Pharmacogenetic comparison of CYP2D6 predictive and measured phenotypes in a South African cohort.

This corrects the article DOI: 10.1038/tpj.2015.76.

Pharmacogenetic comparison of CYP2D6 predictive and measured phenotypes in a South African cohort.

The relationship between genetic variation in CYP2D6 and variable drug response represents a potentially powerful pharmacogenetic tool. However, little is known regarding this relationship in the genetically diverse South African population. The aim was therefore to evaluate the relationship between predicted and measured CYP2D6 phenotype. An XL-PCR+Sequencing approach was used to determine CYP2D6 genotype in 100 healthy volunteers and phenotype was predicted using activity scores. With dextromethorphan as the probe drug, metabolic ratios served as a surrogate measure of in vivo CYP2D6 activity. Three-hour plasma metabolic ratios of dextrorphan/dextromethorphan were measured simultaneously using semi-automated online solid phase extraction coupled with tandem mass spectrometry. Partial adaptation of the activity score system demonstrated a strong association between genotype and phenotype, as illustrated by a kappa value of 0.792, inter-rater discrepancy of 0.051 and sensitivity of 72.7%. Predicted phenotype frequencies using the modified activity score were 1.3% for poor metabolisers (PM), 7.6% for intermediate metabolisers (IM) and 87.3% for extensive metabolisers (EM). Measured phenotype frequencies were 1.3% for PM, 13.9% for IM and 84.8% for EM. Comprehensive CYP2D6 genotyping reliably predicts CYP2D6 activity in this South African cohort and can be utilised as a valuable pharmacogenetic tool.

Lean body weight dosing avoids excessive systemic exposure to proton pump inhibitors for children with obesity.

BACKGROUND: Children with obesity are more likely to suffer gastroesophageal reflux disease, requiring acid-suppression therapy with proton pump inhibitors (PPIs) and no guidelines regarding dosing. OBJECTIVE: To prospectively evaluate lean-body-weight-based (LBW) dosing of the PPI pantoprazole for children with and without obesity. METHODS: Methods: Sixty-two children (6-17 years) received a one-time oral dose of liquid pantoprazole (1.2 mg kg-1 LBW). Plasma pantoprazole concentrations were measured at 10 time points over 8 h and pharmacokinetic (PK) profiles generated using non-compartmental techniques, in order to compare PK parameters of interest between children with and without obesity, while accounting for CYP2C19 genotype. RESULTS: Adjusted for milligram-per-kilogram total body weight (TBW) pantoprazole received, apparent drug clearance (CL/F) was reduced 50% in children with vs. without obesity (p=0.03). LBW-based dosing compensated for this reduction in CL/F (p = 0.15). CONCLUSION: To achieve comparable systemic PPI exposures for children with and without obesity, we recommend using LBW, rather than TBW-based dosing for pantoprazole.

Ontogeny of dextromethorphan O- and N-demethylation in the first year of life.

The exponential increase in the number of drugs used to treat infant and childhood illnesses necessitates an understanding of the ontogeny of drug biotransformation for the development of safe and effective therapies. Healthy infants received an oral dose (0.3 mg/kg) of dextromethorphan (DM) at 0.5, 1, 2, 4, 6, and 12 months of age. DM and its major metabolites were measured in urine. CYP2D6 genotype was determined by polymerase chain reaction-restriction fragment length polymorphism. Genotyping data indicated a strong correlation between CYP2D6 genotype and DM O-demethylation (beta=-0.638; 95% CI: -0.745, -0.532; P<0.001). CYP2D6 activity was detectable and concordant with genotype by 2 weeks of age, showed no relationship with gestational age, and did not change with post natal age up to 1 year. In contrast, DM N-demethylation developed significantly more slowly over the first year of life. Genotype and the temporal acquisition of drug biotransformation are critical determinants of a drug response in infants.

Identification and characterization of CYP2D6*56B, an allele associated with the poor metabolizer phenotype.

A 5-year-old African-American girl presented with a CYP2D6*4xN/*10 genotype that was discordant with her poor metabolizer phenotype determined with the probe drug dextromethorphan. Both phenotype and genotype were confirmed in repeat assessments, suggesting that the CYP2D6*10 allele carried a novel debilitating sequence variation(s). The rationale for this study was to resolve the discordance and to describe the novel non-functional allelic variant of CYP2D6 and its frequency in populations of different ethnic backgrounds.

