Cytochrome P450 2D6 Genotyping

Cytochrome P450 2D6 GenotypingPotential Role in Improving Treatment Outcomes inPsychiatric Disorders

Julia Kirchheiner1 and Cristina Rodriguez-Antona2

1 Institute of Pharmacology of Natural Products and Clinical Pharmacology, University of Ulm,

Ulm, Germany

2 Hereditary Endocrine Cancer Group, Human Cancer Genetics Programme, Spanish National Cancer

Center (CNIO), Madrid, Spain

Abstract The specific reaction toward a given drug varies a lot between individualsand, for many drugs, pharmacogenetic polymorphisms are known to affectbiotransformation and clinical outcome. Estimation of the individual’s drug-metabolizing capacity can be undertaken by genotyping drug-metabolizingenzymes involved in the respective drug metabolism. Consequences that arisefrom genotyping may be the adjustment of dose according to genotype,choice of therapeutic strategy, or even choice of drug.

One of the first fields where the clinical application of pharmacogeneticsmay be used is in that of antipsychotic and antidepressant drug treatmentbecause there is a special need for individualized therapy in psychiatry. Thepharmacokinetics of many TCAs, some SSRIs and other antidepressantdrugs is significantly altered by polymorphisms; however, some controversystill exists as to whether therapeutic efficacy may be improved and/or adverseevents can be prevented by genetically driven adjustment of drug dosage.Pharmacogenetic diagnostics may be an important factor in individualizingdrug treatment according to the genetic make-up of the patients. However,routine application of pharmacogenetic dose adjustment in clinical practicerequires prospective validation.

1. The Need for Individualized Therapyin Psychiatry

The response of individual patients to thesame drug administered in the same dose variesconsiderably. Many patients will experience thedesired drug effect, while others may experienceno effects at all; however, some patients may ex-perience well known adverse drug reactions and,very rarely, a patient will die from severe adverse

effects. It is difficult for physicians to prescribethe optimal drug in the optimal dose for eachpatient, particularly in the field of psychiatricdrug treatment, since prediction of a patient’sresponse to any specific drug is rarely possible.Adverse drug effects are the fifth leading cause ofdeath in the US (after coronary heart disease,cancer, stroke and lung diseases), followed byaccidents, diabetes mellitus and pneumonia.[1]

Similar data have been published in Europe.[2,3]

LEADING ARTICLECNS Drugs 2009; 23 (3): 181-191

1172-7047/09/0003-0181/$49.95/0

ª 2009 Adis Data Information BV. All rights reserved.

To understand why individual patients res-pond differently to a drug, it is helpful to studythe progress of a particular drug from its admin-istration to its observed clinical effect. The effectof a drug will firstly depend on its systemic con-centration and its concentration at the drugtarget site. The systemic concentration of a drugdepends on several pharmacokinetic factors.It has long been known that a wide variety ofindividual patient factors, such as age, body massindex, sex, etc., may influence the pharmacoki-netics of a drug and must therefore be taken intoaccount when determining the dosage for a par-ticular patient. Hereditary variances in drug me-tabolizing enzymes and drug transporters mayalso exert considerable influence on drug con-centrations. Many enzymes involved in drugmetabolism carry genetic variants (polymor-phisms) that can decrease enzyme activity or evenlead to complete deficiency.[4] In most hetero-zygous carriers of genetic variants, the effectsare somewhere between those seen in patientswho are homozygous carriers of the wild-typeallele and those who are homozygous carriers ofthe variant allele. These variants in enzymes ofdrug metabolism can lead to differences in phar-macokinetic parameters, such as higher or lowerblood or tissue concentrations of a drug and itsmetabolites. The so-called phase I reactions inbiotransformation are small molecular modi-fications, such as oxidations and reductions,and are mostly mediated by the cytochromeP450 (CYP) enzyme family. The CYP2D6 en-zyme is involved in the metabolism of manypsychotropic drugs, such as antidepressants andantipsychotics. It is characterized by a high inter-individual variability in catalytic activity that ismainly caused by genetic polymorphisms.[5] Thephenotype determined by the CYP2D6 geno-type is predicted by the number of functionalCYP2D6 alleles, so that the presence of two, oneor no functional CYP2D6 gene copies results inrapid or extensive metabolizer (EM), intermediatemetabolizer (IM) and slow or poor metabolizer(PM) phenotypes, respectively.[6-8] Furthermore,the inheritance of three or more functional allelesby gene duplication or amplification determinesthe ultra-rapid metabolizer (UM) phenotype,

which shows higher than average enzymaticactivity.[6,7,9,10]

The clinical importance of genotyping isgreater in drug therapies in which the individualresponse is otherwise poorly predicted, as well asin drug therapies with narrow therapeutic indiceswhere severe adverse effects can occur. For ex-ample, in antidepressant drug therapy, 30% of thepatients will not respond sufficiently to drug the-rapy, and therapeutic effects are not expected tobe seen until after 2 weeks of drug intake. In thiscase, it is important to know of all the factors thathelp to predict the individual’s drug response(such as the actual metabolic activity or theprevious treatment success or pharmacogeneticfactors). Genetically caused differences in drugmetabolizing enzyme activity may be one im-portant point to consider in order to optimizetreatment.

In this context, the consequences of genotyp-ing may be the adjustment of dose according togenotype, the choice of therapeutic strategy, oreven the choice of another drug.

