| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Drug Monitoring and Toxicology |
1 Institut für Klinische Chemie und Pathobiochemie, and 2
Psychiatrische Klinik und Poliklinik, Klinikum rechts der Isar, Technische Universität München, Munich, Germany.
3 Bezirkskrankenhaus Haar, Haar, Germany.
aAddress correspondence to this author at: Institut für Klinische Chemie und Pathobiochemie, Klinikum rechts der Isar, Technische Universität München, Ismaningerstrasse 22, D-81675 Munich, Germany. Fax 49-89-4140-4875; e-mail Steimer{at}KlinChem.med.TU-Muenchen.de.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: We conducted a prospective, blinded two-center study seeking correlations between CYP2C19 (*2,*3, and *4; conventional PCR) and CYP2D6 genotypes (*1 to *10, *35, and *41; real-time and multiplex PCR) and drug concentrations (Emit® and HPLC) in 50 Caucasians receiving amitriptyline (AT; 75 mg twice a day).
Results: Eighteen CYP2C19 heterozygotes (*1/*2) had higher AT (P = 0.033) and lower nortriptyline (NT; P = 0.059) concentrations than 30 homozygotes (*1/*1). For CYP2D6, we calculated two new indices, i.e., the allele-specific change of concentration on identical background (ASCOC) and a quantitative functional gene dose. The ASCOC describes the change in NT concentration attributable to a mutant allele compared with the wild type. We found significantly higher concentrations for alleles *4 (95.6%; P <0.0001), *10 (63.3%; P <0.001), and *41 (39.8%; P <0.0001) but not for *2 and *35. Assigning of semiquantitative gene doses of 0, 0.5, or 1 to each allele instead of applying the current classification system (predicted phenotypes: 3 intermediate metabolizers, 46 extensive metabolizers, and 1 ultrarapid metabolizer) produced significant NT concentration differences: gene doses of 0.5 (n =3), 1 (n = 14), 1.5 (n = 11), 2 (n = 21) and 3 (n = 1; P <0.00001).
Conclusions: AT and NT concentrations can be predicted within the group of CYP2D6 extensive metabolizers. The ASCOC provides substantial advantages compared with current methods of analysis. CYP2D6 but not CYP2C19 correlates with the sum of both concentrations used to guide AT therapy.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
TCAs show a narrow therapeutic range and produce considerable toxicity at increased concentrations, leading to severe side effects. Concentrations below the therapeutic range may be ineffective in antidepressive treatment(2).
The major metabolic pathway for amitriptyline (AT) starts with demethylation to nortriptyline (NT), mainly by CYP2C19(3). NT is an active compound on its own, which is the reason that the sum of the concentrations of the parent substance and the primary metabolite is used to guide AT therapy. Consecutively, NT is hydroxylated by CYP2D6, yielding 10-OH-NT, a compound generally not believed to contribute significantly to the effects of AT therapy.
Alternative minor pathways described for the metabolism of AT may be primary hydroxylation by CYP2D6 to 10-OH-AT and the involvement of CYP1A2, CYP2C9, and CYP3A4 in demethylation. These pathways, in particular CYP3A4, may be more important at higher concentrations(3). CYP2C19 and CYP2D6 are highly polymorphic, leading to a wide range of enzymatic activities, from absent to normal or even increased activity. On the basis of metabolic activity in relation to enzyme-specific test drugs, individuals can be classified as poor (PM), intermediate (IM), extensive (EM) and ultrarapid metabolizers (UM; CYP2D6 only).
From genotyping results, carriers of two CYP2C19 wild-type alleles are predicted to be EMs (gEM), carriers of two dysfunctional alleles are predicted to be PMs (gPM), and heterozygous carriers are predicted to be IMs (gIM)(4).
