Vous êtes sur la page 1sur 9

nature publishing group

articles

are Circulating Metabolites Important in DrugDrug Interactions?: Quantitative analysis of Risk Prediction and Inhibitory Potency
CK Yeung1, Y Fujioka1, H Hachad1, RH Levy1 and N Isoherranen1
The potential of metabolites to contribute to drugdrug interactions (DDis) is not well defined. The aim of this study was to determine the quantitative role of circulating metabolites in inhibitory DDis in vivo. The area under the plasma concentrationtime curve (auC) data related to at least one circulating metabolite was available for 71% of the 102 inhibitor drugs identified. of the 80 metabolites characterized at steady state, 78% had auCs >10% of that of the parent drug. a comparison of the inhibitor concentration/inhibition constant ([i]/Ki) ratios of metabolites and the respective parent drugs showed that 17 of the 21 (80%) reversible inhibitors studied had metabolites that were likely to contribute to in vivo DDis, with some metabolites predicted to have inhibitory effects greater than those of the parent drug. The in vivo drug interaction risks associated with amiodarone, bupropion, and sertraline could be identified from in vitro data only, when data pertaining to metabolites were included in the predictions. in conclusion, cytochrome p450 (Cyp) inhibitors often have circulating metabolites that contribute to clinically observed Cyp inhibition. It has long been known that metabolites may contribute to drug action and toxicity, act as inhibitors of drug elimination, or displace the drug from plasma or from tissue binding sites. However, the quantitative importance of metabolites in drug therapy has not been well characterized. Classic examples of drugs that have pharmacologically important (active) metabolites include morphine as a metabolite of codeine, phenobarbital as a metabolite of primidone, enalaprilat as a metabolite of enalapril, and desipramine as a metabolite of imipramine. In some cases, metabolites may constitute a risk factor for drug toxicity. In recognition of this, the US Food and Drug Administration issued the so-called MIST (metabolites in safety testing) guidelines (http://www.fda.gov/downloads/ Drugs/GuidanceComplianceRegulator yInformation/ Guidances/ucm079266.pdf), which establish a strategy for toxicology evaluation of circulating metabolites. Emphasis is placed on metabolites that circulate in concentrations >10% of the parent drug concentration at steady state. The International Conference on Harmonisation M3 guideline recently issued by the European Medicines Agency, on the other hand, focuses on metabolites that circulate at 10% of total drug-related material. The International Conference on Harmonisation and MIST guidelines are, however, specific for toxicity testing and do not relate to the study of the metabolites as inhibitors of metabolic enzymes or as displacers of binding. Although metabolites have received consistent attention with respect to their role in pharmacological action and in drug toxicity, their role in inhibitory drugdrug interactions (DDIs) has not been well characterized. This is perhaps due to the fact that there are no established examples of clinical interactions that are caused solely by an inhibitory metabolite in the absence of any inhibitory effect by the parent drug. Several clinically important inhibitors have recently been recognized to have circulating metabolites that have inhibitory effects. For example, hydroxyitraconazole as a metabolite of itraconazole, N-desmethyldiltiazem as a metabolite of diltiazem, and norfluoxetine as a metabolite of fluoxetine appear to contribute to in vivo DDIs.14 It is currently not known whether these three drugs are exceptions or whether they are examples of a more prevalent phenomenon of multiple inhibitors being responsible for a single interaction. In setting out to evaluate the role of inhibitory metabolites in clinically important DDIs, we recently determined that a majority (82%) of clinically used cytochrome P450 (CYP) inhibitors possess circulating metabolites, and that all known potent CYP inhibitors have circulating metabolites.5 In this study, the exposures to metabolites after administration of the parent compound were quantitatively analyzed. The aim of this study was to determine

1Department of Pharmaceutics, University of Washington, Seattle, Washington, USa. Correspondence: N Isoherranen (ni2@u.washington.edu)

Received 8 July 2010; accepted 1 September 2010; advance online publication 1 December 2010. doi:10.1038/clpt.2010.252
CliniCal pharmaCology & TherapeuTiCs | VOLUME 89 NUMBER 1 | jaNUaRy 2011 105

articles
a

10,000 1,000

Multiple dose Single dose 37% 46% 17%

AUCM/AUCP ratio

100 10 1 0.1 0.01 0.001

10,000 1,000

Weak Moderate Potent

AUCM/AUCP ratio

100 10 1 0.1 0.01 0.001

Figure 1 Distribution of metabolite-to-parent aUC ratios for all the inhibitor drugs analyzed. (a) The aUCM/aUCP ratios (y-axis) for all the inhibitor drugs referred to in supplementary Appendixes 1 and 2, with each bar representing a specific metaboliteparent pair. Blue bars refer to steady-state data and red bars to single-dose data. The percentages above the bars indicate the fractions of inhibitors in each group categorized by the values of the aUCM/aUCP ratio (aUCM/ aUCP > 0.1; 0.1 < aUCM/aUCP < 1; aUCM/aUCP 1). (b) The distribution of aUCM/aUCP ratios for the inhibitor drugs for which in vivo drugdrug interaction data of clearance and aUC changes of marker substrates. The drugs were classified according to the system recommended by the US Food and Drug administration (http://www.fda.gov/cder/drug/drugInteractions/) as potent (>5-fold increase in aUC), moderate (>2- but <5-fold increase in aUC), or weak (>1.25- but <2-fold increase in aUC) inhibitors of the target enzyme. aUC, area under the plasma concentrationtime curve; M, metabolite; P, parent.