2. Rationale Behind the Development ofCytochrome P450 (CYP) 2D6 Testing:Genotype-Based Dose

Genetically caused variability in drug meta-bolism is reflected in differences in clearance,half-life and maximum plasma concentrations.These pharmacokinetic differences are highly re-plicable and can be considered as a basis forgenotype-based dose adjustments if a direct cor-relation with clinical outcomes is demonstrated.These dose adjustments can be calculated ac-cording to the principles of bioequivalence, underconsideration of special circumstances such aslinearity of pharmacokinetics, activity of meta-bolites and dose range of the underlying studies.Methods for extracting dose adjustments fromgenotypes have been developed and publishedelsewhere.[11-16]

Differences in pharmacokinetic parameterssuch as oral clearance could then be overcome byadjustment of the drug dose. In figure 1, the theo-retical doses for different genotypes predictingultra rapid, extensive, intermediate and poor

182 Kirchheiner & Rodriguez-Antona

ª 2009 Adis Data Information BV. All rights reserved. CNS Drugs 2009; 23 (3)

metabolic activity are depicted and derived fromthe differences in plasma concentration coursesas schematic genotype-specific doses. These theo-retical dose adjustments make sense for thosekinds of drug therapies where a similar plasmaconcentration-time course also leads to similarclinical effects (such as antibacterials and protonpump inhibitors). However, for some drugs (suchas antidepressants), plasma concentrations cor-relate poorly with clinical efficacy, and empiricdose-finding is used to obtain optimal responseand to avoid adverse drug effects.

In figure 2, the differences in mean oral clear-ance between carriers of none, one, two and threeor more active CYP2D6 genes are expressed aspercentage dose adjustments for antidepressants.In this figure, there is a huge difference betweenindividual drugs with respect to the extent towhich CYP2D6-mediated variability in clearanceis visible. Thus, for each drug, it has to be decidedwhether a genotype-based dose adjustment mayor may not be clinically important, depending onthe extent to which CYP2D6 is involved.

3. Limitations of Pharmacogenetic DoseAdjustments

When calculating quantitative dose adjust-ments from pharmacokinetic data that are de-pendent upon genotypes, several points havebeen taken into account.

First, metabolites possessing their own phar-macological activity often exist, which may causeeither therapeutic activity or adverse effects.Thus, if active metabolites exist in considerableconcentrations in plasma they should be ac-knowledged in the dose adjustments, either bysumming with the concentration of parent drugto give a concentration of the total active moietyor by deciding not to adjust the dose but insteadrecommend a change in drug choice (if metabo-lites are generated that have the potentialfor adverse drug effects). The fate of these meta-bolites generally depends on other (sometimesalso polymorphic) enzymes and therefore theindividual risk for adverse effects may varydepending on the activity of these enzymes.

The sample size and power of the existing datahave to be large enough for deriving dose adjust-ments. In many cases, information is only avail-able from small trials in healthy volunteers,involving only a few homozygous carriers of acertain variant. Furthermore, the dose range ofthe studies from which dose adjustments are de-rived should be in the clinical range. However,many studies are conducted in healthy volunteersat lower doses. In this case, dose recommenda-tions cannot be automatically extrapolated to thedose range used in patients.

Co-medication is another factor that has to betaken into account when calculating dose ad-justments. Different genotypes lead to differentconsequences from interacting substances. Forexample, a person who is genetically a PM cannothave their enzyme activity increased or decreasedby substances that are inducers or inhibitorsof that enzyme. In contrast, EMs or UMs canconvert to phenotypically PMs by strong enzymeinhibitors. Thus, genotype-based dose adjust-ments have to be weighed up against changesin phenotype that are caused, for example, byco-medication.

t

c

t

c

t

c

t

c

0

50

100

150

200

250

Dos

e (%

)

Standard doseGenotype-specific dose

PM IM EM UM

Fig. 1. Principles of calculating genotype-specific dose adjust-ments based upon differences in pharmacokinetic parameters suchas clearance and area under the plasma concentration-time curve.The theoretical doses for genetically caused subgroups of poor(PM), intermediate (IM), extensive (EM) and ultra-rapid (UM) meta-bolizers are depicted as schematic genotype-specific doses in orderto obtain equal plasma concentration-time courses (reproduced fromKirchheiner and Seeringer,[17] with permission from Elsevier. Copy-right 2007). c= concentration; t = time.

Cytochrome P450 2D6 Genotyping 183


In addition, other pharmacogenetic variantsmay influence the effects of the CYP2D6 var-iants. Drug transport[18] or genetic variability indrug targets may confound the effects caused byCYP2D6 variants and may modify dose selectionin the individual patient.

Prospective validation of genotype-based doseadjustments is necessary and several studies arenow being performed comparing therapy withpharmacogenetic diagnostics to standard therapyin a randomized controlled fashion. These pro-spective trials, as well as other future trials, haveto provide the necessary steps to demonstrate thevalidity, utility and cost-effectiveness of phar-macogenetic testing, and may provide the basisfor reimbursement programmes to be recognizedin routine clinical practice.[19]

4. Clinical Implications ofPharmacogenetic Testing in Psychiatry

Fifty-two percent of psychiatric, 49% of psycho-geriatric and 46% of geriatric patients use at least

one drug metabolized by CYP2D6, with 62%of these classified as antidepressants or anti-psychotics.[20] Thus, CYP2D6 has a specialrelevance in the treatment of depression andschizophrenia. Three different scenarios can ap-pear: (i) CYP2D6 catalyzes the inactivation of anactive drug (e.g. doxepin, trimipramine, mirta-zapine); (ii) CYP2D6 catalyzes the activation of aprodrug (e.g. codeine, tramadol); or (iii) CYP2D6converts the drug to a metabolite that has similartherapeutic activity to the parent drug and, to-gether, they are referred to as the ‘active moiety’(e.g. risperidone, venlafaxine). In the first case,UMs metabolize the active drug faster than EMsand therefore UMs may not achieve thera-peutic plasma concentrations at ordinary drugdosages, while PMs metabolize the active drugat a lower rate and consequently have a higherrisk for adverse drug reactions. In the secondcase, a lack of CYP2D6 activity (in PMs) mayresult in reduced conversion of the prodrug andlower effectiveness of drug therapy, and UMsmay have a greater risk of adverse drug reactions.