The situation with CYP2D6 is more complicated. Genotyping allows identification of close to 100% of all PMs (6–8% in a Caucasian population)(5). However, only 20% of phenotypic UMs (5–10% in a Caucasian population) can be explained by multiplication of an active gene(6)(7)(8). Until recently, the metabolic capacity of a large subgroup of patients with impaired but residual function (IM; 10–15%) and enzyme function within the largest group (EM) could not be predicted with sufficient precision by genotyping, and a clear-cut gene-dose effect could not be established(7)(8)(9)(10).
Two alleles have been reported from in vitro studies as possibly playing a key role in the IM and ultraextensive metabolizer phenotype (CYP2D6*41 and *35)(9)(11)(12)(13). Both alleles have been studied in only a few clinical populations being treated with CYP2D6 substrates(14)(15)(16). The cost/benefit ratio of pretherapeutic CYP2D6 genotyping suffers from the fact that, at present, only 
8–10% of patients genotyped (gPMs and gUMs) would potentially benefit from the procedure(17)(18). In particular, the ability to identify IMs and predict metabolic function within the EM group could change cost/benefit estimations of pretherapeutic genotyping and have substantial clinical impact(10)(17)(18). Few data are available from single-dose in vivo studies with test drugs, and even fewer data are available for clinically treated patients(14)(15). We therefore determined the most important CYP2C19 and CYP2D6 polymorphisms, including *41 and *35, and plasma concentrations of AT and NT in a Caucasian population of 50 depressed inpatients treated for 3 weeks with a fixed dose of AT.
The study sought to answer the following questions:
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Neither genotype nor serum concentrations were known to the treating physicians or the patients during the first 3 weeks. Serum concentrations were available, however, to a study supervisor to ensure that no potentially toxic concentrations under the fixed-dose regimen would go unrecognized.
The clinical data and genotype outcome correlations are described elsewhere (Steimer et al., submitted for publication).
dosing
The AT dose was increased over the first 2 days and was then given at a fixed dose of 150 mg/day (75 mg twice a day; 12-h dosing interval) for the first 3 weeks of treatment. The patients took their medication under supervision to reduce noncompliance. Accompanying medication was allowed, although substances interfering with CYP2D6 or CYP2C19 metabolism were avoided.
blood sampling
Blood samples for measurement of serum AT and NT were taken from a peripheral vein (Serum Monovette®; Sarstedt) immediately before the morning dose at 0830 on days 0, 7, 14, 18, and 21 to assure that steady state had been reached. Measurements were performed within a maximum of 72 h after sampling on centrifuged samples stored at 4 °C. After 72 h, all samples were frozen at –20 °C.
measurement of serum concentrations
Serum concentrations were measured by either the Emit® immunoassay specific for AT and NT (Syva; center 1) or a commercial HPLC assay (Bio-Rad Diagnostics Group; center 2 and center 1 for confirmatory measurements). Accuracy was ensured for both centers by regular and successful contributions to an international proficiency testing scheme (Heathcontrol), where both methods performed equally [mean differences between center 1 and center 2 were 1.1% (AT) and 5.5% (NT) vs 1.3% (AT) and 5.4% (NT)]. The limit of quantification used routinely for both assays (AT and NT) was 20 µg/L. All results measured were >20 µg/L. The calibration range of the Emit assay was 0–250 µg/L (AT and NT), and for the HPLC it was a one-point calibration (258 and 262 µg/L for AT and NT, respectively). The between-day CVs for single routine measurements were as follows:
To compensate for the differences in precision, all measurements in patient samples by Emit were performed in duplicate.
genotyping
Each patient gave 2.7 mL of EDTA blood for genotype analysis. Genotyping of the dysfunctional cytochrome P450 2C19 alleles *2, *3, and *4 was performed according to published methods(21)(22)(23). CYP2D6 genotype was assessed as described elsewhere(24)(25) by real-time PCR. The most important alleles in a Caucasian population were assessed: the fully functional alleles [*1 and *2 (formerly *2G)]; the completely dysfunctional alleles (*3, *4, *5, *6, *7, and *8); the alleles with reduced function [*9, *10, and *41 (formerly *2C)]; and duplicated alleles and *35 with potentially enhanced function.
