiotensin converting enzyme inhibitor-associated angioedema. Clin. Pharmacol. Ther. 60, 8–13 (1996).

  64. Brown, N. J., Snowden, M. & Griffin, M. R. Recurrent angiotensin-converting enzyme inhibitor-associated angioedema. J. Am. Med. Assoc. 278, 232–233 (1997).

  65. Scandinavian Simvastatin Survival Study Group. Randomised trial of cholesterol lowering in 4,444 patients with coronary heart disease: the Scandinavian Simvastatin Survival Study (4S). Lancet 344, 1383–1389 (1994).

  66. Shepherd, J. et al. Prevention of coronary heart disease with pravastatin in men with hypercholesterolemia. West of Scotland Coronary Prevention Study Group. N. Engl. J. Med. 333, 1301–1307 (1995).

  67. Ucar, M., Mjorndal, T. & Dahlqvist, R. HMG-CoA reductase inhibitors and myotoxicity. Drug Saf. 22, 441–457 (2000).

  68. Kleyn, P. W. & Vesell, E. S. Genetic variation as a guide to drug development. Science 281, 1820–1821 (1998).

  69. Mamiya, K. et al. The effects of genetic polymorphisms of CYP2C9 and CYP2C19 on phenytoin metabolism in Japanese adult patients with epilepsy: studies in stereoselective hydroxylation and population pharmacokinetics. Epilepsia 39, 1317–1323 (1998).

  70. Sullivan-Klose, T. H. et al. The role of the CYP2C9 Leu359 allelic variant in the tolbutamide polymorphism. Pharmacogenetics 6, 341–349 (1996).

  71. Wedlund, P. J., Aslanian, W. S., McAllister, C. B., Wilkinson, G. R. & Branch, R. A. Mephenytoin hydroxylation deficiency in caucasians: frequency of a new oxidative drug metabolism polymorphism. Clin. Pharmacol. Ther. 36, 773–780 (1984).

  72. Dahl, M.-L., Johansson, I., Bertilsson, L., Ingelman-Sundberg, M. & Sjoeqvist, F. Ultrarapid hydroxylation of debrisoquine in a Swedish population: analysis of the molecular genetic basis. J. Pharmacol. Exp. Ther. 274, 516–520 (1995).

  73. Edeki, T. I., He, H. & Wood, A. J. Pharmacogenetic explanation for excessive beta-blockade following timolol eye drops. Potential for oral-ophthalmic drug interaction. J. Am. Med. Assoc. 274, 1611–1613 (1995).

  74. Zhou, H. H. & Wood, A. J. Stereoselective disposition of carvedilol is determined by CYP2D6. Clin. Pharmacol. Ther. 57, 518–524 (1995).

  75. Idle, J. R., Mahgoub, A., Lancaster, R. & Smith, R. L. Hypotensive response to debrisoquine and hydroxylation phenotype. Life Sci. 22, 979–984 (1978).

  76. Ellard, G. A., Mitchison, D. A., Girling, D. J., Nunn, A. J. & Fox, W. The hepatic toxicity of isoniazid among rapid and slow acetylators of the drug. Am. Rev. Respir. Dis. 118, 628–629 (1978).

  77. Stolk, J. N. et al. Reduced thiopurine methyltransferase activity and development of side effects of azathioprine treatment in patients with rheumatoid arthritis. Arthritis Rheum. 41, 1858–1866 (1998).

  78. Israel, E. et al. The effect of polymorphisms of the beta(2)-adrenergic receptor on the response to regular use of albuterol in asthma. Am. J. Respir. Crit. Care Med. 162, 75–80 (2000).

  79. Lutucuta, S., Ballantyne, C. M., Elghannam, H., Gotto, A. M.Jr, & Marian, A. J. Novel polymorphisms in promoter region of ATP binding cassette transporter gene and plasma lipids, severity, progression, and regression of coronary atherosclerosis and response to therapy. Circ. Res. 88, 969–973 (2001).

  80. Kuivenhoven, J. A. et al. The role of a common variant of the cholesteryl ester transfer protein gene in the progression of coronary atherosclerosis. The Regression Growth Evaluation Statin Study Group. N. Engl. J. Med. 338, 86–93 (1998).

Author affiliations

  1. Departments of Medicine and Pharmacology, Vanderbilt University School of Medicine, 532 Robinson Research Building, Nashville, Tennessee 37232, USA.

Correspondence to: Dan M. Roden1 Email: Dan.Roden@ MCMail.Vanderbilt.edu

You have created an error.

Extra navigation

Search PubMed for

natureevents

Advertisement