the relative importance of circulating metabolites in clinical DDIs. The University of Washington Metabolism and Transport Drug Interaction Database was used to collect the relevant in vitro enzyme inhibition data and clinical interaction data, and in vivo interaction potential was analyzed in accordance with US Food and Drug Administration guidelines.
Results Prevalence of circulating metabolites

The metabolite-to-parent area under the plasma concentration time curve (AUC) ratios were calculated for 72 of the 102 previously identified inhibitors5 on the basis of the availability of data. Because some of the inhibitors have several circulating metabolites, this resulted in 120 data sets of AUCM/AUCP ratios (Figure 1). Within these data, 37% of the metabolites had an AUC greater than that of the parent drug, and an additional 46% had an AUC that was between 10 and 100% of that of the parent drug. Only 17% of all the metabolites were determined to be
106

insignificant in terms of their AUCs in relation to the AUC of the parent drug (AUCM/AUCP <0.1). Figure 1 depicts the distribution of the inhibitors based on metabolite-to-parent AUC ratio with steady-state and single-dose data. The individual metabolite-to-parent pairs and the AUC values used for the analysis are summarized in Supplementary Appendix 1 online (steady-state data) and Supplementary Appendix 2 online (single-dose data). Supplementary Appendix 3 online shows the list of circulating metabolites for which no pharmacokinetic characterization was available. When the data were analyzed separately for weak, moderate, and potent inhibitors, the distribution of AUCM/ AUCP ratios was similar to that of all the inhibitors together.
Prediction of relative contributions of metabolites to in vivo DDIs

Quantitative predictions of the importance of metabolites in in vivo DDIs were done by calculating the IPM/IPP ratios (inhibitory potency of the metabolite as [I]/Ki, over the inhibitory potency of
VOLUME 89 NUMBER 1 | jaNUaRy 2011 | www.nature.com/cpt

articles
table 1 the use of IPM/IPP ratios to determine the relative importance of the metabolites in comparison to the respective parent drugs
p parent (p) amiodarone Bupropion metabolite (m) N-Desethylamiodarone Threobupropion Hydroxybupropion Erythrobupropion Sulfinpyrazone Venlafaxine atorvastatin amodiaquine Clomipramine Sertraline Risperidone Sertraline Haloperidol Risperidone Fluoxetine Citalopram Propranolol Sulfinpyrazone Itraconazole Ranolazine Chloroquine Imatinib Sulfinpyrazone sulfide N-Desmethylvenlafaxine atorvastatin lactone Desethylamodiaquine N-Desmethylclomipramine N-Desmethylsertraline Paliperidone N-Desmethylsertraline Reduced haloperidol Paliperidone Norfluoxetine Didesmethylcitalopram Desmethylcitalopram 4-Hydroxypropranolol Sulfinpyrazone sulfone Hydroxyitraconazole O-Demethyl ranolazine Monodesethylchloroquine N-Desmethylimatinib Cyp 2C9 2D6 2D6 2D6 2C9 2D6 3a 2D6 2D6 2D6 3a 3a4 2D6 2D6 2D6 2C19 2C19 2D6 2C9 3a4 3a4 2D6 2C9 3a 2D6 Carbamazepine atomoxetine Omeprazole Quinidine Carbamazepine-10,11-epoxide N-Desmethylatomoxetine Omeprazole sulfone Omeprazole sulfone 3-Hydroxyquinidine Quinidine N-oxide
CyP, cytochrome P450. *The specific values and their matching references. aUS Food and Drug administration. Clinical Reviews: Ranolazine. Drugs@FDa 2007. bUS Food and Drug administration. Clinical Review: Imatinib. Drugs@FDa 2002.