statistics
Statistical analyses were performed with SPSS 11.5. We analyzed data with the Kruskal–Wallis H or the Mann–Whitney U nonparametric ANOVA to establish differences between groups. The reported P values are always two-tailed.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
serum concentrations
On day 21, the mean (SD) concentration per dose unit of AT + NT was 77.3 (31.4) µg/L · kg/mg (range, 30.6–190.8 µg/L · kg/mg). The corresponding concentrations for NT and AT were as follows: NT, 34.5 (20.0) µg/L · kg/mg (range, 10.1–100.2 µg/L · kg/mg); AT, 42.9 (19.4) µg/L · kg/mg (range, 13.2–94.9 µg/L · kg/mg). We observed no significant differences between days 7 and 21. The mean CV of the serum concentrations (AT + NT) for all 50 patients between days 7 and 21 was 14%. The mean concentrations on days 7 and 21 were as follows: for AT + NT, 158 µg/L on day 7 and 166 µg/L on day 21; for AT, 85 and 91 µg/L on days 7 and 21, respectively; for NT, 72 and 75 µg/L on days 7 and 21, respectively. These values indicate that steady state had been achieved.
CYP2C19
The DNA of one patient was not suitable for PCR analysis of CYP2C19 after several assays and thus could not be included. Of the remaining 49 patients, 1 patient (2%) was identified as a carrier of two null alleles (*2/*2; predicted phenotype, PM), 18 patients (36%) were identified as carriers of one null allele (*1/*2; predicted phenotype, IM), and 30 patients (60%) were identified as carriers of two wild-type alleles (*1/*1; predicted phenotype, EM). The allele frequency of the dysfunctional *2 allele was 20.4%. No dysfunctional alleles other than *2 were detected, which is in agreement with previous publications for Caucasian populations(26)(27). Patients with reduced CYP2C19 function (gPMs and gIMs) had higher concentrations of AT but lower concentrations of NT and lower NT/AT ratios than gEMs (see the "Total" rows in Table 4
). This is plausible because CYP2C19 is believed to be the major enzyme responsible for demethylation of AT to NT(3). The observed differences between gEMs and gIMs were statistically significant regarding the serum concentrations of AT (P = 0.033) and the NT/AT ratio (P = 0.001). Differences in NT just missed statistical significance (P = 0.059). There was no significant trend toward higher total concentrations of the sum of both active compounds (AT + NT) in patients with dysfunctional alleles, neither in the gIM population (P = 0.650) nor in the one gPM patient observed. This can be expected because the effect of CYP2C19 points in opposite directions for AT and NT. CYP2C19 thus influences the ratio but not the summed concentration of both compounds.
|
CYP2D6
The following CYP2D6 allele frequencies were found: *1 (40%), *2 (28%), *4 (16%), *5 (1%), *1XN (1%), *41 (11%), and *10 (3%). The *35 G
A polymorphism was found in five patients. Table 1
shows the frequencies of the 11 genotypes (without *35) and the mean concentrations per dose unit per weight and corrected for CYP2C19 genotypes measured between days 7 and 21. These were used for all further analyses.
|
CYP2D6 allele function
The high number of functionally relevant polymorphisms in CYP2D6 provide a multitude of allele combinations with few individuals for each allele combination. This hampers the statistical analysis of relationships between genotype and drug concentration data.
To overcome this problem, we assessed CYP2D6 function on the allele level rather than the genotype level. Groups of patients identical for one allele but different for the second allele were compared and analyzed to assess whether the difference in the second allele caused a difference in steady-state drug concentrations. For analysis of the effects of dysfunctional alleles (*4, *10, and *41), the two alleles with normal function (*1 or *2) were used as reference.