Cytochrome P4502D6 (CYP2D6) gene locus heterogeneity: characterization of gene duplication events.

Duplications and multiplications of active CYP2D6 genes can cause ultrarapid drug metabolism and lead to therapeutic failure. Multiple functional and non-functional duplication alleles have been further characterized. Duplications were detected by long-range polymerase chain reaction (PCR), PCR-restriction fragment length polymorphism, and sequence analysis. A PCR fragment encompassing the entire duplicated gene was utilized for detailed characterization. Duplications occurred at 1.3, 5.75, and 2.0% in Caucasian, African American, and racially mixed populations, respectively (n=887 total). Of those 28, 47, and 17% were non-functional CYP2D6*4 x N. Twelve unique duplication alleles were detected: *1 x N, *2 x N, *4 x N, *6 x N, *10 x N, *17 x N, *17 x N[spacer], *29 x N, *35 x N, *43 x N, *45 x N, and a novel non-functional tandem arrangement of a chimeric 2D7/2D6 and *1 gene. All novel duplications except *35 x N were found in African Americans. Accurate identification of gene duplication events is essential to avoid false-positive ultrarapid metabolism assignments and thus, overestimation of predicted activity and increased risk for unwanted adverse events.

Ontogeny of Hepatic Drug Transporters as Quantified by LC-MS/MS Proteomics.

Protein expression of major hepatic uptake and efflux drug transporters in human pediatric (n = 69) and adult (n = 41) livers was quantified by liquid chromatography / tandem mass spectroscopy (LC-MS/MS). Transporter protein expression of OCT1, OATP1B3, P-gp, and MRP3 was age-dependent. Particularly, significant differences were observed in transporter expression (P < 0.05) between the following age groups: neonates vs. adults (OCT1, OATP1B3, P-gp), neonates or infants vs. adolescents and/or adults (OCT1, OATP1B3, and P-gp), infants vs. children (OATP1B3 and P-gp), and adolescents vs. adults (MRP3). OCT1 showed the largest increase, of almost 5-fold, in protein expression with age. Ontogenic expression of OATP1B1 was confounded by genotype and was revealed only in livers harboring SLCO1B1*1A/*1A. In livers >1 year, tissues harboring SLCO1B1*14/*1A showed 2.5-fold higher (P < 0.05) protein expression than SLCO1B1*15/*1A. Integration of these ontogeny data in physiologically based pharmacokinetic (PBPK) models will be a crucial step in predicting hepatic drug disposition in children.

Single dose, CYP2D6 genotype-stratified pharmacokinetic study of atomoxetine in children with ADHD.

The effect of CYP2D6 genotype on the dose-exposure relationship for atomoxetine has not been well characterized in children. Children 6-17 years of age diagnosed with attention-deficit hyperactivity disorder (ADHD) were stratified by CYP2D6 genotype into groups with 0 (poor metabolizers [PMs], n = 4), 0.5 (intermediate metabolizers [IMs], n = 3), one (extensive metabolizer [EM]1, n = 8) or two (EM2, n = 8) functional alleles and administered a single 0.5 mg/kg oral dose of atomoxetine (ATX). Plasma and urine samples were collected for 24 (IM, EM1, and EM2) or 72 hours (PMs). Dose-corrected ATX systemic exposure (area under the curve [AUC]0-∞ ) varied 29.6-fold across the study cohort, ranging from 4.4 ± 2.7 µM*h in EM2s to 5.8 ± 1.7 µM*h, 16.3 ± 2.9 µM*h, and 50.2 ± 7.3 µM*h in EM1s, IMs, and PMs, respectively (P < 0.0001). Simulated steady state profiles at the maximum US Food and Drug Administration (FDA)-recommended dose suggest that most patients are unlikely to attain adequate ATX exposures. These data support the need for individualized dosing strategies for more effective use of the medication.

Pharmacogenetic allele nomenclature: International workgroup recommendations for test result reporting.

This article provides nomenclature recommendations developed by an international workgroup to increase transparency and standardization of pharmacogenetic (PGx) result reporting. Presently, sequence variants identified by PGx tests are described using different nomenclature systems. In addition, PGx analysis may detect different sets of variants for each gene, which can affect interpretation of results. This practice has caused confusion and may thereby impede the adoption of clinical PGx testing. Standardization is critical to move PGx forward.

Crystal structure of human catecholamine sulfotransferase.

