Clinical Chemistry AACC Online Job Center
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Clinical Chemistry 43: 254-266, 1997;
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit an electronic Letter to
the Editor about this paper
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via ISI Web of Science (93)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Linder, M. W.
Right arrow Articles by Valdes, R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Linder, M. W.
Right arrow Articles by Valdes, R., Jr.
Related Collections
Right arrow Laboratory Management
Right arrow Evidence Based Laboratory Medicine and Test Utilization
Right arrow Drug Monitoring and Toxicology
(Clinical Chemistry. 1997;43:254-266.)
© 1997 American Association for Clinical Chemistry, Inc.

Review

Pharmacogenetics: a laboratory tool for optimizing therapeutic efficiency

Mark W. Linder1, Russell A. Prough2 and Roland Valdes, Jr.1,2,a

Departments of
1 Pathology and
2 Biochemistry, University of Louisville School of Medicine, Louisville, KY 40292.
a Address correspondence to this author at: Department of Pathology, University of Louisville, Louisville, KY 40292. Fax 502-852-1771; e-mail ROVALDO1{at}ULKYVM.Louisville.edu


   Abstract
Top
 Abstract
Introduction
 enzymes involved in drug...
 cytochrome p-450 superfamily of...
 phenotyping vs genotyping
 CYP Enzymes
 References
 
Pharmacogenetics is the study of the linkage between an individual's genotype and that individual's ability to metabolize a foreign compound. Differences in metabolism of therapeutics can lead to severe toxicity or therapeutic failure by altering the relation between dose and blood concentration of the pharmacologically active drug. Phenotypes exhibiting poor and ultraextensive metabolism result from genetic variance (polymorphism) of enzymes involved in metabolism. Thus, in pharmacogenetic studies one applies genotyping of polymorphic alleles encoding drug-metabolizing enzymes to the identification of an individual's drug metabolism phenotype. This knowledge, when applied to dosing or drug selection, can avoid adverse reactions or therapeutic failure and thus enhance therapeutic efficiency. More than 25 commonly prescribed medicines are metabolized by the cytochrome P-4502D6 (CYP2D6) isoenzyme, and polymorphism of the CYP2D6 gene affects the therapeutic management of up to 17% of individuals in some ethnic groups. In this review, we summarize and update information concerning drug-metabolizing genotypes with emphasis on CYP2D6 genotyping techniques that can be applied by the clinical laboratory for linking human genetics to therapeutic management.


Key Words: indexing terms: cytochrome P-450 • drug metabolism • human genetics • outcomes research • preventive medicine • therapeutic drug monitoring • tricyclic antidepressants


   Introduction
Top
 Abstract
 Introduction
enzymes involved in drug...
 cytochrome p-450 superfamily of...
 phenotyping vs genotyping
 CYP Enzymes
 References
 
Most xenobiotics, including pharmaceutical agents, are metabolized to some extent. Metabolism results in detoxification and elimination of the drug or activation of the prodrug to the biologically active therapeutic or toxin. When the pharmacological activity of a drug or toxin is linked to the catalytic activity of a specific enzyme, factors that influence the activity of the enzyme will affect the clinical response to the agent. Enzymes responsible for the activation and metabolism of drugs and other compounds in humans show wide interindividual variation in their protein expression or catalytic activity, resulting in unique drug metabolism phenotypes. This variation can be due to transient causes such as enzyme inhibition and induction, or to a permanent cause such as genetic mutation or gene deletion. When specific gene mutations or deletions are maintained within the population, the gene is said to be polymorphic or having multiple forms. Genetic polymorphism has been linked to three classes of phenotypes based on the extent of drug metabolism. Extensive metabolism (EM) of a drug is characteristic of the normal population; poor metabolism (PM) is associated with accumulation of specific drug substrates and is typically an autosomal recessive trait requiring mutation and (or) deletion of both alleles for phenotypic expression; and ultraextensive metabolism (UEM) results in increased drug metabolism and is an autosomal dominant trait arising from gene amplification.1 For some classes of therapeutic agents and environmental/occupational carcinogens, there is good evidence that genetic polymorphism of drug-metabolizing enzymes plays a significant role in adverse effects of therapeutic agents or incidence of exposure-linked cancer. Thus, determination of these genetic polymorphisms may be of clinical value in predicting adverse or inadequate response to certain therapeutic agents and in predicting increased risk of environmental or occupational exposure-linked disease.

Some relevant questions concerning the clinical application of genotyping/phenotyping data as it applies to the clinical laboratory and to the clinical practitioner are: When is genotyping clinically indicated? What are the advantages of adding genotyping to therapeutic drug monitoring in assessing the clinical status of the patient? How is genotyping information most efficiently applied to dosage adjustments and (or) choice of therapeutic? Do these strategies provide cost-effective healthcare paradigms?

In this review we describe the biochemical evidence linking observed genetic mutations with altered drug metabolism. We also provide a list of drugs for which metabolism is affected by genetics and recommend genotyping strategies that are most readily integrated into a clinical laboratory environment. The rationale behind this approach is that genotyping/phenotyping will lead to increased therapeutic efficiency, improved patient outcomes, and thus more cost-effective medicine.


   enzymes involved in drug metabolism
Top
 Abstract
 Introduction
 enzymes involved in drug...
cytochrome p-450 superfamily of...
 phenotyping vs genotyping
 CYP Enzymes
 References
 
Enzymes involved in drug metabolism are classified as either phase I (oxidative) or phase II (conjugative). These two reaction types often complement each other in function. For example, through catalysis of oxygenation, oxidation, reduction, and hydrolysis reactions, phase I enzymes generate functional groups that may subsequently serve as a site for conjugation to glucuronic acid, sulfate, or glutathione, catalyzed by phase II enzymes (1). Table 1 lists the principal phase I and phase II enzymes found in human liver.


View this table:
[in this window]
[in a new window]
  		Table 1. Principle phase I and phase II enzymes in human liver.


   cytochrome p-450 superfamily of heme thiolate proteins
Top
 Abstract
 Introduction
 enzymes involved in drug...
 cytochrome p-450 superfamily of...
phenotyping vs genotyping
 CYP Enzymes
 References
 
The major route of phase I drug metabolism is oxidation by cytochrome P-450 (CYP) mixed-function monooxygenases located within the endoplasmic reticulum. Thirty or more different forms of P-450s have been characterized in humans, each with distinct catalytic specificity and unique regulation. Because of the diversity of this family of heme thiolate proteins, a nomenclature system based on sequence identity has been developed to assist in unifying scientific efforts in this area and to provide a basis for nomenclature of newly recognized members of this gene superfamily. For example, CYP1A2 is isoform 2 of subfamily A included in the 1 CYP family. The gene encoding this enzyme is designated CYP1A2. Assignment of family and subfamily categories is based on amino acid sequence homology. For a review of the most recent nomenclature, refer to Nelson et al. (2). Table 2 lists the specific and relative content of the individual CYP enzymes involved in drug metabolism in humans.