For example, patients with *1/*41 (test group) were analyzed vs patients with *1/*1 (reference group)(11) (see Table 2
). The individual concentrations of all patients in the reference group (n = 8) and test group (n = 4) were expressed as the percentage deviation from the mean concentration of the reference group (24.7 µg/L·kg/mg; see Table 1
). These percentage values were termed allele-specific change of concentration on identical background (ASCOC). In the reference group, the ASCOC values scatter around 0% because their own mean value was used as the reference concentration. In the test group, the ASCOC values represent the relative (percentage) change of concentration attributable to the test allele (e.g., *41; see Fig. 1
). On the basis of the observed genotypes, the effects of *41 were studied on three backgrounds (*1, *2, and *4) and vs the two functionally equivalent alleles, *1 and *2 (six comparisons in total; see Table 2
). In contrast to absolute concentrations, ASCOC values allow the combined statistical analysis of all observations from these six comparisons because they are standardized and represent the relative (percentage) change of concentration.
|
|
This procedure was repeated for the other dysfunctional alleles. This analysis indicated that alleles *41, *10, and *4 have a significant impact on NT concentrations (Fig. 1
).
The *35 allele.
We could not identify a uniform trend when we compared dose-corrected NT serum concentrations of heterozygous carriers of the *35 allele with patients not carrying the allele. The mean change of concentration (ASCOC) attributable to the *35 A allele when present in heterozygous individuals was 1.7% (P = 0.851, not significant). The *35 allele was disregarded in the following analysis.
The *2 allele.
We compared all groups identical in one allele and displaying either *1 or *2 as the second allele (Table 2
) as well as *1/*1 with *2/*2. The mean change of concentration (ASCOC) attributable to the *2 allele compared with the *1 wild-type allele was 2.4% (P = 0.957). Overall, we could detect no significant difference in the activity of alleles *1 and *2 (Fig. 1
). Therefore, *1 and *2 were considered equal and were both used as references in the further analysis.
The *41 allele.
All groups identical in one allele and displaying either *41 (test group) or *1 or *2 (reference group) as the second allele were compared (Table 2
). Individuals with the *41 allele always had higher NT concentrations per dose unit and when corrected for CYP2C19 than did individuals with the *1 or *2 alleles. All comparisons on the background of a functional allele yielded either highly or at least borderline significant differences. Both comparisons without the background of a functional allele (other allele, *4) did not yield significant results. This was possibly because there were only two patients in that group.
The average effect (ASCOC) of the *41 allele compared with *1 or *2 on NT concentrations per dose unit and corrected for CYP2C19 was 39.8% (P = 0.000001; Fig. 1
), which is equivalent to a 1.95-fold change in the metabolic ratio (see Discussion, 1.3982 = 1.95; 36.0% when not corrected for CYP2C19). This indicates clinical significance not only when *41 is combined with a nonfunctional allele, as described in the current literature(9)(11), but also in combination with functional alleles.
The *10 allele.
All three patients with the *10 allele had higher NT concentrations per dose unit and corrected for CYP2C19 than did those with the *1 or *2 alleles. The average effect (ASCOC) of the *10 allele compared with *1 or *2 on NT concentrations per dose unit and corrected for CYP2C19 was 63.3% (P = 0.0009; Fig. 1
), which is equivalent to a 2.67-fold change in metabolic ratio (MR; 1.6332 = 2.67; 68.5% when not corrected for CYP2C19).
The *4 allele.
Individuals with the *4 allele always had higher NT concentrations per dose unit and corrected for CYP2C19 than did those with the *1 or *2 alleles (Table 2
). All comparisons on functionally equal background yielded either highly significant differences or showed a trend toward significance in most cases. The mean effect (ASCOC) of the *4 allele on NT concentrations per dose unit and corrected for CYP2C19 was 95.6% (P = 5 x 10–13), which is equivalent to a 3.83-fold change in MR (1.9562 = 3.83; 101.0% when not corrected for CYP2C19). This clearly demonstrates clinical significance not only when combined with a second dysfunctional allele but also in combination with functional alleles.