Sulfonation, like phosphorylation, can modify the activity of a variety of biological molecules. The sulfotransferase enzymes sulfonate neurotransmitters, drugs, steroid hormones, dietary carcinogens and proteins. SULT1A3 specifically sulfonates catecholamines such as dopamine, adrenaline and noradrenaline. The crystal structure of SULT1A3 with a sulfate bound at the active site, has been determined at 2.4 A resolution. Although the core alpha/beta fold is like that of estrogen and heparan sulfotransferases, major differences occur in and around the active site. Most notably, several regions surrounding the active site, including a section of 40 residues, are disordered in SULT1A3. Regions that are topologically equivalent to the disordered parts of SULT1A3 are involved in substrate and cofactor binding in estrogen and heparan sulfotransferase. Flexibility in these regions suggests that ligand binding elicits a disorder-order transition in and around the active site of sulfotransferases and might contribute to the broad substrate specificity of these enzymes.

Age-Related Changes in MicroRNA Expression and Pharmacogenes in Human Liver.

Developmental changes in the liver can significantly impact drug disposition. Due to the emergence of microRNAs (miRNAs) as important regulators of drug disposition gene expression, we studied age-dependent changes in miRNA expression. Expression of 533 miRNAs was measured in 90 human liver tissues (fetal, pediatric [1-17 years], and adult [28-80 years]; n = 30 each). In all, 114 miRNAs were upregulated and 72 were downregulated from fetal to pediatric, and 2 and 3, respectively, from pediatric to adult. Among the developmentally changing miRNAs, 99 miRNA-mRNA interactions were predicted or experimentally validated (e.g., hsa-miR-125b-5p-CYP1A1; hsa-miR-34a-5p-HNF4A). In human liver samples (n = 10 each), analyzed by RNA-sequencing, significant negative correlations were observed between the expression of >1,000 miRNAs and mRNAs of drug disposition and regulatory genes. Our data suggest a mechanism for the marked changes in hepatic gene expression between the fetal and pediatric developmental periods, and support a role for these age-dependent miRNAs in regulating drug disposition.

Clinical Pharmacogenetics Implementation Consortium (CPIC) Guideline for CYP2D6 and CYP2C19 Genotypes and Dosing of Selective Serotonin Reuptake Inhibitors.

Selective serotonin reuptake inhibitors (SSRIs) are primary treatment options for major depressive and anxiety disorders. CYP2D6 and CYP2C19 polymorphisms can influence the metabolism of SSRIs, thereby affecting drug efficacy and safety. We summarize evidence from the published literature supporting these associations and provide dosing recommendations for fluvoxamine, paroxetine, citalopram, escitalopram, and sertraline based on CYP2D6 and/or CYP2C19 genotype (updates at www.pharmgkb.org).

Characterization of the microsomal epoxide hydrolase gene in patients with anticonvulsant adverse drug reactions.

Therapy with the aromatic anticonvulsants phenytoin, phenobarbital and carbamazepine has been associated with the occurrence of rare idiosyncratic hypersensitivity reactions. These drugs are thought to be activated to potentially reactive arene oxide (epoxide) metabolites by cytochrome P450-dependent monooxygenation, while liver microsomal epoxide hydrolase (mEH) plays a detoxifying role by converting such reactive intermediates to non-toxic dihydrodiols. Evidence from in vitro lymphocyte toxicity tests and enzyme inhibitor studies has suggested that an inherited defect in mEH function may be responsible for the enhanced drug toxicity observed in affected individuals. To test this hypothesis we designed methods to directly compare mEH gene structure in patients presenting with anticonvulsant adverse reactions and in control subjects in which no in vivo or in vitro toxicity to anticonvulsants could be demonstrated. Southern analysis of peripheral lymphocyte DNA using a full-length mEH cDNA as hybridization probe revealed no gross differences in mEH gene structure in selected patients when compared with DNA samples from unaffected control subjects. The human mEH gene was then cloned and characterized from a control individual. Nine exons were identified within a 22 kb region and sequences of selected regions, including all exons, were determined. Single strand conformation polymorphism (SSCP) analysis was performed on all exonic regions in genomic DNA from each of 26 subjects, including six unrelated patients with previous toxicity to anticonvulsant therapy and seven siblings (three of whom had displayed toxicity). Several distinct SSCP patterns could be observed among the subjects tested, each corresponding to a specific point mutation within one of the amplified fragments of the mEH gene. However, none of the SSCP patterns reflecting point mutations was correlated with the occurrence of anticonvulsant toxicity. From these observations we conclude that a genetic defect altering the structure and function of the mEH protein is unlikely to be responsible for predisposing patients to anticonvulsant adverse reactions.