View this table:
[in this window]
[in a new window]
  		Table 2. Specific content of individual CYP enzymes involved in oxidation of foreign compounds in human liver.


   phenotyping vs genotyping
Top
 Abstract
 Introduction
 enzymes involved in drug...
 cytochrome p-450 superfamily of...
 phenotyping vs genotyping
CYP Enzymes
 References
 
The objective is to distinguish the three classes of metabolizers (PM, EM, or UEM) prospectively to allow for appropriate modifications in patient management. The specific drug metabolism phenotype may be identified by either phenotyping or genotyping approaches. Phenotyping is accomplished by administration of a test drug (the metabolism of which is known to be solely dependent on the function of a specific drug-metabolizing enzyme) followed by measurement of the metabolic ratio (MR, defined as the ratio of drug dosage or unchanged drug to metabolite measured in serum or urine). Table 3 lists the polymorphic CYP enzymes and test drugs that may be used for in vivo measurement of their activity (phenotyping). Defining the individual's phenotype, relative to a reference substrate, allows the drug metabolism phenotype for other substrates of that enzyme to be predicted (3). In pharmacokinetic studies, phenotyping has the advantage over genotyping in revealing drug–drug interactions or defects in the overall process of drug metabolism. Phenotyping has several drawbacks in that it is hampered by complicated protocols of testing, risks of adverse drug reactions (4), problems with incorrect phenotype assignment due to coadministration of drugs (5)(6), and confounding effects of disease (7)(8).


View this table:
[in this window]
[in a new window]
  		Table 3. Drug substrates for CYP phenotyping.

Genotyping involves identification of defined genetic mutations that give rise to the specific drug metabolism phenotype. These mutations include genetic alterations that lead to overexpression (gene duplication), absence of an active protein product (null allele), or production of a mutant protein with diminished catalytic capacity (inactivating allele). One method that has proven useful in screening for genetic mutations associated with altered metabolism of drugs and (or) cancer susceptibility (9)(10) is amplification of a specific region of the gene of interest by PCR followed by digestion of the amplified DNA product with restriction endonucleases. Restriction endonucleases have the capacity to digest DNA with a high degree of nucleotide sequence specificity. Thus, point mutations within the recognition sequence of a specific restriction endonuclease may be detected through determining whether the DNA of interest serves as a substrate for that endonuclease. These studies are routinely carried out by comparing the size of digestion products generated from a DNA substrate amplified from control subject DNA vs study subject DNAs. Differences in the size of DNA fragments generated as a result of endonuclease digestion is commonly referred to as a restriction fragment length polymorphism (RFLP) (11)(12)(13). The size of the digestion products are easily evaluated by agarose gel electrophoresis with ethidium bromide staining and UV transillumination.

A second approach for detection of specific mutations within a gene of interest is through allele-specific PCR amplification where oligonucleotides specific for hybridizing with the common or variant alleles are utilized in parallel amplification reactions. Only the oligonucleotide that precisely hybridizes to the target sequences produces an amplification product. Analysis for the presence or absence of the appropriate amplified product is also accomplished by agarose gel electrophoresis. The best example of this approach is the identification of the A and B variants of CYP2D6 (14), which is described in more detail later in this article. These genotyping methods require small amounts of blood or tissue (15), are not affected by underlying disease or by drugs taken by the patient, and provide results within 48–72 h, allowing for rapid intervention.


   CYP Enzymes
Top
 Abstract
 Introduction
 enzymes involved in drug...
 cytochrome p-450 superfamily of...
 phenotyping vs genotyping
 CYP Enzymes
References
 
The following sections focus on the various polymorphic CYP enzymes with emphasis on clinical implications and testing strategies.

cyp2d6
CYP2D6 is by far the best characterized P-450 enzyme that demonstrates polymorphic expression in humans. The earliest evidence of polymorphic expression was identified during clinical trials of the antihypertensive drug debrisoquin (1). Since then, debrisoquin hydroxylase activity has been the standard for phenotypic analysis of this polymorphism in clinical studies and has revealed individuals of the EM, PM, and UEM phenotypes. Several additional drugs have been identified for use in phenotyping studies, including sparteine and dextromethorphan, and more recently propafenone (Table 3Up ) (16) .

Clinical significance of the CYP2D6 genetic polymorphism.
The clinical significance of this drug metabolism polymorphism has been the subject of numerous clinical studies over the past decade (2)(14)(17)(18)(19). Brosen and Gram (19) suggest that clinical significance can be evaluated by asking the following questions: Does the kinetics of active principle of a drug depend significantly on a specific enzyme? Is the resulting pharmacokinetic variability of any clinical importance? Can the variation in response be assessed by direct clinical or paraclinical measurements? On the basis of these criteria, significance exists for those drugs for which plasma concentration measurements are considered useful and for which the elimination of the drug and (or) its active metabolite is mainly determined by an enzyme (e.g., CYP2D6) whose polymorphic expression has been characterized. On the basis of the above criteria, frequency of use, and a narrow therapeutic index, Brosen and Gram concluded that the polymorphism of CYP2D6 was of clinical significance for tricyclic antidepressants, certain neuroleptics, and antiarrhythmics (20). Table 4 lists clinically important CYP2D6 substrates, and use of these drugs may form a basis for requirement of pharmacogenetic analysis.


View this table:
[in this window]
[in a new window]
  		Table 4. Substrates of CYP4502D6.

Examples involving tricyclic antidepressants and cardiac antiarrhythmics.
The N-demethylation of imipramine and most likely amitriptyline to their respective pharmacologically active desmethyl metabolites, desipramine and nortriptyline, is catalyzed primarily by CYP2C19 (21)(22) and CYP1A2 (5), whereas the 2-hydroxylation of desipramine and nortriptyline, which results in pharmacologically inactive metabolites, is catalyzed by CYP2D6 (20). The polymorphic nature of the CYP2D6-dependent 2-hydroxylation is evident in the wide range (Table 5 ) of elimination half-lives of the desmethyl metabolites vs the elimination half-lives of the parent drugs, which reflects the sum of the N-demethylation and 2-hydroxylation processes (23). For tricyclic antidepressants, both the PM and UEM phenotypes of CYP2D6 are at risk of adverse reactions. PM individuals given standard doses of these drugs will develop toxic plasma concentrations, potentially leading to unpleasant side effects including dry mouth, hypotension, sedation, and tremor, or in some cases life-threatening cardiotoxicity (24). Table 6 demonstrates the relative concentrations of imipramine and its active metabolite desipramine in EM vs PM subjects treated with identical dosing regimens of imipramine. Note that the absolute concentrations of both the parent drug (imipramine) as well as the desmethyl metabolite (desipramine) are greater in PM individuals and that because of the accumulation of the desmethyl metabolite, the ratio of parent drug to metabolite is much lower in PM individuals. Administration of CYP2D6 substrates to UEM individuals may result in therapeutic failure because plasma concentrations of active drug at standard doses are far too low (24). Although these patients may be successfully treated with higher concentrations of these drugs, metabolites from alternate metabolic pathways may accumulate and contribute to toxicity (25). The clinical presentation of UEM and PM patients are at times similar, leading to confusion in understanding the basis for the adverse drug reaction. Because of lack of dose individualization, patients are subject to recurrent depressive episodes and may not respond to treatment (24)(26). Although the relation between dose and therapeutic response may be addressed through traditional therapeutic drug monitoring, prior knowledge of the drug metabolism potential, through genotyping techniques, could predict the outcome and allow for appropriate dosage adjustments to prevent the initial adverse event from occurring (27)(28)(29)(30). Patients requiring treatment with antidepressant or antipsychotic substrates of CYP2D6 may begin the normal treatment regimen. Because of the long half-life of these drugs, toxic concentrations may take 5 to 7 weeks to develop (31). Likewise, patients requiring antiarrhythmics could not await the results of genotyping but may be initiated on the treatment regimen while genotyping results are pending, and adjustments in dosing or therapeutic selection may be made once the data are available. In this case, knowledge of the CYP2D6 genotype would assist in anticipation of toxic effects or in the appropriate selection of an alternative therapeutic on the basis of knowledge of the metabolism of the alternative drug. Thus, there appears to be ample time to evaluate the genotype of the patient and apply the information to dose adjustments or change in therapeutic to a drug that is not a substrate for CYP2D6 before the onset of adverse symptoms.