functional gene dose
From the results reported above, it is possible to calculate a quantitative "functional gene dose" (FGD). We defined FGD as 2 divided by the relative concentration expected for a specific allele combination. For the wild type with two functional alleles, this gives a functional gene dose of 2, which is, in this case, set arbitrarily for two functional genes divided by 1, which is the relative concentration expected for the homozygous wild type. If FGD is similarly calculated for other allele combinations, the results are as follows:
For a combination of *4 and *41, one would expect a combined, possibly multiplicative, effect of both above-mentioned ASCOC values, which give a combined ASCOC of 173% and a FGD of 0.73. The two patients with *4/*41 genotype in our population had a combined ASCOC of 160%.
To allow comparison with published data, we assigned a semiquantitative gene dose (SGD) of 1 for functional alleles, 0 for completely dysfunctional null alleles (e.g., *4), and 0.5 for the *41 and *10 alleles. On the basis of conventional classification (genotype-predicted phenotype), our patient population consisted of 1 gUM, 3 gIMs, and 46 gEMs. Because of the small numbers in two of the three groups, only borderline significance was found when NT concentrations were analyzed (P = 0.047). Instead, five groups could be formed based on SGDs (Table 3
), which led to highly significant concentration differences (P = 0.000008).
|
combined effects of cyp2c19 and cyp2d6
The combined influence of CYP2C19 gene dose and CYP2D6 SGD is shown in Table 4
. A significant difference in the mean values for NT, AT + NT, and the NT/AT ratio, but not AT, was observed. Particularly, NT and the NT/AT ratio displayed uniform patterns with higher values for low CYP2D6 SGDs and high CYP2C19 gene doses not only in the total values but in every column and line (Table 4
). There was one exception, a patient with a CYP2D6 SGD of 3 who displayed unusually high AT concentrations. The total concentration of AT + NT is determined by CYP2D6 SGD but not CYP2C19 gene dose. A multivariate linear regression model including age, sex, CYP2D6 SGD, and CYP2C19 gene dose showed that only CYP2D6 SGD (P = 0.00001; partial 
2 Pillai’s trace = 0.361) and CYP2C19 gene dose (P = 0.02; partial 2 Pillai’s trace = 0.187) contributed significantly to the model predicting NT concentrations per dose unit and weight.
dose recommendations
The mean concentrations achieved between days 7 and 21 (corrected for the five patients receiving doses other than 150 mg/day) were 91 µg/L AT, 78 µg/L NT, and 169 µg/L AT + NT. The corresponding values corrected for dose and body weight were 41.6, 34.1, and 75.4 µg/L · kg/mg, respectively. If one assumes that our patients achieved a representative mean concentration, dose recommendations can be derived from SGD-specific concentrations per dose unit. On the basis of AT + NT and the CYP2D6 SGD, the following doses relative to the applied standard dose can be calculated: for SGD of 0.5, dose = 74% of standard; for SGD of 1.0, dose = 84% of standard; for SGD of 1.5, dose = 100% of standard; for SGD of 2, dose = 126% of standard. Targeting just NT concentrations, as in NT therapy, the recommended doses would be as follows: for SGD of 0.5, dose = 57% of standard; for SGD of 1.0, dose = 75% of standard; for SGD of 1.5, dose = 99% of standard; for SGD of 2, dose = 152% of standard. In the future, ASCOC values and FGDs may be used to directly derive allele-specific dose recommendations.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Most importantly, the genetic basis of the large variation within the EM phenotype, including the subgroup of IMs, has not been obvious. The phenotype ranges (MRs of test drugs) for carriers of one or two functional alleles overlap extensively, and a clear-cut gene-dose effect has not been established(7)(8)(9). Although alleles with reduced function associated with the IM phenotype have been identified (*9 and *10)(31)(32), their frequency was too low to provide a sufficient explanation in Caucasians. Recently, a new mutation in the CYP2D6 5'-flanking region associated with a functional bimodality of the frequent *2 allele was described (*41)(9)(11). Together with the alleles *9 and *10, the presence of *41 may explain the majority of IMs in Caucasian populations when present in combination with a completely dysfunctional 0 allele. However, the gIM subgroup, as defined above, accounts for only a small portion (10–15%) of the total EMs. All carriers of one or two functional alleles and combinations of alleles with reduced and normal function are allocated to the gEM group without further differentiation. Moreover, there seems to be no agreement, as yet, regarding the use of the term IM (predicted from genotype). Some authors follow the definition from above(5), whereas others have used the term IM (predicted from genotype) for carriers of one functional and one completely dysfunctional allele (gEM according to the above classification)(14)(33)(34).