0

20

40

60

80

100

120

140

160

180

200

Imipr

amine

Doxep

in

Map

rotili

ne

Trim

ipram

ine

Desipr

amine

Nortri

ptyli

ne

Clomipr

amine

Parox

etine

Venla

faxine

Amitr

iptyli

ne

Mian

serin

Per

cent

age

of a

vera

ge d

ose

PM doseIM doseEM doseUM dose

Fig. 2. Cytochrome P450 (CYP) 2D6 genotype-dependent quantitative changes in the pharmacokinetics of antidepressant drugs expressedas percentage dose adjustments. Dose adjustments were calculated based on published pharmacokinetic data dependent on the CYP2D6genotype, as described in the review by Kirchheiner and Seering.[17] Dose adjustments are based on an average dose of 100% and are aimedat the Caucasian population. Data from studies in Asiatic, African or other populations were not incorporated as PM data are lacking (re-produced from Kirchheiner and Seering,[17] with permission from Elsevier. Copyright 2007). EM =extensive metabolizer; IM = intermediatemetabolizer; PM = poor metabolizer; UM= ultra-rapid metabolizer.



In the third scenario, the enzymes involvedin deactivation of the active metabolite (notCYP2D6) harbour the potential to affect plasmaand target site concentrations, which may in turntranslate into clinical efficacy and safety mani-festations.

5. Impact of CYP2D6 Polymorphisms onDosage of Antidepressants

Depression is the fourth leading cause of dis-ease burden worldwide and demands long-term,often life-long, drug treatment.[21]

Pharmacotherapy of depression, one of themajor psychiatric disorders, is characterized by along duration of drug therapy, relatively narrowtherapeutic index and poor predictability of in-dividual response. Approximately 30% of all pa-tients with depression do not respond sufficientlywell to the first antidepressant drug given.[22]

Failure to respond to antidepressant drug ther-apy, as well as intolerable adverse effects, notonly leads to personal suffering in both in-dividuals and their families, but also imposesconsiderable costs on society. At present, there isno reliable way to predict the individual’s re-sponse probability before initiation of a certaindrug treatment.

As shown in several studies,[23-25] differences inpharmacokinetic parameters caused by geneticpolymorphisms may impact the outcome and riskof adverse drug reactions of antidepressants.In a German study evaluating the effect ofthe CYP2D6 genotype on adverse effects andnonresponse during treatment with CYP2D6-dependent antidepressants, PMs and UMs weresignificantly over-represented compared with thecontrol population in the group of patients ex-periencing adverse drug reactions (4-fold) andnonresponders (5-fold).[23] At the same time,Grasmader et al.[24] found no influence of theCYP2D6 and CYP2C19 genotypes on anti-depressant drug response, although the incidenceof relevant adverse effects tended to be higher inPMs of CYP2D6. Furthermore, a prospective 1-year clinical study of 100 psychiatric inpatientssuggested a trend toward longer duration ofhospital stay and higher treatment costs in UMs

and PMs of CYP2D6.[25] In conclusion, whilesome reports show associations between theCYP2D6 polymorphism and adverse drug reac-tions with antidepressants, this is still con-troversial and requires further investigation inlarge studies.[26]

5.1 TCAs

The group of TCAs undergoes biotransforma-tion in the liver, withCYP2D6 catalyzing hydroxy-lation or demethylation reactions.[27]

PMs of CYP2D6 have greatly (‡50%) de-creased clearances for amitriptyline, clomipra-mine, desipramine, imipramine, nortriptyline,doxepin and trimipramine[28] (figure 2). In-dividuals who are prescribed TCAs could benefitfrom CYP2D6 genotyping if the dose is adjustedfor the group of PMs and UMs of CYP2D6.

In addition to CYP2D6, another highly poly-morphic enzyme, CYP2C19, also takes part inthe biotransformation of some TCAs such asamitriptyline, imipramine and clomipramine.[29]

Thus, genotyping for both CYP2D6 andCYP2C19may be useful for optimization of anti-depressant drug treatment with TCAs.[30]

5.2 SSRIs

While TCAs exhibit a relatively narrow thera-peutic index, they are no longer extensively pre-scribed in most countries. SSRIs are the first-linetherapy for depression and are characterized byrelatively broad therapeutic windows. SomeSSRIs, such as fluoxetine, fluvoxamine and par-oxetine, are potent inhibitors of CYP2D6 activ-ity. Therefore, after several doses of these drugs,auto-inhibition of CYP2D6 occurs and conver-sion from EM to PM phenotype and from UMto EM has been described.[31,32] In the case offluvoxamine, differences in area under the plasmaconcentration-time curve (AUC) were describedafter single doses,[33,34] whereas multiple dosesresulted in similar AUCs in PMs and EMs,indicating a strong inhibitory effect on CYP2D6in EMs.[35]

Although paroxetine is also a CYP2D6 in-hibitor, 2-fold higher AUCs of the drug were



observed in CYP2D6 PMs than in EMs aftermultiple doses.[36] In another study, one UMcarrying at least three functional CYP2D6 genecopies had undetectable drug concentrations.[32]

In contrast, no influences of CYP2D6 poly-morphisms on the pharmacokinetic parametersof sertraline and citalopram have been observed.Moclobemide, sertraline and citalopram are alsometabolized by CYP2C19, and differences inAUCs in CYP2C19 PMs were approximately2-fold (moclobemide and citalopram) or less(sertraline) compared with EMs.[37-39]