View this table:
[in this window]
[in a new window]
  		Table 5. Range of elimination half-lives reported for amitriptyline, nortriptyline, imipramine, and desipramine in humans.


View this table:
[in this window]
[in a new window]
  		Table 6. Relative disposition of imipramine and desipramine in individuals of PM vs EM phenotypes of debrisoquin hydroxylase.

Biochemical basis for genetically determined CYP2D6 phenotypes.
After identification of CYP2D6 as the gene responsible for expression of debrisoquine hydroxylase, methods became available for analyzing specific mutations of this gene associated with the PM and UEM phenotypes with respect to debrisoquin. The CYP2D6 gene resides in the CYP2D6-8 cluster on chromosome 22 in association with the CYP2D7P and CYP2D8Ppseudogenes (32)(33). Fig. 1 shows the structure of the CYP2D6-8 gene cluster and its most common polymorphic haplotypes evaluated by digestion of genomic DNA with the restriction endonuclease XbaI. A 29-kb XbaI fragment indicates the presence of the normal locus with two pseudogenes CYP2D7P and CYP2D8P and one copy of the CYP2D6 gene (18). The 42- and 44-kb alleles involve gene duplications of either an active gene (42 kb) or inactive gene (44 kb), and the 11.5-kb XbaI fragment indicates deletion of the CYP2D6 gene (26). The first evidence of genetic polymorphism of the CYP2D6 gene was observation of homozygous CYP2D6 gene deletion (CYP2D6D allele) by Southern transfer in individuals demonstrating the PM phenotype for debrisoquine hydroxylase (19)(34).



View larger version (22K):
[in this window]
[in a new window]
  		Figure 1. Structure of the CYP2D6-8 gene cluster and its most common alleles.

Schematic representation of the structural organization of chromosome 22 including the relative position of CYP2D6, CYP2D7, and CYP2D8P. XbaI sites used for mapping are indicated by x. Numbers indicate the XbaI restriction fragment length of each allele in kb. Mutations associated with each allele are indicated by a vertical slash under the affected gene. Modified with permission from Spurr et al., Methods Enzymol 1991;206:149–90 and Academic Press.

Subsequent to identification of the CYP2D6D allele, studies were initiated to correlate the presence of the CYP2D6D allele with the PM phenotype. Although these studies were successful in correlating the CYP2D6D genotype with the PM phenotype, they also revealed individuals with the PM phenotype who were apparently homozygous for the wild-type (wt) allele. Thus there remained one or more unidentified mutations in the CYP2D6 gene that gave rise to the PM phenotype. Several CYP2D6 null and inactivating alleles have since been identified. These allele designations, biochemical alterations, and incidence of occurrence are listed in Table 7 . To date, a total of six null alleles, A, B, D, E, F, and T, and an additional two variants that encode protein products with decreased catalytic activity (CYP2D6C and J) have been characterized.


View this table:
[in this window]
[in a new window]
  		Table 7. Mutant CYP2D6 alleles.

The UEM phenotype results from duplication or amplification of a unique CYP2D6L allele on chromosome 22. The most common gene duplication results in two copies of the CYP2D6L gene on one chromosome (CYP2D6L2) (Fig. 1Up , 42-kb XbaI fragment) (35). However, a 12-fold amplification of this gene has been observed (21). The gene product apparently has similar catalytic activity to the wt enzyme, and thus increased activity in individuals with this allele is directly related to gene dose. Refer to Agundez et al. (36) for an elegant demonstration of this gene dose effect. This unique allele has only been identified in the heterozygous state and results in overexpression of CYP2D6 enzyme. This overexpression has a dramatic effect on the metabolism and elimination of CYP2D6 substrates and may lead to subtherapeutic serum concentrations of parent drug at standard doses (37)(38). Therefore, subjects of the UEM phenotype may require megadoses of certain drugs (particularly those drugs metabolized to inactive compounds) to obtain therapeutic efficacy (23).

Prevalence of CYP2D6 phenotypes and associated genotypes.
There are significant interethnic differences in the prevalence of the PM phenotype of debrisoquine hydroxylase. For example, in North American and European Caucasian populations, the prevalence of the PM phenotype is 5–10% (39)(40)(41). In contrast, the prevalence is 1.8% in American blacks (42), 1.2% in native Thai (43), 1.0% in Chinese (44), 2.1% in a native Malay population (45), and apparently absent in the Japanese population (46). In the Caucasian population, the PM phenotype is the consequence of various mutant CYP2D6 genotypes. Table 8 lists the frequency of the various CYP2D6 genotypes and their associated phenotypes in Caucasian individuals. Thirty-five percent of Caucasians are heterozygous for an inactivating mutant CYP2D6 allele and thus may demonstrate an intermediate phenotype. The prevalence of the UEM phenotype in Caucasians is 7%, and is the result of the CYP2D6L2 allele in >95% of cases (20). This phenotypic group has been relatively ignored by the literature. However, the clinical consequence of the UEM phenotype may be as significant as PM phenotype in terms of therapeutic efficacy with respect to certain therapeutics, including the tricyclic antidepressants.


View this table:
[in this window]
[in a new window]
  		Table 8. Prevalence of the most common CYP2D6 genotypes among unrelated Caucasian subjects.