To date, the focus has been on finding correlations between CYP2D6 genotypes and the four populations that can be differentiated by several test drugs. Phenotyping with test drugs has been a valuable research tool, but it has been shown that there are substantial differences in the discriminatory power for CYP2D6 activity among the different test drugs in use(7)(8), and it is plausible to assume that therapeutic drugs are also different in that regard. It may therefore be more promising to directly study correlations between CYP2D6 genotype and its effects on therapeutic drugs rather than taking the roundabout way of assigning four predicted phenotypes with regard to test drugs. Dose recommendations could possibly be given more precisely when they are derived directly from the exact genotype rather than from test-drug-related phenotype or genotype-predicted phenotype, as suggested in a recent review(19).
In preselected populations designed to detect CYP2D6-related pharmacokinetic differences(33)(34), trimipramine and doxepin clearance differed significantly between carriers of two or only one functional allele. However, alleles with reduced function (*9, *10, and *41) were not considered, and single-dose studies in highly controlled populations of volunteers may not predict the real therapeutic relevance of genomic variations when patients are treated chronically with drugs because of nonlinear kinetics, tolerance, and/or compensatory homeostatic mechanisms. Clinical studies must be performed in the actual clinical environment in which the pharmacogenetic test is likely to be used(17). Two studies have quantitatively examined the correlation between CYP2D6 gene dose (number of functional genes) and NT concentrations(35)(36). Murphy et al.(36) also tested for an allele with reduced function (*10), arbitrarily assigned a mutation score of 0.5 to this allele, and showed a correlation between a mutation score and NT concentrations. However, both studies had not tested the *41 allele.
Studies in clinical patients have concentrated on PMs and UMs because potential benefits from pretherapeutic genotyping were expected to be most pronounced in individuals with extreme pharmacokinetics and because precise prediction of metabolic capacity within EMs has not been possible up to now(7)(8)(9). Because of their low prevalence (8–10% in a Caucasian population), 10 individuals must be genotyped to predict 1 PM or UM. This caused difficulty in demonstrating cost-effectiveness because the cost of therapeutic failure or adverse events in these patients is ill defined at present(17)(37). Therefore, except for a few specialized academic centers, pretherapeutic genotyping for CYP2D6 is currently not an accepted clinical tool for individualizing psychoactive drug therapy(37)(38).
The high number of alleles affecting enzyme function and the resulting plethora of genotypes has made it extremely challenging to assess the effects of CYP2D6 polymorphisms on therapeutic drug concentrations on a genotype level. Even in large studies, only a few individuals share the exact same allele combination(5)(8). To increase group size, many clinical studies involving therapeutic drugs report correlations between genotype-predicted phenotypes or the number of functional alleles and concentration(39)(40)(41). This practice, however, sacrifices important information inherent to all studies using genotyping and serum concentrations.