In conclusion, within the group of SSRIs,CYP inhibition poses a problem in terms ofdrug interactions, but the CYP2D6 gene poly-morphism has less effect and CYP2D6-baseddose adjustments do not seem to be useful for thisgroup of antidepressants (with the exception ofparoxetine). This conclusion is supported byrecent meta-analyses of the data on CYP2D6testing in patients being treated with SSRIs.[26,40]

5.3 Other Antidepressants

For mirtazapine, the CYP2D6 genotype hasbeen shown to have a significant influence on thevariability in plasma concentrations; however,when comparing UMs to EMs, the magnitude ofconcentration differences was only moderate.[41]

CYP2D6 is responsible for the transformationof venlafaxine to the equipotent O-desmethyl-venlafaxine.[42-45] However, a higher risk for car-diotoxic events and other adverse drug effectsmay exist in PMs[46] as cases of severe arrhythmiahave been reported in four patients treated withvenlafaxine who were all PMs of CYP2D6.[47]

The CYP2D6 polymorphism seems to haveno major influence on the metabolism of duloxe-tine, nefazodone, moclobemide, reboxetine, andtrazodone.[28]

5.4 Summary

There is considerable evidence to indicate thatCYP2D6 and, to a lesser extent, CYP2C19,polymorphisms affect the pharmacokineticsof antidepressant drugs. Insufficient data existconcerning the effects of polymorphisms ontherapeutic outcome and adverse drug effects.

The usefulness of genotyping procedures in de-pressed patients has not been confirmed inprospective clinical trials; therefore, this ap-proach is currently limited to a few hospitalsand to some patients experiencing adverse effectsor who are not achieving therapeutic effects.Nevertheless, it could be expected that withthe recent approval of the pharmacogenetictest that assesses both polymorphic genes(CYP2D6 and CYP2C19) [AmpliChip�; RocheMolecular Systems, Inc., Branchburg, NJ, USA]by the US FDA, a new era of validation stu-dies for personalized therapy of depression maybegin.

6. Impact of CYP2D6 Polymorphisms onDosage of Antipsychotics

The specific extrapyramidal adverse effects ofantipsychotic drugs show dose dependency, sug-gesting that testing for genetic polymorphismsmay be more beneficial in antipsychotic drugtreatment than for antidepressant drugs. For ex-ample, 33% of patients who developed severeadverse effects during the first few days of treat-ment with the typical antipsychotics phenothia-zine or haloperidol were CYP2D6 PMs.[48]

The antipsychotic drugs can be divided intotypical or atypical drugs, depending on theiraffinity for blocking the dopamine D2 receptorsite. The atypical agents have largely replacedtheir older relatives because of their lower ratesof extrapyramidal adverse effects and tardivedyskinesia. CYP2D6 is the main metabolic path-way for many typical antipsychotics, while foratypical antipsychotics, CYP1A2 and CYP3A4also play a role, and the individual activityof these enzymes may impact on the optimaldrug dosage.[49]

6.1 Atypical Antipsychotics

Hydroxylation of risperidone at the 9-positionis the most important metabolic pathway and ismediated mainly by CYP2D6.[50-53] The total sumof plasma risperidone and 9-hydroxyrisperidoneconcentrations is usually used as a determi-nant for risperidone pharmacological activity in



therapeutic drug monitoring.[54] Since the sum ofplasma risperidone and 9-hydroxyrisperidoneconcentrations does not differ between CYP2D6PMs and EMs, the CYP2D6 polymorphism washypothesized to not be important.[54-57] How-ever, several studies have found an impact of9-hydroxyrisperidone plasma concentrations andCYP2D6 activity on the effects of risperidone.For example, in a recent report on children withpervasive developmental disorder treated withrisperidone, the serum prolactin level was posi-tively correlated with the number of functionalCYP2D6 genes and serum 9-hydroxyrisperidoneconcentrations.[58] This and other studies haveled to the suggestion that the plasma profile forCYP2D6 PMs (characterized by higher risper-idone than 9-hydroxyrisperidone concentrations)may bemore ‘toxic’ than for other phenotypes.[59]

Accordingly, in one large study on adverse drugeffects in 360 patients, PM patients had a >3-foldhigher risk (odds ratio [OR] = 3.4) of significantrisperidone adverse drug reactions, and a 6-foldgreater risk (OR = 6.0) of discontinuing risper-idone because of adverse drug reactions thanEMs.[60] Thus, in the case of risperidone, PMsmayneed to be treated with an alternative antipsy-chotic drug or with a lower dose of risperidone.

The atypical antipsychotic drug aripiprazolewas studied for polymorphic CYP2D6 metabo-lism prior to marketing and 60% higher exposureto the total active moieties (parent drug anddehydroaripiprazole) was detected in PMs com-pared with EMs.[61] Similar changes in drugconcentrations were detected for flupentixol aswell as perazine in a naturalistic study of schizo-phrenic patients receiving different doses.[62]

A single-dose study of olanzapine in healthyvolunteers showed no CYP2D6-mediated differ-ences in pharmacokinetics,[63] whereas in a studyin patients, steady-state concentrations differedaccording to the CYP2D6 genotype.[64]

However, data from the large CATIE studycould not replicate any associations betweendrug-metabolizing enzyme polymorphisms anddosage, tolerability or efficacy for the five leadingantipsychotic drugs (perphenazine, risperidone,olanzapine, quetiapine, ziprasidone) in the clin-ical trial situation.[65]

6.2 Typical Antipsychotics

The typical antipsychotic drug thioridazine issubstantially affected by the CYP2D6 poly-morphism and PMs probably only need 30% ofthe average dose,[66] which corresponds to datafrom a study in healthy volunteers administeredsingle doses.[67] In another study in patients,smaller differences were found.[68]