Methods for determining CYP2D6 genotypes.
The CYP2D6 gene deletion can only be detected through RFLP analysis of genomic DNA and Southern transfer by using a cDNA probe. Digestion of genomic DNA with the restriction endonuclease XbaI followed by Southern transfer yields either a >=29-kb fragment in subjects without the deletion or an 11.5-kb fragment in individuals with the CYP2D6D allele (Fig. 1Up ). The labor intensity of this approach makes this method much less attractive to clinical application compared with PCR-based methods. The most common null allele is CYP2D6B, representing 29% of all CYP2D6 alleles and 70% of the null CYP2D6 alleles. The second most common null allele is the D allele, followed by the A allele, representing 12% and 3% of all CYP2D6 alleles respectively. Analysis of the A and B variants is typically carried out with an allele-specific amplification method developed by Heim and Meyer (14), which includes an initial amplification strategy to specifically amplify a region of the CYP2D6 gene, followed by allele-specific amplification to identify the A and B mutant alleles. The second step of this approach has recently been automated on the basis of ligase chain reaction technology (47), which should advance efforts to reduce the time and expense associated with testing. Studies that have utilized this method to prospectively identify poor metabolizers have demonstrated 95% sensitivity and 100% specificity, and the phenotype of 99% of randomly selected subjects can be determined by this genotyping strategy (12)(48)(49). Recently, a more rapid method has been developed by Douglas et al. that allows for the simultaneous evaluation of both A and B variants with a single amplification step and restriction digest (50). A limitation of this approach is that it requires careful interpretation of the restriction digestion patterns. A model of the diagnostic sensitivity of the various genotyping approaches for prediction of phenotype is demonstrated in Table 9 (25). It is important to recognize that the UEM phenotype or genotype was not taken into consideration in this model. From these data, it is evident that for clinical purposes of identifying phenotypically PM individuals, analysis of the A and B variants will predict the PM phenotype with >95% accuracy and the EM phenotype with >99% accuracy (40)(41). It is therefore highly likely that these analyses could significantly affect the risk–benefit ratio and increase the therapeutic efficacy of drugs that are substrates for CYP2D6. When analysis of the A and B variants fail to corroborate the apparent phenotype, analysis for the CYP2D6D gene deletion would be the next logical step. The complete absence of an amplification product during the analysis of the A and B variants would lend support for confirming the gene deletion by Southern transfer. However, in the event the A and B analysis yields the apparently wt PCR product, the best approach may be to pursue characterization of the other null or inactivating alleles for which PCR methods are available (Table 7Up ). On the basis of the prevalence of the various mutations, analysis of the T allele may be of considerable value (51).


View this table:
[in this window]
[in a new window]
  		Table 9. Sensitivity of various genotyping procedures for detection of PMs.1

The CYP2D6 gene duplication may be detected by observation of a 42-kb fragment in the XbaI RFLP analysis of genomic DNA (Fig. 1Up ) (21). An allele-specific PCR method has been developed and remains to be validated as a screening tool for diagnosis of the UEM phenotype (23).

Interpretation of genotyping data.
Interpretation of the PCR-based analysis of CYP2D6 genotype requires the laboratorian to be aware that individuals homozygous for the gene deletion will not yield a PCR product in standard assays for the various mutations. Homozygosity for the gene deletion is rare and accounts for <4% of PMs (0.4% of the total population) (25). Heterozygotes for the gene deletion may appear to be homozygous for the wt allele. On the basis of the metabolic criteria for phenotype assignment, individuals heterozygous for the gene deletion are categorized as EM and thus there is concordance between the observed genotype with phenotype assignment. Individuals heterozygous for the gene deletion and one of the other inactivating alleles will be correctly classified as PM. However, the true genotype can be proven only through the Southern transfer approach. For clinical purposes this is not necessary, but Southern transfer must be carried out in all population studies designed to characterize allele frequencies.

Clinical significance of heterozygosity for an inactivating or null allele.
In the early process of correlating genetic mutations with metabolic phenotype, the phenotyping methods did not discriminate homozygous vs heterozygous genotypes. As a result, it has been accepted that these mutations are recessive and that only homozygotes for the mutant allele will demonstrate altered pharmacokinetics for a given drug substrate, and individuals genotyped as heterozygotes are empirically classified as EMs. This concept has been challenged (46)(52) and may await more sensitive techniques to identify phenotypic differences in drug metabolism between the homozygous common and heterozygous mutant genotypes. The clinical consequence of the heterozygous genotype remains to be determined. As demonstrated in Fig. 2 , the metabolic ratios of heterozygous individuals with respect to the test drug debrisoquin approaches that of the homozygous mutant individuals. The difference in metabolic ratios indicates that the relative concentrations of parent drug in the homozygous vs heterozygous individuals may differ by more than fivefold. Therefore, higher plasma concentrations of CYP2D6 substrates may be expected in heterozygous individuals. Considering that 35% of individuals may be of the heterozygous genotype (Table 8Up ), this fact may have significant bearing on the frequency of unpleasant side effects of CYP2D6 substrates. The relative concentrations of therapeutic substrates in homozygous vs heterozygous individuals at a fixed dose has not been determined in a study designed to accurately detect all of the mutant CYP2D6 alleles. In a recent study evaluating schizophrenic patients treated with neuroleptics, there was a statistically significant correlation between the debrisoquin MR in CYP2D6 heterozygotes and the severity of extrapyramidal side effects during neuroleptic treatment (53). Recently Madsen et al. (6) addressed the relation between imipramine metabolism and CYP2D6 polymorphisms. The studies involved dosing healthy Caucasian individuals with imipramine, followed by determination of imipramine metabolism through measurement of imipramine, desipramine, and the 2-hydroxy metabolites of each in urines. The hydroxylation ratios were significantly different between the homozygous and heterozygous EMs, with the heterozygous EM demonstrating intermediate hydroxylation ratios between the homozygous EM and PM populations. The relative serum concentrations of the drugs were not determined in these studies, nor was the study designed to evaluate the accumulation of drug over time as a result of the heterozygous vs homozygous genotypes.



View larger version (23K):
[in this window]
[in a new window]
  		Figure 2. Distribution of metabolic ratios among various CYP2D6 genotypes.

The relation between metabolic ratio (molar ratio of debrisoquine:4-hydroxydebrisoquine in 0–8-h urine) and genotype within the volunteer group. On the basis of RFLP and PCR analysis, subjects were classed as either homozygous EM (solid), heterozygous EM (striped), or PM (hatched). Taken with permission from ref. 38.

Of interest in the current literature is the interaction of many selective serotonin reuptake inhibitors with the CYP2D6 enzyme. The most notable example of this is fluoxetine (54). Through competition with CYP2D6 substrates, these drugs precipitate a drug-induced PM phenotype (54)(55). One can reasonably predict that the effects of CYP2D6 inhibitors on the metabolism of CYP2D6 substrates would be more pronounced in heterozygous EMs. However, this has not been definitively demonstrated.

cyp2c19
A second well-characterized CYP-related drug metabolism polymorphism in humans is associated with the 4'-hydroxylation of the S-enantiomer of the anticonvulsant mephenytoin (56). Like the 2D6 isoenzyme, specific genetic mutations of the gene encoding S-mephenytoin hydroxylase lead to a PM phenotype with respect to several common therapeutic drugs listed in Table 10 . In contrast to the debrisoquine polymorphism, no UEM phenotype has been demonstrated for this polymorphic enzyme.


View this table:
[in this window]
[in a new window]
  		Table 10. Pharmaceutical substrates of CYP2C19.