We demonstrate here that accurate predictions of serum concentrations are possible from CYP2D6 genotyping data even within the group of gEMs. It is startling that this can be demonstrated on the background of extra variability by another polymorphic enzyme (CYP2C19). Calculating ASCOC values increases the statistical power of genotype-based comparisons substantially by combining allele comparisons on various background alleles. Table 5
summarizes some key issues to consider when comparing the different methods to predict serum concentrations of therapeutic drugs from genotyping data. In our clinical population, the calculation of ASCOC values confirmed functional equivalence of the *2 and the *1 wild-type alleles as described previously(11). No change in concentration was found for *35, an allele that has been suggested to contribute to enhanced ultrarapid metabolism(13). In contrast, we report significantly higher NT concentrations for the dysfunctional alleles *41, *4, and *10 regardless of the second allele present.
|
Instead of counting the number of "functional" alleles and designating that number as the gene dose, we used the ASCOC values to calculate a quantitative FGD corresponding to the expected mean effect of an allele combination on the concentration of NT.
To facilitate comparison with published results, the quantitative data from ASCOC values and FGDs were simplified and SGDs were assigned. Highly significant differences in NT concentrations could be detected between the four subgroups of patients conventionally classified as gEMs and gIMs. In agreement with previous in vitro(11) and in vivo findings in NT-treated patients(36), we show highly significant and possibly clinically relevant differences for carriers of one or two functional alleles. We also provide the first evidence that carriers of a fully functional allele combined with an allele with reduced function can be distinguished from individuals with two functional genes (SGD, 1.5 vs 2.0).
In the future, ASCOC values could be used to individually predict serum concentrations for each allele combination rather than for (semiquantitative) gene-dose groups or genotype-predicted phenotypes. Once exact ASCOC values are established, they can be used to give dose recommendations, even taking into account subtle concentration differences as observed between *41 and *10 (concentrations, 40% or 63% of the standard). The concept facilitates metaanalyses, allows detailed and precise comparisons of the effects of polymorphic enzymes on different drugs, and could be useful for assessing interactions among drugs competing for the same polymorphic enzyme (comparison of comedication on identical genetic background)(42). This approach could generate more meaningful data than comparing genetically mixed populations and could help to limit study population sizes.
Two points need mentioning: (a) the results of this study are derived almost exclusively from EMs and IMs and cannot be extrapolated to PM or UM populations at present; and (b) when comparing our results for NT with MRs from the application of test drugs, the fact that MRs are calculated as the concentration of a test drug divided by the concentration of the metabolite formed must be considered. Therefore, MRs overestimate the effect of a functional change of an enzyme because this change is taken into account twice. A change in enzyme function increasing the parent drug concentration consecutively leads to a reduced metabolite concentration. Therefore, the best estimate for the effect of a change in the MR for a therapeutic drug is 
rather than MR or logMR.
combined effects of CYP2C19 and CYP2D6
Both CYP2D6 (SGD) and CYP2C19 gene dose influence AT, NT, and AT + NT steady-state concentrations per dose unit and the NT/AT ratio. The sum concentration of AT + NT is influenced predominantly by the change in NT concentration resulting from CYP2D6 polymorphisms. CYP2C19 influences both AT and NT concentrations, but any change in activity increases the concentration of one compound while decreasing the other, thus exerting no net effect on the total concentration of both compounds (Table 4
). CYP2C19 predominantly determines the ratio between those two compounds.
The sum AT + NT is used to guide AT therapy by therapeutic drug monitoring(2)(43), which implicates that CYP2D6 but not CYP2C19 genotyping may be useful in AT therapy. However, our Caucasian population consisted almost exclusively of CYP2C19 heterozygotes and only one patient with two dysfunctional alleles. CYP2C19 may have a greater impact in Asian populations with a higher number of PMs, where only little NT is formed and other pathways of elimination may predominate(39).
dose recommendations
The results of this study build on the sparse data on AT available in a recent report discussing CYP2C19- and CYP2D6-dependent dose recommendations(19). Following the current practice of guiding therapy with the sum concentration of AT + NT, the calculated optimum doses vary from 74% to 126% of the standard dose (CYP2D6 SGD of 0.5 and 2.0). Because no influence of CYP2C19 on the total concentration of AT + NT could be shown, dose recommendations regarding this enzyme do not seem to be relevant for gIMs and gEMs.