With haloperidol, PMs may be sufficientlytreated with 60–70% of the average dose,[69,70]

although two studies reported no significant dif-ferences between EMs and PMs.[71,72] At thesame time, a significantly higher risk of extra-pyramidal adverse effects in PMs was observed,probably due to higher concentrations of reducedhaloperidol.[70] Patients with the UM genotypehad the worst clinical outcome, as measured bythe Positive and Negative Symptoms Scale.[70]

For perphenazine, thioridazine and zuclo-penthixol, pharmacokinetic differences due tothe CYP2D6 genotype were less marked understeady-state conditions compared with single-dose studies (PM dose 30–40% according tosingle-dose studies versus 60–70% according tomultiple-dose studies).[28] The only study evalu-ating the effects of CYP2D6 genotypes on thezuclopenthixol depot formulation revealed thatCYP2D6 PMs should get only 70% of an averagedose.[73] Thus, particularly at the beginning ofantipsychotic treatment, knowledge regardingthe PM status may help to reduce adverse eventsby initiating therapy with lower doses in thissubgroup of the population.

6.3 Summary

Numerous data for antipsychotic drugs existshowing an influence of CYP2D6 genotype ondrug clearance, blood concentrations and theblood concentration to dose ratio. However,analyses under real-life conditions in the clinicalsituation were only able to show a weak associa-tion between CYP2D6 genotype and drug dosagerequirements, tolerability and efficacy measure-ments.[19,74] Therefore, large prospective studiesare warranted in order to truly evaluate the utilityof genetic testing in the real-life setting.



7. Impact of CYP2D6 Polymorphisms onAtomoxetine in Attention-DeficitHyperactivity Disorder in Children

Atomoxetine, a known CYP2D6 substrate,has been approved for the treatment of chil-dren with attention-deficit hyperactivity disorder(ADHD). This drug was developed in the phar-macogenomic era, which means that pharmaco-genomic data have been assessed during clinicaldrug development. Drug labelling informationindicates that the AUC of atomoxetine in PMs isapproximately 10-fold greater and the maximumsteady-state plasma drug concentration is ap-proximately 5-fold greater than in EMs.[75] It ismentioned that laboratory tests are available toidentify CYP2D6 PMs, but no recommendationsfor dose adjustment are given.[75] Recent analysesby the manufacturer (Eli Lilly and Company,Indianapolis, IN, USA) show that clinicians wereable to dose atomoxetine to comparable efficacyand safety levels in EMs and PMs without know-ledge of genotype metabolizer status.[76] How-ever, reports exist regarding the adverse effects ofatomoxetine in PMs and its dose-related toxi-city,[77,78] which contradict the conclusion thatCYP2D6 genotype testing is unnecessary duringroutine clinical management.

8. Conclusion

While genotyping for CYP2D6 in psychiatricpatients is currently mostly performed for re-search purposes, it may potentially be usedin clinical practice as a predictor of drug meta-bolism, which would allow for optimizing thedosage to prevent adverse drug reactions. It isnoteworthy that there are currently no reliablebiomarkers for psychiatric disorders and, thus,prescribing is often based on a ‘trial and error’approach so that the first medication prescribedis often changed for a second or even third choicefollowing the patient’s nonresponse to treatment.

It should be emphasized that before the ap-plication of CYP2D6 testing in clinical practice,there are several prerequisites in order to gainfurther evidence for the utility and consequencesof such testing. Most existing data on CYP2D6

influences in psychopharmacology are pharma-cokinetic data, often from studies in healthy vo-lunteers. Some reports on the effects of CYP2D6polymorphisms on adverse drug reactions exist,but no prospective studies have been conductedto date, and only very limited data support apositive correlation towards an influence on drugresponse. Thus, prospective, randomized studiesare required to determine whether CYP2D6genotyping prior to the start of drug treatment isbeneficial in personalizing dosage. These studiesshould include data on adverse drug effects andclinical response. However, no prospective stu-dies of this type with antidepressants or anti-psychotics have been conducted to date.

Acknowledgements

No sources of funding were used to assist in the prepara-tion of this review. The authors have no conflicts of interestthat are directly relevant to the content of this review.

References1. Lazarou J, Pomeranz BH, Corey PN. Incidence of adverse

drug reactions in hospitalized patients: a meta-analysis ofprospective studies. JAMA 1998; 279 (15): 1200-5

2. Schneeweiss S, Hasford J, Gottler M, et al. Admissionscaused by adverse drug events to internalmedicine and emer-gency departments in hospitals: a longitudinal population-based study. Eur J Clin Pharmacol 2002; 58 (4): 285-91

3. Pirmohamed M, James S, Meakin S, et al. Adverse drugreactions as cause of admission to hospital: prospectiveanalysis of 18 820 patients. BMJ 2004; 329 (7456): 15-9

4. Evans WE, Relling MV. Pharmacogenomics: translatingfunctional genomics into rational therapeutics. Science1999; 286 (5439): 487-91

5. Bertilsson L, Dahl ML, Dalen P, et al. Molecular genetics ofCYP2D6: clinical relevance with focus on psychotropicdrugs. Br J Clin Pharmacol 2002; 53 (2): 111-22

6. Sachse C, Brockmoller J, Bauer S, et al. Cytochrome P4502D6 variants in a Caucasian population: allele frequenciesand phenotypic consequences. Am J Hum Genet 1997; 60(2): 284-95

7. Zanger UM, Fischer J, Raimundo S, et al. Comprehensiveanalysis of the genetic factors determining expression andfunction of hepatic CYP2D6. Pharmacogenetics 2001; 11(7): 573-85