The poor metabolizer phenotype is inherited in an autosomal recessive manner (57). Like the CYP2D6 polymorphisms, there are significant interethnic differences in the prevalence of the PM phenotype. The PM phenotype occurs in 2–5% of Caucasian (57) and black Zimbabwean Shona populations (58), and 18–23% in Oriental populations (59). Until recently the specific CYP associated with this polymorphic drug metabolism was unknown. Recent studies have shown that CYP2C19 is the enzyme primarily responsible for the 4'-hydroxylation of mephenytoin in humans (60)(61). The extensive metabolizer phenotype comprises both the homozygous dominant and heterozygous recessive genotypes. The principal genetic defect in poor metabolizers of S-mephenytoin is a single G->A mutation in exon 5, which creates a novel aberrantly spliced CYP2C19 mRNA. Translation of this mRNA transcript results in the production of an inactive truncated protein. This null allele of CYP2C19 is designated m1. The correlation of this allele with the S-mephenytoin metabolism phenotype in unrelated Swiss and American Caucasians was evaluated with a PCR method. The m1 mutation accounted for 75% of CYP2C19 alleles in the Caucasian PMs genotyped. The sensitivity of this test is 63%, with 100% specificity. In this same study including unrelated Japanese subjects, the mutant gene m1 accounted for 74% of the alleles (62) in the Japanese PM group. The sensitivity of this test is 59% with 100% specificity. The m1 allele did not account for all PMs regardless of race. Further evaluation of PM subjects with an apparently wt CYP2C19 genotype revealed a second mutant allele, designated CYP2C19 m2, unique to the Japanese individuals. This mutation is a G636->A mutation, which results in a premature stop codon. All Japanese PMs whose phenotype could not be explained by the m1 mutations were found to be either homozygous or heterozygous (m1m2) for the mutant alleles. All Japanese EMs had at least one wt allele. Thus, the m1, o2 mutations accounted for all of the alleles in Japanese PMs, and the combined sensitivity and specificity of the m1 and m2 mutations for identification of the PM phenotype in Japanese subjects is 100% (63). The m2 allele has also been identified in the Chinese population, and in conjunction with the m1 allele accounted for 100% of alleles in Chinese PMs. The frequency of PMs in this population was ~11% (64), which reflects the frequency observed in the Japanese population. In a Shona population in Zimbabwe, CYP2C19 m1 genotyping correctly identified three of four poor metabolizers in a study cohort of 84 subjects. No subjects were inappropriately predicted as a PM and one individual was incorrectly identified as an EM on the basis of the genotyping data. Thus, in this population the m1 genotyping method demonstrated 75% sensitivity and 100% specificity. Analysis of the m2 mutation was not carried out.

The clinical consequence of the CYP2C19 polymorphism has not been fully described. Considering the relative abundance of this isoenzyme (Table 2Up ) and the identification of an increasing number of pharmaceutical substrates (Table 10Up ), it is likely that the clinical relevance of this drug metabolism polymorphism will be revealed in the near future. Of particular interest are the tricyclic antidepressant substrates, which are clinically important with respect to the CYP2D6 polymorphism, and the antimalarial prodrugs proguanil and chlorproguanil, which require CYP2C19-dependent bioactivation for therapeutic efficacy (65)(66). The proton pump inhibitor omeprazole, used in the short-term treatment of active duodenal ulcer and esophagitis, is primarily metabolized by the CYP2C19 enzyme, and thus its metabolism is subject to this genetic polymorphism. Because of the wide therapeutic range of this drug, severe toxicity is not observed in PMs, and omeprazole may serve as a safer alternative for CYP2C19 phenotyping compared with S-mephenytoin (67). One recent study on the causes of drug-induced hepatitis revealed a relation between Atrium® (a combination preparation of phenobarbital, febarbamate, and difebarbamate)-induced hepatitis and the PM phenotype of mephenytoin hydroxylase (68). The influence of CYP2D19 genotype must be further evaluated in this context because of the observation that liver disease can phenocopy the S-mephenytoin/omeprazole PM phenotype in genotypically wt individuals (8).

The following two sections provide an overview of two additional CYP enzymes for which there is some evidence of genetic heterogeneity in the population but the exact molecular basis and clinical significance have not been defined.

cyp2e1
CYP2E1 is responsible for the metabolism and bioactivation of many procarcinogens (69) and certain drugs, including ethanol and acetaminophen (70). CYP2E1 is encoded by a single gene in humans located on chromosome 10 (71). Two alleles of this gene have been identified in humans, C and c2. The nomenclature for these alleles differ from the system used to describe the CYP2D6 system. For each location on the gene where polymorphic mutations have been observed, there is a designation for the wt allele as well as for the mutant allele. For example, the C allele designates a single point mutation located in intron 6 of CYP2E1 that can be identified by RFLP analysis after digestion by DraI endonuclease. The common wt allele with respect to this mutation is designated D. The absence of the rare allele (C) has been associated with lung cancer in a Japanese case–control study (72). Similar studies involving Caucasians and African Americans in the US found no relation between allelic variance of the DraI polymorphism and increased risk of lung cancer (73). An additional RFLP revealed by RsaI endonuclease is located in the 5' transcriptional regulatory region of this gene (74). The mutation disrupts a RsaI restriction endonuclease site, generating the rare allele c2 (RsaI-). This polymorphic RsaI site is located within the binding site (cis-responsive element) for the liver-specific transcription factor HNF1 (LF-B1). Expression of chimeric chloramphenicol acetyltransferase gene constructs containing the native vs c2 CYP2E1 promoter sequences in situ demonstrated overexpression of the reporter gene construct containing the promoter from the c2 CYP2E1 allele (75). In addition, the expression of CYP2E1 mRNA was increased relative to controls in peripheral lymphocytes of individuals either homozygous or heterozygous for the c2 mutant CYP2E1 allele (76). Thus, this mutation may potentially lead to increased expression of functional protein and result in increased metabolism of CYP2E1 substrates.

Chlorzoxazone, a skeletal muscle relaxant, is reported to provide a measure of CYP2E1 activity in vivo (75). The use of this in vivo probe of CYP2E1 remains unsettled (77). However, the majority of studies favor its use. Recent studies involving this probe to correlate chlorzoxazone 6-hydroxylase activity (phenotype) with the known polymorphisms of CYP2E1 failed to demonstrate a relation between altered CYP2E1 genotype and the capacity to metabolize chlorzoxazone. Further, the distribution of oral and fractional clearance values was not bimodal, suggesting that a single CYP2E1 allele is predominant in the population studied. An unfortunate limitation of this study is that no individuals homozygous for the c2 variant were identified in the study cohort of Caucasian subjects (78). The lack of c2 alleles identified in this study is most likely due to the interracial differences in the prevalence of the c2 allele, which was first described in the Japanese (76). Typically heterozygotes have the EM phenotype, which cannot be discriminated by phenotyping approaches from the homozygous common genotype. Additionally, it is not clear that phenotyping approaches are a sensitive means of identifying the UEM phenotype, which would be the expected consequence of the c2 mutation. In summary, there is some evidence of genetic polymorphism of CYP2E1 in the human population; however, the molecular mechanisms remain to be further characterized.

cyp3a
The CYP3A subfamily in humans comprises the 3A3, 3A4, and 3A5 isoenzymes in adults and the 3A7 isoenzyme in fetal liver. The predominant 3A isoenzyme in adult human liver is the 3A4 isoenzyme, which accounts for 20–40% of the total hepatic CYP in humans (Table 3Up ) (79)(80)(81). Clinically relevant substrates for the CYP3A isoenzymes are listed in Table 11 .