Should it be demonstrated that AT or NT alone correlates with clinical outcome, it may be useful to devise CYP2C19-based dose recommendations for AT and NT or CYP2D6-based recommendations for NT. For NT-based therapy, similar numbers may be expected. It is interesting to note that individuals with a CYP2D6 SGD of 1.5 displayed the exact mean concentration obtained for the whole population, indicating that the standard dose was optimal for this group of patients. Average doses recommended by the manufacturer may be considered as the pragmatic results of large-scale studies performed with genetically mixed populations. In Caucasians, this probably leads to doses suboptimal for carriers of two fully functional genes because of the unrecognized presence of PMs and IMs in the study populations(19).
One point regarding dose recommendations should not be forgotten. It is clear that an optimum initial dosing strategy will have to include not only genotypes but also body weight and possibly other factors.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
1 Nonstandard abbreviations: TCA, tricyclic antidepressant; AT, amitriptyline; NT, nortriptyline; PM, poor metabolizer; IM, intermediate metabolizer; EM, extensive metabolizer; UM, ultrarapid metabolizer; gEM, genotype-predicted extensive metabolizer; gPM, genotype-predicted poor metabolizer; gIM, genotype-predicted intermediate metabolizer; gUM, genotype-predicted ultrarapid metabolizer; ASCOC, allele-specific change of concentration on identical background; MR, metabolic ratio; FGD, functional gene dose; and SGD, semiquantitative gene dose. 
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  N. Hosono, M. Kato, K. Kiyotani, T. Mushiroda, S. Takata, H. Sato, H. Amitani, Y. Tsuchiya, K. Yamazaki, T. Tsunoda, et al. CYP2D6 Genotyping for Functional-Gene Dosage Analysis by Allele Copy Number Detection Clin. Chem., August 1, 2009; 55(8): 1546 - 1554. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. von Ahsen and M. Oellerich Chip-Based Genotyping: Translation of Pharmacogenetic Research to Clinical Practice Clin. Chem., July 1, 2007; 53(7): 1186 - 1187. [Full Text] [PDF]
|
|
|
|
|
|
J. K. Blievernicht, E. Schaeffeler, K. Klein, M. Eichelbaum, M. Schwab, and U. M. Zanger MALDI-TOF Mass Spectrometry for Multiplex Genotyping of CYP2B6 Single-Nucleotide Polymorphisms Clin. Chem., January 1, 2007; 53(1): 24 - 33. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. E. Kootstra-Ros, M. J. M. Van Weelden, J. W. J. Hinrichs, P. A. G. M. De Smet, and J. van der Weide Therapeutic drug monitoring of antidepressants and cytochrome p450 genotyping in general practice. J. Clin. Pharmacol., November 1, 2006; 46(11): 1320 - 1327. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. J. Gardiner and E. J. Begg Pharmacogenetics, Drug-Metabolizing Enzymes, and Clinical Practice Pharmacol. Rev., September 1, 2006; 58(3): 521 - 590. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Laika, S. Leucht, and W. Steimer ABCB1 (P-Glycoprotein/MDR1) Gene G2677T/A Sequence Variation (Polymorphism): Lack of Association with Side Effects and Therapeutic Response in Depressed Inpatients Treated with Amitriptyline. Clin. Chem., May 1, 2006; 52(5): 893 - 895. [Full Text] [PDF]
|
|
|
|
|
|
W. Steimer, K. Zopf, S. von Amelunxen, H. Pfeiffer, J. Bachofer, J. Popp, B. Messner, W. Kissling, and S. Leucht Amitriptyline or Not, That Is the Question: Pharmacogenetic Testing of CYP2D6 and CYP2C19 Identifies Patients with Low or High Risk for Side Effects in Amitriptyline Therapy Clin. Chem., February 1, 2005; 51(2): 376 - 385. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Valdes Jr and M. W. Linder Fine-Tuning Pharmacogenetics: Paradigm for Linking Laboratory Results to Clinical Action Clin. Chem., September 1, 2004; 50(9): 1498 - 1499. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