8. Home page of the human cytochrome P450 (CYP) allelenomenclature committee [online]. Available from URL:http://www.cypalleles.ki.se/ [Accessed 2008 Dec 2]

9. Griese EU, Zanger UM, Brudermanns U, et al. Assess-ment of the predictive power of genotypes for the in-vivocatalytic function of CYP2D6 in a German population.Pharmacogenetics 1998; 8 (1): 15-26



10. Gaedigk A, Ndjountche L, Divakaran K, et al. CytochromeP4502D6 (CYP2D6) gene locus heterogeneity: character-ization of gene duplication events. Clin Pharmacol Ther2007; 81 (2): 242-51

11. Kirchheiner J, Brockmoller J. Clinical consequences ofcytochrome P450 2C9 polymorphisms. Clin PharmacolTher 2005; 77 (1): 1-16

12. Kirchheiner J, Brøsen K, Dahl ML, et al. CYP2D6 andCYP2C19 genotype-based dose recommendations forantidepressants: a first step towards subpopulation-specificdosages. Acta Psychiatr Scand 2001; 104 (3): 173-92

13. McLeod HL, Siva C. The thiopurine S-methyltransferasegene locus: implications for clinical pharmacogenomics.Pharmacogenomics 2002; 3 (1): 89-98

14. Wu AH, Wang P, Smith A, et al. Dosing algorithm forwarfarin using CYP2C9 and VKORC1 genotyping from amulti-ethnic population: comparison with other equations.Pharmacogenomics 2008; 9 (2): 169-78

15. Wen MS, Lee M, Chen JJ, et al. Prospective study of war-farin dosage requirements based on CYP2C9 andVKORC1 genotypes. Clin Pharmacol Ther 2008 Jul;84 (1): 83-9

16. Sconce EA, Khan TI, Wynne HA, et al. The impact ofCYP2C9 and VKORC1 genetic polymorphism andpatient characteristics upon warfarin dose requirements:proposal for a new dosing regimen. Blood 2005; 106 (7):2329-33

17. Kirchheiner J, Seeringer A. Clinical implications of phar-macogenetics of cytochrome P450 drug metabolizingenzymes. Biochim Biophys Acta 2007; 1770 (3): 489-94

18. Uhr M, Tontsch A, Namendorf C, et al. Polymorphisms inthe drug transporter gene ABCB1 predict antidepressanttreatment response in depression. Neuron 2008; 57 (2): 203-9

19. Grossman I. Routine pharmacogenetic testing in clinicalpractice: dream or reality? Pharmacogenomics 2007; 8 (10):1449-59

20. Mulder H, Heerdink ER, van Iersel EE, et al. Prevalence ofpatients using drugs metabolized by cytochrome P450 2D6in different populations: a cross-sectional study. AnnPharmacother 2007; 41 (3): 408-13

21. Ustun TB, Ayuso-Mateos JL, Chatterji S, et al. Globalburden of depressive disorders in the year 2000. Br J Psy-chiatry 2004; 184: 386-92

22. Bauer M, Whybrow PC, Angst J, et al. World Federation ofSocieties of Biological Psychiatry (WFSBP) guidelines forbiological treatment of unipolar depressive disorders: part1. Acute and continuation treatment of major depressivedisorder. World J Biol Psychiatry 2002; 3 (1): 5-43

23. Rau T, Wohlleben G, Wuttke H, et al. CYP2D6 genotype:impact on adverse effects and nonresponse during treat-ment with antidepressants-a pilot study. Clin PharmacolTher 2004; 75 (5): 386-93

24. Grasmader K, Verwohlt PL, Rietschel M, et al. Impact ofpolymorphisms of cytochrome-P450 isoenzymes 2C9,2C19 and 2D6 on plasma concentrations and clinicaleffects of antidepressants in a naturalistic clinical setting.Eur J Clin Pharmacol 2004; 60 (5): 329-36

25. Kawanishi C, Lundgren S, Agren H, et al. Increasedincidence of CYP2D6 gene duplication in patients withpersistent mood disorders: ultrarapid metabolism of anti-

depressants as a cause of nonresponse. A pilot study. Eur JClin Pharmacol 2004; 59 (11): 803-7

26. Matchar DB, Thakur ME, Grossman I, et al. Testingfor cytochrome P450 polymorphisms in adults with non-psychotic depression treated with selective serotonin re-uptake inhibitors (SSRIs). Evid Rep Technol Assess (FullRep) 2007; 146: 1-77

27. Baumann P, Jonzier-Perey M, Koeb L, et al. Amitriptylinepharmacokinetics and clinical response: II. Metabolic poly-morphism assessed by hydroxylation of debrisoquine andmephenytoin. Int Clin Psychopharmacol 1986; 1 (2): 102-12

28. Kirchheiner J, Nickchen K, Bauer M, et al. Pharmaco-genetics of antidepressants and antipsychotics: the contri-bution of allelic variations to the phenotype of drugresponse. Mol Psychiatry 2004; 9 (5): 442-73

29. Brøsen K, Gram LF. Pharmacokinetic and clinicalsignificance of genetic variability in psychotropic drugmetabolism. Psychopharmacol Ser 1989; 7: 192-200

30. Steimer W, Zopf K, von Amelunxen S, et al. Amitriptylineor not, that is the question: pharmacogenetic testing ofCYP2D6 and CYP2C19 identifies patients with low or highrisk for side effects in amitriptyline therapy. Clin Chem2005; 51 (2): 376-85

31. Laine K, Tybring G, Hartter S, et al. Inhibition of cyto-chrome P4502D6 activity with paroxetine normalizes theultrarapid metabolizer phenotype as measured by nor-triptyline pharmacokinetics and the debrisoquin test. ClinPharmacol Ther 2001; 70 (4): 327-35