View this table:
[in this window]
[in a new window]
  		Table 11. Pharmaceutical substrates of CYP3A.

Although a large degree of interindividual variability in the expression of CYP3A isoenzymes has been demonstrated in human liver (>20-fold), no genetic basis for this polymorphic expression has been defined to date. However, in a recent report of a human liver graft that demonstrated undetectable CYP3A protein by Western blotting, the recipient developed renal failure with 24 h of initiating FK-506 therapy, which resolved upon withdrawal of FK-506. After treatment with corticosteroids and FK-506, liver biopsy revealed the expression of CYP3A (82). Expression may have been induced in the recipient subsequent to combined FK-506 and corticosteroid therapy (83)(84). However, CYP3A4 expression was not evaluated subsequent to discontinuation of FK-506 and corticosteroids to demonstrate reversal in expression. The absence of CYP3A in a human liver specimen has been previously reported (85); the molecular mechanism(s) of this null variant have not been elucidated, but the ability to induce expression may suggest a mutation resident in cis elements of the genes of trans-acting transcription factors responsible for the basal expression of this gene.

In addition to the potential for genetic variability in expression or activity, CYP3A activity is known to be induced on exposure to barbiturates and glucocorticoids and inhibited by macrolide antibiotics such as erythromycin that may influence variability in the in vivo estimates of CYP3A activity. In addition, extrahepatic expression of CYP3A can influence phenotyping approaches (86), depending on the route of test drug administration (87).

In this review we addressed several questions concerning the clinical application of genotyping/phenotyping data as they apply to the clinical laboratory as well as to the clinical practitioner. Genotyping may be indicated in each instance when the therapeutic of choice is a substrate for a polymorphic enzyme. Alternatively, genotyping is indicated when individuals demonstrate suboptimal response to drugs that are substrates for polymorphic enzymes. The advantage of combining genotyping with therapeutic drug monitoring is that genotyping can predict the PM or UEM drug metabolism phenotypes, and this information can be used a priori in dose adjustment or selection of an alternative therapeutic that is not a substrate for the polymorphic enzyme. The cost/healthcare effectiveness of these paradigms has not been extensively studied. Although there would be considerable cost associated with screening all individuals before dosing with these drugs, this cost may be offset by a reduction in costs associated with toxic episodes or therapeutic failure and subsequent intervention.


   Acknowledgments
 
This work was supported in part by HHS grants NIH P20 ES0632 and NIH HL 36172.


   Footnotes
 
1 Nonstandard abbreviations: EM; extensive metabolizer; PM, poor metabolizer; UEM; ultraextensive metabolizer; CYP; cytochrome P-450; MR, metabolic ratio; RFLP; restriction fragment length polymorphism; and wt, wild type.


   References
Top
 Abstract
 Introduction
 enzymes involved in drug...
 cytochrome p-450 superfamily of...
 phenotyping vs genotyping
 CYP Enzymes
 References
 