Non-monooxygenase cytochromes P450 as potential human autoantigens in anticonvulsant hypersensitivity reactions.

Antibodies recognizing rat cytochrome P450 (CYP) 3A1 but not the closely related human CYPs 3A4/5 have been identified in the sera of patients with hypersensitivity reactions to phenytoin and carbamazepine. Comparison of the mapped epitope to the comparable region in CYP3A4 revealed that Leu361 was essential for antibody recognition because of L361V mutation (mimicking human EYLDMVVNETLRL) abolished immunoreactivity. To identify alternative human autoantigens, a site-directed mutagenesis strategy was employed to identify amino acids critical for antibody recognition. A protein database search with the consensus sequence, DxVLxETLxx, from immunoblot analysis produced CYP8 (prostacyclin synthase), CYP5A1 (thromboxane synthase), CYP27 and CYP7A1 (cholesterol 7 alpha-hydroxylase) as possible candidates; considerable homology was also observed with the fungal CYP52A subfamily. Immunoblotting with patient sera and fragments of each candidate autoantigen expressed as Escherichia coli gene 10 fusion proteins confirmed CYP8 and CYP5A1 as possible antigens, and revealed the presence of IgG1 and IgG3 antibodies against a construct mimicking fungal CYP52A10. All patient sera contained IgG4 antibodies against CYP8, CYP5A1 and the fungal mimic suggestive of continual antigenic challenge. In genetically susceptible individuals, prior infectious challenge may be a determinant of risk for the development of anticonvulsant hypersensitivity reactions and has been incorporated into a model investigating the pathogenesis of these events.

Comparison of three CYP2D6 probe substrates and genotype in Ghanaians, Chinese and Caucasians.

The ability to metabolize CYP2D6 substrates sparteine, debrisoquine, and dextromethorphan was studied in healthy Caucasian (n = 20), Ghanaian (n = 21), and Chinese (n = 22) CYP2D6 extensive metabolizers. Genotype analysis for the CYP2D6*1, *3, *4, *5, *9, *10, and *17 alleles was performed. Interethnic differences in the disposition of the probe drugs were found among the extensive metabolizers; extensive metabolizer status was confirmed by phenotype and genotype analysis. The mean metabolic rate was lower for Caucasians than for Ghanaians for sparteine (P < 0.02) and for both Ghanaians and Chinese for debrisoquine (P < 0.02). Correlation comparisons resulted in lower pairwise correlation coefficients in Ghanaians compared with Chinese and Caucasians for every combination of probe substrates. In addition, in Chinese and Caucasians, metabolic rates for each pair of probe drugs were significantly correlated (P < 0.002), but in Ghanaians the dextromethorphan metabolic rates were not correlated to either sparteine or debrisoquine (P < 0.05). Even when only those with a CYP2D6*1/*1 genotype were included in the correlation calculations, the Ghanaians had very low correlation coefficients (r(s) - 0.02-0.2, n = 9); much lower than those found in Caucasian (r(s) 0.78-0.92, n = 14) or Chinese (r(s) 0.54-0.96, n = 7) individuals. Quinidine had significantly less affect on sparteine metabolic rates in Ghanaians than both Caucasians and Chinese (P < 0.02). In addition, five of the 21 Ghanaian individuals had dextromethorphan metabolic ratios which were unaffected by quinidine. These individuals also had differences in urinary recovery of dextromethorphan and its metabolites when compared to the other Ghanaian individuals. These results confirm the large ethnic differences in probe drug metabolism and quinidine sensitivity among these ethnic groups. They also suggest that the Ghanaians have an additional unidentified allele(s) with altered substrate specificity and quinidine sensitivity which is currently genotyped as CYP2D6*1.

Limitations of dextromethorphan N-demethylation as a measure of CYP3A activity.