32. Lam YW, Gaedigk A, Ereshefsky L, et al. CYP2D6 inhibi-tion by selective serotonin reuptake inhibitors: analysis ofachievable steady-state plasma concentrations and theeffect of ultrarapid metabolism at CYP2D6. Pharmaco-therapy 2002; 22 (8): 1001-6

33. Carrillo JA, Dahl ML, Svensson JO, et al. Disposition offluvoxamine in humans is determined by the polymorphicCYP2D6 and also by the CYP1A2 activity. Clin Pharma-col Ther 1996; 60 (2): 183-90

34. Spigset O, Granberg K, Hagg S, et al. Relationship betweenfluvoxamine pharmacokinetics and CYP2D6/CYP2C19phenotype polymorphisms. Eur J Clin Pharmacol 1997; 52(2): 129-33

35. Spigset O, Granberg K, Hagg S, et al. Non-linear fluvox-amine disposition. Br J Clin Pharmacol 1998; 45 (3): 257-63

36. Sindrup SH, Brøsen K, Gram LF. Pharmacokinetics of theselective serotonin reuptake inhibitor paroxetine: non-linearity and relation to the sparteine oxidation poly-morphism. Clin Pharmacol Ther 1992; 51 (3): 288-95

37. Gram LF, Guentert TW, Grange S, et al. Moclobemide, asubstrate of CYP2C19 and an inhibitor of CYP2C19,CYP2D6, and CYP1A2: a panel study. Clin PharmacolTher 1995; 57 (6): 670-7

38. Wang JH, Liu ZQ, Wang W, et al. Pharmacokinetics ofsertraline in relation to genetic polymorphism ofCYP2C19. Clin Pharmacol Ther 2001; 70 (1): 42-7

39. Sindrup SH, Brøsen K, Hansen MG, et al. Pharmacoki-netics of citalopram in relation to the sparteine and themephenytoin oxidation polymorphisms. Ther Drug Monit1993; 15 (1): 11-7

40. Thakur M, Grossman I, McCrory DC, et al. Review of evi-dence for genetic testing for CYP450 polymorphisms in



management of patients with nonpsychotic depression withselective serotonin reuptake inhibitors. Genet Med 2007; 9(12): 826-35

41. Kirchheiner J, Henckel HB, Meineke I, et al. Impact of theCYP2D6 ultrarapid metabolizer genotype on mirtazapinepharmacokinetics and adverse events in healthy volunteers.J Clin Psychopharmacol 2004; 24 (6): 647-52

42. Fukuda T, Yamamoto I, Nishida Y, et al. Effect of theCYP2D6*10 genotype on venlafaxine pharmacokinetics inhealthy adult volunteers. Br J Clin Pharmacol 1999; 47 (4):450-3

43. Fukuda T, Nishida Y, Zhou Q, et al. The impact of theCYP2D6 and CYP2C19 genotypes on venlafaxinepharmacokinetics in a Japanese population. Eur J ClinPharmacol 2000; 56: 175-80

44. Otton SV, Ball SE, Cheung SW, et al. Venlafaxine oxidationin vitro is catalysed by CYP2D6. Br J Clin Pharmacol 1996;41 (2): 149-56

45. Veefkind AH, Haffmans PM, Hoencamp E. Venlafaxineserum levels and CYP2D6 genotype. Ther Drug Monit2000; 22 (2): 202-8

46. Shams ME, Arneth B, Hiemke C, et al. CYP2D6polymorphism and clinical effect of the antidepressantvenlafaxine. J Clin Pharm Ther 2006; 31 (5): 493-502

47. Lessard E, Yessine M, Hamelin B, et al. Influence ofCYP2D6 activity on the disposition and cardiovasculartoxicity of the antidepressant agent venlafaxine in humans.Pharmacogenetics 1999; 9: 435-43

48. Spina E, Ancione M, Di Rosa AE, et al. Polymorphic deb-risoquine oxidation and acute neuroleptic-induced adverseeffects. Eur J Clin Pharmacol 1992; 42 (3): 347-8

49. de Leon J, Armstrong SC, Cozza KL. The dosing of atypicalantipsychotics. Psychosomatics 2005; 46 (3): 262-73

50. Prior TI, Chue PS, Tibbo P, et al. Drug metabolism andatypical antipsychotics. Eur Neuropsychopharmacol 1999;9 (4): 301-9

51. Fang J, Bourin M, Baker GB. Metabolism of risperidone to9-hydroxyrisperidone by human cytochromes P450 2D6and 3A4. Naunyn Schmiedebergs Arch Pharmacol 1999;359 (2): 147-51

52. Jung SM, Kim KA, Cho HK, et al. Cytochrome P450 3Ainhibitor itraconazole affects plasma concentrations ofrisperidone and 9-hydroxyrisperidone in schizophrenicpatients. Clin Pharmacol Ther 2005; 78 (5): 520-8

53. Mahatthanatrakul W, Nontaput T, Ridtitid W, et al.Rifampin, a cytochrome P450 3A inducer, decreases plas-ma concentrations of antipsychotic risperidone in healthyvolunteers. J Clin Pharm Ther 2007; 32 (2): 161-7

54. Huang ML, Van Peer A, Woestenborghs R, et al. Pharma-cokinetics of the novel antipsychotic agent risperidone andthe prolactin response in healthy subjects. Clin PharmacolTher 1993; 54 (3): 257-68

55. Olesen OV, Licht RW, Thomsen E, et al. Serum con-centrations and side effects in psychiatric patientsduring risperidone therapy. Ther Drug Monit 1998; 20 (4):380-4