  1. Poulsen HE, Loft S. The impact of genetic polymorphisms in risk assessment of drugs. Arch Toxicol Suppl 1994;16:211-212. [Medline] [Order article via Infotrieve]
  2. Nelson DR, Kamataki T, Waxman DJ, Guengerich P, Estabrook RW, Feyereisen R, et al. The P450 superfamily: update on new sequences, gene mapping, accession numbers, early trivial names of enzymes and nomenclature. DNA Cell Biol 1993;12:1-51. [ISI][Medline] [Order article via Infotrieve]
  3. Gonzalez FJ, Skoda RC, Dimura S, Umeno M, Zanger UM, Nebert DW, et al. Characterization of the common genetic defect in humans deficient in debrisoquin metabolism. Nature 1988;331:442-446. [Medline] [Order article via Infotrieve]
  4. Relling MV, Cherrie J, Schell MJ, Petros WP, Meyer WH, Evans WE. Lower prevalence of the debrisoquin oxidative poor metabolizer phenotype in American black versus white subjects. Clin Pharmacol Ther 1991;50:308-313. [ISI][Medline] [Order article via Infotrieve]
  5. LLerna A, Herraiz AG, Cobaleda J, Johansson I, Dahl ML. Debrisoquin and mephenytoin hydroxylation phenotypes and CYP2D6 genotype in patients treated with neuroleptic and antidepressant agents. Clin Pharmacol Ther 1993;54:606-611. [ISI][Medline] [Order article via Infotrieve]
  6. Madsen H, Kramer Nielsen K, Brosen K. Imipramine metabolism in relation to the sparteine and mephenytoin oxidation polymorphisms—a population study. Br J Clin Pharmacol 1995;39:433-439. [ISI][Medline] [Order article via Infotrieve]
  7. Corporaso N, Tucker MA, Hoover R, Hayes RB, Pickel LW, Issaq HJ, et al. Lung cancer and the debrisoquin metabolic phenotype. J Natl Cancer Inst 1990;82:1264-1274. [Abstract/Free Full Text]
  8. Rost KL, Brockmoller J, Esdorn F, Roots I. Phenocopies of poor metabolizers of omeprazole caused by liver disease and drug treatment. J Hepatol 1995;23:268-277. [ISI][Medline] [Order article via Infotrieve]
  9. Sheilds PG, Harris CC. Molecular epidemiology and the genetics of environmental cancer. JAMA 1991;266:681-687. [Abstract]
  10. Shields PG. Inherited factors and environmental exposures in cancer risk. J Occup Med 1993;35:34-41. [ISI][Medline] [Order article via Infotrieve]
  11. Harris CC. Interindividual variation among humans in carcinogen metabolism, DNA adduct formation and DNA repair. Carcinogenesis (Oxford) 1989;10:1563-1566. [Abstract/Free Full Text]
  12. Nakachi K, Imai K, Hayashi S, Watanabe J. Genetic susceptibility to squamous cell carcinoma of the lung in relation to cigarette smoking dose. Cancer Res 1991;51:5177-5180. [Abstract/Free Full Text]
  13. Kawajiri K, Nakachi K, Imai K, Yoshii A, Shinoda N, Watanabe J. Identification of genetically high risk individuals to lung cancer by DNA polymorphisms of the cytochrome P450IA1 gene. FEBS Lett 1990;263:131-133. [ISI][Medline] [Order article via Infotrieve]
  14. Heim M, Meyer UA. Genotyping of poor metabolizers of debrisoquin by allele-specific PCR amplification. Lancet 1990;336:529-532. [ISI][Medline] [Order article via Infotrieve]
  15. Bowie LJ, Reddy PL, Nagabhushan M, Sevigny P. Detection of {alpha}-thalassemias by multiplex polymerase chain reaction. Clin Chem 1994;40:2260-2266. [Abstract/Free Full Text]
  16. Botsch S, Heinkele G, Meese CO, Eichelbaum M, Kroemer HK. Rapid determination of CYP2D6 phenotype during propafenone therapy by analyzing urinary excretion of propafenone glucuronides. Eur J Clin Pharmacol 1994;46:133-135. [ISI][Medline] [Order article via Infotrieve]
  17. Gonzalez FJ, Idle JR. Pharmacogenetic phenotyping and genotyping—present status and future potential. Clin Pharmacokinet 1994;26:59-70. [ISI][Medline] [Order article via Infotrieve]
  18. Eichelbaum M, Gross AS. The genetic polymorphism of debrisoquine/sparteine metabolism—clinical aspects. Pharmacol Ther 1990;46:377-394. [ISI][Medline] [Order article via Infotrieve]
  19. Brosen K, Gram LF. Clinical significance of the sparteine/debrisoquine oxidation polymorphism. Eur J Clin Pharmacol 1989;36:537-547. [ISI][Medline] [Order article via Infotrieve]
  20. Brosen K, Gram LF. First-pass metabolism of imipramine and desipramine: impact of the sparteine oxidation phenotype. Clin Pharmacol Ther 1988;43:400-406. [ISI][Medline] [Order article via Infotrieve]
  21. Skjelbo E, Gram LF, Brosen K. The N-demethylation of imipramine correlates with the oxidaton of S-mephenytoin (S/R-ratio)—a population study. Br J Clin Pharmacol 1993;35:331-334. [ISI][Medline] [Order article via Infotrieve]
  22. Breyer-Pfaff U, Pfandl B, Nill K, Nusser E, Monney E, Jonzier-Perey M, et al. Enantioselective amitriptyline metabolism in patients phenotyped for two cytochrome P450 isoenzymes. Clin Pharmacol Ther 1992;52:350-358. [ISI][Medline] [Order article via Infotrieve]
  23. Ellenhorn MJ, Barceloux DG. Medical toxicology; diagnosis and treatment of human poisoning 1988:405 Elsevier New York. .
  24. Balant-Gorgia AE, Balant LP, Garrone G. High blood concentrations of imipramine or clomipramine and therapeutic failure: a case report study using drug monitoring data. Ther Drug Monit 1989;11:415-420. [ISI][Medline] [Order article via Infotrieve]
  25. Bertilsson L, Aberg-Wistedt A, Gustafsson LL, Nordin C. Extremely rapid hydroxylation of debrisoquine: a case report with implication for treatment with nortriptyline and other tricyclic antidepressants. Ther Drug Monit 1985;7:478-480. [ISI][Medline] [Order article via Infotrieve]
  26. Gaedigk A, Blum M, Gaedigk R, Eichelbaum M, Meyer UA. Deletion of the entire cytochrome P4502D6 gene as a cause of impaired drug metabolism in poor metabolizers of the debrisoqine/sparteine polymorphism. Am J Hum Genet 1991;48:943-950. [ISI][Medline] [Order article via Infotrieve]
  27. Horowitz LN, Morganroth J. Second generation antiarrhythmic agents, have we reached antiarrhythmic nirvana?. J Am Coll Cardiol 1987;9:459-463. [ISI][Medline] [Order article via Infotrieve]
  28. Whitford EG, McGovern B, Schoenfeld MH, Garan H, Newell JB, McElroy M, et al. Long-term efficacy of mexiletine alone and in combination with class Ia antiarrhythmic drugs for refractory ventricular arrhythmias. Am Heart J 1988;115:360-366. [ISI][Medline] [Order article via Infotrieve]
  29. Buchert E, Woosley RL. Clinical implications of variable antiarrhythmic drug metabolism. Pharmacogenetics 1992;2:2-11. [ISI][Medline] [Order article via Infotrieve]
  30. Haefeli WE, Bargetxi MJ, Follath F, Meyer UA. Potent inhibition of cytochrome P450 IID6 (debrisoquine 4-hydroxylase) by flecainide in vitro and in vivo. J Cardiovasc Pharmacol 1990;15:776-779. [ISI][Medline] [Order article via Infotrieve]
  31. Melstrom B, Bertilsson L, Lou Y-C, Sawe J, Sjoqvist F. Amitriptyline metabolism: relationship to polymorphic debrisoquine hydroxylation. Clin Pharmacol Ther 1988;34:516-520.
  32. Eichelbaum M, Baur MP, Dengler HJ, Osikowska-Evers BO, Tieves G, Zekorn C, et al. Chromosomal assignment of human cytochrome P450 (debrisoquine/sparteine type) to chromosome 22. Br J Clin Pharmacol 1987;23:455-458. [ISI][Medline] [Order article via Infotrieve]
  33. Kimura S, Umeno M, Skoda RC, Meyer UA, Gonzalez FP. The human debrisoquine 4-hydroxylase (CYP2D) locus: sequence and identification of the polymorphic CYP2D6 gene, a related gene, and a pseudogene. Am J Hum Genet 1989;45:889-904. [ISI][Medline] [Order article via Infotrieve]