We evaluated the utility of the 3-methoxymorphinan/dextromethorphan (3MM/DM) urinary ratio to reflect baseline CYP3A activity, and its ability to discriminate moderate CYP3A inhibition during fluvoxamine therapy. For 4 months, oral dextromethorphan 30 mg and intravenous midazolam 0.025 mg/kg were administered to nine men every 14 days, and to 10 premenopausal women during the follicular and luteal phases of their menstrual cycles. Phenotyping during the first 3 months or cycles established baseline CYP3A activity. During the fourth month, individuals were given fluvoxamine 150 mg/day. CYP3A activity was expressed as both the urinary 3MM/DM molar ratio and midazolam plasma clearance (MDZ CL). 3MM/DM ratios were independent of dextromethorphan CYP2D6 phenotype (r = 0.13, P = 0.6). Intraindividual variability in baseline CYP3A activity (median, 25-75th percentile), as determined by coefficients of variation, was 48.3% (36.8-68.8%) for 3MM/DM and 10.3% (8.3-11.8%) for MDZ CL. No significant correlation between 3MM/DM and MDZ CL either at baseline (r = -0.22, P = 0.4) or during fluvoxamine therapy (r = -0.15, P = 0.6) was noted. With fluvoxamine 150 mg/day, median percentage change in the 3MM/DM ratios was -50.0% (-105.6-6.0%; P = 0.7), and median percentage change in MDZ CL was -33.7% (-27.0-39.3%; P < 0.0001). Only MDZ CL consistently indicated moderate inhibition of hepatic CYP3A activity. In addition, there was a lack of correlation between the magnitudes of fluvoxamine-induced change in 3MM/DM and MDZ CL (r = 0.41, P = 0.1). The large intraindividual variability of the 3MM/DM urinary ratio, as well as the inability to discriminate moderate CYP3A inhibition, makes this a suboptimal method for accurately assessing CYP3A activity.

Quantification of intraindividual variability and the influence of menstrual cycle phase on CYP2D6 activity as measured by dextromethorphan phenotyping.

Intraindividual variability and the effects of menstrual cycle phase on CYP2D6 activity were evaluated by dextromethorphan phenotyping in 20 Caucasian normal volunteers. Dextromethorphan 30 mg was administered to 10 men every 14 days for 3 months, and to 10 premenopausal women during the mid-follicular and mid-luteal phases of each menstrual cycle for three complete cycles. Urinary dextromethorphan/dextrorphan molar ratios were obtained after an overnight urine collection. Ten women and nine men were extensive metabolizer phenotypes, and one man was a poor metabolizer phenotype (confirmed by genotyping). There was no difference in dextromethorphan metabolic ratios between the mid-follicular (mean +/- SD: 0.00728+/-0.00717) and mid-luteal (0.00745+/-0.00815) phases of the menstrual cycle (P = 0.88). Also, no significant difference was found in the intraindividual variability of the metabolic ratios between the two phases (P = 0.80). No statistically significant sex difference in CYP2D6 activity was found between men (0.00537+/-0.00431) and women (0.00737+/-0.00983) extensive metabolizers (P = 0.84). For all individuals, intraindividual variability in dextromethorphan ratios ranged from 12.1-136.6% with a median of 36.7%. Because hormonal fluctuations within the mid-follicular and mid-luteal phases of the menstrual cycle do not appear to affect CYP2D6 activity, pharmacokinetic or clinical investigations of CYP2D6 substrate activity may not require menstrual cycle phase stratification. Because baseline metabolic ratios may fluctuate an average of 37%, repeat baseline and treatment phenotyping assessments should be obtained for accurate determination of a given drug's effect on CYP2D6 activity when measured by dextromethorphan.

Assessment of the predictive power of genotypes for the in-vivo catalytic function of CYP2D6 in a German population.

The polymorphic cytochrome P450 CYP2D6 catalyses the biotransformation of at least 40 drugs. The CYP2D6 genetic polymorphism is responsible for pronounced interindividual differences in plasma concentrations and, hence, in drug action and side-effects after administration of the same dose. Provided there is a close relationship between CYP2D6 genotypes and catalytic function, genotyping could be used in the clinical setting for individualization of drug dose. In the present study, we evaluated the relationship between the in-vivo enzyme activity and 35 different genotypes in order to determine whether genotyping can be used to predict a person's metabolic capacity for CYP2D6-catalysed drug oxidation using sparteine as a probe drug. One hundred and ninety-five Caucasian individuals were genotyped for seven nonfunctional (CYP2D6 x 3, x 4, x 5, x 6, x 7, x 8, x 16) and eight functional alleles (CYP2D6 x 1, x 2, x 2 x 2, x 2B, x 2B x 2, x 9, x 10, x 17). The metabolic ratio distribution for sparteine showed trimodality, with 15 poor metabolizers, 21 intermediate metabolizers, and 1.59 extensive and ultrarapid metabolizers. All poor metabolizers were unambiguously identified as carriers of two nonfunctional alleles. In contrast, the most frequent functional genotypes extensively overlapped and, with few exceptions, genotype was not a useful predictor of function. Gene dose effects among homozygotes and heterozygotes of the major functional alleles were not significant and could not explain the wide variations. Only a minor fraction of phenotypical ultrarapid metabolizers, arbitrarily defined as individuals with a metabolic ratio < 0.2, could be identified as carriers of three functional gene copies, including duplicated CYP2D6 x 2 x 2 alleles. Similarly, only a minor fraction of the intermediate metabolizers had predictive genotypes involving alleles coding for enzyme with impaired function. Thus, genotyping correctly identifies poor metabolizers, but quantitative prediction of drug metabolism capacity among extensive metabolizers is not possible.