56. RohHK, Chung JY, OhDY, et al. Plasma concentrations ofhaloperidol are related to CYP2D6 genotype at low, butnot high doses of haloperidol in Korean schizophrenicpatients. Br J Clin Pharmacol 2001; 52 (3): 265-71

57. Scordo MG, Spina E, Facciola G, et al. Cytochrome P4502D6 genotype and steady state plasma levels of rispe-ridone and 9-hydroxyrisperidone. PsychopharmacologyBerl 1999; 147 (3): 300-5

58. Troost PW, Lahuis BE, Hermans MH, et al. Prolactin re-lease in children treated with risperidone: impact and roleof CYP2D6 metabolism. J Clin Psychopharmacol 2007; 27(1): 52-7

59. Bork JA, Rogers T, Wedlund PJ, et al. A pilot study onrisperidone metabolism: the role of cytochromes P450 2D6and 3A. J Clin Psychiatry 1999; 60 (7): 469-76

60. de Leon J, Susce MT, Pan RM, et al. The CYP2D6 poormetabolizer phenotype may be associated with risperidoneadverse drug reactions and discontinuation. J Clin Psy-chiatry 2005; 66 (1): 15-27

61. Abilify� prescribing information. Available from URL:http://packageinserts.bms.com/pi/pi_abilify.pdf [Accessed2008 Dec 12]

62. Walter S. Bedeutung der erblichen Polymorphismen vonCytochrom-P450-2D6 fur den Metabolismus und diePharmakokinetik von Antipsychotika. Berlin: HumboldtUniversitat zu Berlin, 2000

63. Hagg S, Spigset O, Lakso HA, et al. Olanzapine dispositionin humans is unrelated to CYP1A2 and CYP2D6 pheno-types. Eur J Clin Pharmacol 2001 Sep; 57 (6-7):493-7

64. Carrillo JA, Herraiz AG, Ramos SI, et al. Role of thesmoking-induced cytochrome P450 (CYP) 1A2 and poly-morphic CYP2D6 in steady-state concentration of olan-zapine. J Clin Psychopharmacol 2003; 23 (2): 119-27

65. Grossmann I, Liu Y,Walley N. Pharmacogenetic analysis ofantipsychotics comprehensive analysis of pharmacokineticvariants. 56th Annual Meeting of the American Society ofHuman Genetics; 2006 Oct 10-13; New Orleans (LA)

66. Berecz R, de la Rubia A, Dorado P, et al. Thioridazinesteady-state plasma concentrations are influenced by to-bacco smoking and CYP2D6, but not by the CYP2C9genotype. Eur J Clin Pharmacol 2003; 59 (1): 45-50

67. von Bahr C, Movin G, Nordin C, et al. Plasma levels ofthioridazine and metabolites are influenced by the debri-soquin hydroxylation phenotype. Clin Pharmacol Ther1991; 49 (3): 234-40

68. Eap CB, Guentert TW, Schaublin Loidl M, et al. Plasmalevels of the enantiomers of thioridazine, thioridazine2-sulfoxide, thioridazine 2-sulfone, and thioridazine5-sulfoxide in poor and extensive metabolizers of dex-tromethorphan and mephenytoin. Clin Pharmacol Ther1996; 59 (3): 322-31

69. Llerena A, Dahl ML, Ekqvist B, et al. Haloperidol disposi-tion is dependent on the debrisoquine hydroxylationphenotype: increased plasma levels of the reduced meta-bolite in poor metabolizers. Ther Drug Monit 1992; 14 (3):261-4

70. Brockmoller J, Kirchheiner J, Schmider J, et al. The impactof the CYP2D6 polymorphism on haloperidol pharmaco-kinetics and outcome. Clin Pharmacol Ther 2002; 72:438-52

71. Gram LF, Debruyne D, Caillard V, et al. Substantial rise insparteine metabolic ratio during haloperidol treatment.Br J Clin Pharmacol 1989; 27 (2): 272-5



72. Young D, Midha KK, Fossler MJ, et al. Effect of quinidineon the interconversion kinetics between haloperidol andreduced haloperidol in humans: implications for the in-volvement of cytochrome P450IID6. Eur J Clin Pharmacol1993; 44 (5): 433-8

73. Jaanson P, Marandi T, Kiivet RA, et al. Maintenance ther-apy with zuclopenthixol decanoate: associations betweenplasma concentrations, neurological side effects andCYP2D6genotype. Psychopharmacology (Berl) 2002; 162 (1): 67-73

74. Brockmoller J, Kirchheiner J, Schmider J, et al. The impactof the CYP2D6 polymorphism on haloperidol pharmaco-kinetics and on the outcome of haloperidol treatment. ClinPharmacol Ther 2002; 72 (4): 438-52

75. Strattera� prescribing information. Available from URL:http://pi.lilly.com/us/strattera-pi.pdf [Accessed 2008Dec 12]

76. Trzepacz PT, Williams DW, Feldman PD, et al. CYP2D6metabolizer status and atomoxetine dosing in children and

adolescents with ADHD. Eur Neuropsychopharmacol2008; 18 (2): 79-86

77. Lim JR, Faught PR, Chalasani NP, et al. Severe liver injuryafter initiating therapy with atomoxetine in two children.J Pediatr 2006; 148 (6): 831-4

78. Reimers A, Langsetmo HK. Combined overdose of ato-moxetine and methylphenidate in a cytochrome P450 2D6poor metabolizer. J Clin Psychopharmacol 2007; 27 (1):110-1

Correspondence: Prof. Dr Julia Kirchheiner, Institute ofPharmacology of Natural Products and Clinical Pharma-cology, University Ulm, Helmholtzstr. 20, 89081 Ulm,Germany.E-mail: [email protected]



Documents

Cytochrome P450 2D6 Genotyping