  34. Zanger UM, Vilbois F, Hardwick JP, Meyer UA. Absence of hepatic cytochrome P450 bufI causes genetically deficient debrisoquine oxidation in man. Biochemistry 1988;27:5447-5454. [Medline] [Order article via Infotrieve]
  35. Johansson I, Lundqvist E, Bertilsson L, Dahl M-L, Sjoqvist F, Sundberg MI. Inherited amplification of an active gene in the cytochrome P450 CYP2D locus as a cause of ultrarapid metabolism of debrisoquine. Proc Natl Acad Sci U S A 1993;90:11825-11829. [Abstract/Free Full Text]
  36. Agundez JA, Ledesma MC, Ladero JM, Benitez J. Prevalence of CYP2D6 gene duplication in a white population. Clin Pharmacol Ther 1995;57:265-269. [ISI][Medline] [Order article via Infotrieve]
  37. Bertilsson L, Dahl ML, Sjoqvist F, Aberg-Wistedt A, Humble M, Johansson I, et al. Molecular basis for rational megaprescribing in ultrarapid hydroxylators of debrisoquine. Lancet 1993;341:63.[ISI][Medline] [Order article via Infotrieve]
  38. Daly AK, Armstrong M, Monkman SC, Idle ME, Idle JR. Genetic and metabolic criteria for the assignment of debrisoquine 4-hydroxylation (cytochrome P4502D6) phenotypes. Pharmacogenetics 1991;1:33-41. [Medline] [Order article via Infotrieve]
  39. Brosen K. Recent developments in hepatic drug oxidation: implications for clinical pharmacokinetics. Clin Pharmacokinet 1990;18:220-239. [ISI][Medline] [Order article via Infotrieve]
  40. Alvan G, Bechtel P, Iselius L, Gundert-Remy U. Hydroxylation polymorphisms of debrisoquin and mephenytoin in European populations. Eur J Clin Pharmacol 1990;39:533-537. [ISI][Medline] [Order article via Infotrieve]
  41. Marinac JS, Foxworth JW, Willsie SK. Dextromethorphan polymorphic hepatic oxidation (CYP2D6) in healthy black American adult subjects. Ther Drug Monit 1995;17:120-124. [ISI][Medline] [Order article via Infotrieve]
  42. Evans WE, Relling MV, Rahman A, McLeod HL, Scott EP, Lin J-S. Genetic basis for a lower prevalence of deficient CYP2D6 oxidative drug metabolism phenotypes in Black Americans. J Clin Invest 1993;91:2150-2154.
  43. Wanwimolruk S, Patamasucon P, Lee EDJ. Evidence for the polymorphic oxidation of debrisoquine in the Thai population. Br J Clin Pharmacol 1990;29:244-247. [ISI][Medline] [Order article via Infotrieve]
  44. Bertilsson L, Lou Y-Q, Du YL, Liu Y, Kuang T-Y, Lia X-M, et al. Pronounced differences between native Chinese and Swedish populations in the polymorphic hydroxylations of debrisoquine and S-mephenytoin. Clin Pharmacol Ther 1992;51:388-397. [ISI][Medline] [Order article via Infotrieve]
  45. Lee EDJ, Nam YP, Hee GN. Oxidation phenotyping in Chinese and Malay populations. Clin Exp Pharmacol Physiol 1988;15:889-891. [ISI][Medline] [Order article via Infotrieve]
  46. Panserat S, Mura C, Gèrard N, Vincent-Viry M, Galteau MM, Jacqz-Aigrain E, et al. DNA haplotype-dependent differences in the amino acid sequence of debrisoquine 4-hydroxylase (CYP2D6): evidence for two major allozymes in extensive metabolizers. Hum Genet 1994;94:401-406. [ISI][Medline] [Order article via Infotrieve]
  47. Hansen TS, Petersen NE, Iitia A, Blaabjerg O, Petersen PH, Hørder M. Robust nonradioactive oligonucleotide ligation assay to detect a common point mutation in the CYP2D6 gene causing abnormal drug metabolism. Clin Chem 1995;41:413-418. [Abstract/Free Full Text]
  48. Broly F, Gaedik A, Heim M, Eichelbaum M, Morike K, Myer UA. Debrisoquine/sparteine hydroxylation genotype and phenotype: analysis of common mutations and alleles of CYP2D6 in a European population. DNA Cell Biol 1991;10:545-558. [ISI][Medline] [Order article via Infotrieve]
  49. Zimmerman T, Schlenk R, Pfaff G, Lach P, Wildfeuer A. Prediction of phenotype for dextromethorphan O-demethylation by using polymerase chain reaction in healthy volunteers. Arzneimittelforschung 1995;45:41-43. [Medline] [Order article via Infotrieve]
  50. Douglas AM, Atchison BA, Somogyi AA, Drummer OH. Interpretation of a simple PCR analysis of the CYP2D6(A) and CYP2D6(B) null alleles associated with the debrisoquine/sparteine genetic polymorphism. Pharmacogenetics 1994;4:154-158. [ISI][Medline] [Order article via Infotrieve]
  51. Saxena R, Shall GL, Relling MV, Frame JN, Moir D, Evans WE, et al. Identification of a new variant CYP2D6 allele with a single base deletion in exon 3 and its association with the poor metabolizer phenotype. Hum Mol Genet 1994;3:923-926. [Abstract/Free Full Text]
  52. Musa MN, Miescke KJ. Pharmacogenetics of desipramine metabolism. Int J Clin Pharmacol Ther 1994;32:126-130. [ISI][Medline] [Order article via Infotrieve]
  53. Arthur H, Dahl ML, Siwers B, Sjoqvist F. Polymorphic drug metabolism in schizophrenic patients with tardive dyskinesia. J Clin Psychopharmacol 1995;15:211-216. [ISI][Medline] [Order article via Infotrieve]
  54. Goodnick PJ. Influence of fluoxetine on plasma levels of desipramine [Letter]. Am J Psychiatry 1989;146:152.
  55. Riesenman C. Antidepressant drug interactions and the cytochrome P450 system: a critical appraisal. Pharmacotherapy 1995;15:84S-99S. [Medline] [Order article via Infotrieve]
  56. Kupfer A, Presig R. Pharmacogenetics of mephenytoin: a new drug hydroxylation polymorphism in man. Eur J Clin Pharmacol 1984;26:753-759. [ISI][Medline] [Order article via Infotrieve]
  57. Ward SA, Goto F, Nakamura K, Jacqz E, Wilkinson GR, Branch RA. S-mephenytoin 4-hydroxylase is inherited as an autosomal recessive trait in Japanese families. Clin Pharmacol Ther 1987;42:96-99. [ISI][Medline] [Order article via Infotrieve]
  58. Masimirembwa C, Bertilsson L, Johnansson I, Hasler JA, Ingelman-Sundberg M. Phenotyping and genotyping of S-mephenytoin hydroxylase (cytochrome P450 2C19) in a Shona population of Zimbabwe. Clin Pharmacol Ther 1995;57:656-661. [ISI][Medline] [Order article via Infotrieve]
  59. Nakamura K, Goto F, Ray WA, McAllister CB, Jacqz E, Wilkinson GR, et al. Interethnic differences in genetic polymorphism of debrisoquin and mephenytoin hydroxylation between Japanese and Caucasian populations. Clin Pharmacol Ther 1985;38:402-408. [ISI][Medline] [Order article via Infotrieve]
  60. Wrighton SA, Steven JC, Becker GW, VandenBranden M. Isolation and characterization of human liver cytochrome P450 2C19: correlation between 2C19 and S-mephenytoin 4'-hydroxylation. Arch Biochem Biophys 1993;306:240-245. [ISI][Medline] [Order article via Infotrieve]
  61. Goldstein JA, Faletto MB, Romes-Sparks M, Sullivan T, Kitareewan S, Raucy JL, et al. Evidence that CYP2C19 is the major (S)-mephenytoin 4'-hydroxylase in humans. Biochemistry 1994;33:1742-1753.
  62. de Morias SM, Wilkinson GR, Blaisdell J, Nakamura K, Meyer UA, Goldstein JA. The major genetic defect responsible for the polymorphism of S-mephenytoin metabolism in humans. J Biol Chem 1994;269:15149-15152.
  63. de Morais SM, Wilkinson GR, Blaisdell J, Meyer UA, Nakamura K, Goldstein JA. Identification of a new genetic defect responsible for the polymorphism of (S)-mephenytoin metabolism in Japanese. Mol Pharmacol 1994;46:594-598. [Abstract]
  64. de Morais SM, Goldstein JA, Xie H-G, Huang S-L, Lu Y-Q, Xia H, et al. Genetic analysis of the S-mephenytoin polymorphism in a Chinese population. Clin Pharmacol Ther 1995;58:404-411. [ISI]<