NAD(P)H:quinone oxidoreductase: polymorphisms and allele frequencies in Caucasian, Chinese and Canadian Native Indian and Inuit populations.

NAD(P)H:quinone oxidoreductase (NQO1) catalyses the two-electron reduction of quinone compounds. NQO1 is involved in the reductive bioactivation of cytotoxic antitumour quinones such as mitomycin C, but also plays a protective role against the carcinogenicity and mutagenicity of quinones, their precursors and metabolites. Three alleles have been identified in the human population: the functional Arg139/Pro187 allele (which we have termed NQO1*1); the nonfunctional allele Arg139/Ser187 (NQO1*2) and the Trp139/Pro187 allele (NQO1*3), which is associated with a diminished activity. We applied polymerase chain reaction-based genotyping assays to characterize interethnic variability in the frequency of NQO1 alleles in Caucasian (n = 575), Canadian Native Indian (n = 110), Canadian Inuit (n = 83) and Chinese (n = 86) populations. The NQO1*2 allele was found at significantly higher frequencies in Chinese (0.49) and Native North American populations (Inuit 0.46; Canadian Native Indians 0.40) compared with Caucasians (0.16). The NQO1*3 allele was not observed in Inuit individuals, and occurred at a lower frequency than the NQO*2 allele in Caucasians (0.05), Chinese (0.04) and Canadian Native Indians (0.01). Our results predict that a greater proportion of Orientals and related ethnic groups lack, or have reduced, NQO activity relative to Caucasians. Affected individuals may not only exhibit resistance to quinone-based cancer therapy because of a decreased production of cytotoxic drug metabolites, but may also be more susceptible to toxicities associated with toxicants.

Optimization of cytochrome P4502D6 (CYP2D6) phenotype assignment using a genotyping algorithm based on allele frequency data.

Cytochrome P4502D6 (CYP2D6) is a highly polymorphic gene locus with > 50 variant alleles which lead to a wide range in enzymatic activity. So called poor metabolizers are carriers of any two non-functional alleles of the CYP2D6 gene. CYP2D6 genotyping is cumbersome and the question of how much genotyping is necessary for an accurate phenotype prediction is still debated. The goal of this study was to determine the optimum amount of genotyping required to accurately predict the phenotype at a reasonable cost in a white North American population. To address this issue, we designed a polymerase chain reaction (PCR)/restriction fragment length polymorphism-based genotyping strategy to detect 'key' mutations linked to extensive metabolizer or poor metabolizer associated alleles in combination with extra-long PCR (XL-PCR). All mutations with the exception of gene deletions and duplications are detectable by simple restriction digestion analysis and agarose gel electrophoresis. In addition, we utilized a genotyping algorithm based on our own and published allele frequency data and phenotype analysis to calculate the probability of a correct genotype (and thus, phenotype) assignment. As little as one XL-PCR reaction followed by a maximum of six reamplification reactions allows an accurate prediction of an individual's genotype to 99.15%. As few as four reamplification reactions identify 97.9% of poor metabolizer individuals. We evaluated our model in 208 white North Americans by testing for the presence of 'key' mutations linked to CYP2D6*2, *3, *4, *6, *7, *8, *9, *10, *11, *12, *15, *17 and *18 alleles and the *5, *13 and *16 gene deletions. For all individuals, the correct phenotype has been predicted. Discordant phenotype assignment occurred in only two individuals which subsequently was attributed to CYP2D6 inhibition by concomitant drug therapy.