m [i]/Ki 0.026 0.014 [i] (mol/l) Ki (mol/l) 1.5 1.9 3.2 0.37 2.3* 5.4 13 1.7 27 20 0.9 4.1 7.9 16 80 3.5 0.24 16 0.2 7.7 56 1.0 73 26 104 15.5 40.3 13.7 13.5 4.8 0.019 0.38 0.30 0.60 53.5 16.0 8.2 188 2.3 43.5 [i]/Ki 0.67 0.36 0.24 0.22 0.19 0.047 0.004 0.20 0.062 0.009 0.001 0.15 0.033 0.003 2.0 0.002 0.002 0.046 0.031 0.042 0.009 0.013 0.034 0.099 0.10 0.089 0.001 0.046 0.002 0.13 0.014 ipm/ipp reference(s) 25.3 24.8 17.0 15.3 6.0 6.0 5.5 5.1 4.2 2.4 2.0 1.6 1.3 1.0 0.92 0.85 0.85 0.49 0.40 0.38 0.35 0.32 0.23 0.092 0.088 0.086 0.069 0.058 0.0025 0.0019 0.0002 10,11* 12 12 12 13 14 15,16* 17 18 19 20 21,22 23 20 24 25 25 26,27* 13 3 a 28 b,29* b,29* b,29* 30 31 32 33 34 34

[i] (mol/l) Ki (mol/l) 2.5 0.3 94.6* 21

17.8 0.23 0.06* 0.08 0.24 0.086 0.020 0.33 0.023 0.020 0.37 0.20 0.24 0.42 17.8 0.94 2.2 0.62 8.6*

229 30 90 2.1 16 23 67 3.5 0.89 6.9 0.2 87.3 87.3 4.5* 229 8.5 90.0 15.3 59.0 8 7.5

0.031 0.0078 0.0007 0.039 0.014 0.0038 0.0003 0.094 0.026 0.0028 2.2 0.002 0.003 0.094 0.078 0.11 0.025 0.040 0.15 1.1 1.1 1.0 0.017 0.80 0.008 70

5.0 0.93 0.003* 0.82 0.49 0.14 0.046 0.53 0.008 0.046 0.38 0.015 0.13 0.046 2.3 1.09 0.89 0.20 1.4*

2C19 3a 2C19 3a 2D6 2D6

30.5 0.59 1.6 1.9

29.6 34.0 2.0 201.0 0.03

the parent as [I]/Ki) for all inhibitors for which circulating concentrations of the metabolite and parent were available and for which in vitro Ki values had been determined. Table 1 shows a summary of the calculated IPM/IPP ratios. The metabolites were predicted to be most significant in vivo inhibitors for 11 of the 21 (52%) inhibitor drugs analyzed, whereas five metabolites belonging to four inhibitors (19%) were predicted not to contribute to the in vivo interaction (IPM/IPP ratio <0.1). Notably, the two inhibitors that are commonly recognized to have inhibitory metabolites, itraconazole and fluoxetine, had IPM/IPP ratios <1. For three of the potent or moderate inhibitors analyzed, namely, amiodarone, amodiaquine, and bupropion, the IPM/IPP ratios ranged from 5
CliniCal pharmaCology & TherapeuTiCs | VOLUME 89 NUMBER 1 | jaNUaRy 2011

to 25, demonstrating that the metabolites are predicted to be the main contributors to interactions. On the other hand, omeprazole sulfone, quinidine N-oxide, and 3-hydroxy quinidine had IPM/ IPP ratios <0.01, corresponding to a prediction of insignificant contribution to the in vivo interactions.
Role of circulating metabolites in risk analysis of inhibitory interactions

Given that metabolites were predicted to quantitatively contribute to selected in vivo DDIs, the effect of circulating metabolites on preclinical risk analysis of potential DDIs was determined. Only inhibitors for which in vivo interaction data with accepted probes
107

articles

108

table 2 evaluation of potential drugdrug interaction risk using parent and metabolite [I]/Ki data according to the FDA guidance
D+m [i]/Ki 0.53 1.02 R R D m P P P P 16 1.2 0.099 29 0.44a 0.22 0.18a 19 0.056 0.25a 0.33 51 1.3 0.25 0.011 0.095 0.223 0.008 0.09 3a4 2D6 3a4 2C9 2D6 50 mg, t.i.d., 3 days Imipramine 200 mg, b.i.d., 13 days (s)-Warfarin 1.9 200 mg, q.d., 29 days Pimozide 1.4 150 mg, q.d., 8 days Imipramine 1.7 1,000 mg, b.i.d., 7 days Simvastatin 1.9 5.0 2.7 0.32a 0.53 36 8.1 0.027 0.054 0.001 0.004 0.039 0.026 0.020 0.15 2.1 0.11 0.001 0.010 R 0.082 0.031 0.25 2.4 R R R P 1.2 89 0.67 0.004 19 L L L L P R L L P R L R R R R R R P L P R R 49 50 R R P L c 45 22 47 c 46 22 48 L L L L P R L 37 29 b 39 40 42 43 38 29 29 39 41 41 44 FDa guidance D+m P L reference auC 35 12 [i] 36 12 Cyp 2C9 2D6 4.0 2.3 2D6 8.1 0.10 2D6 3a 3a4 2C19 40 mg, q.d., 7 days 3a4 2D6 50 mg, single Dextromethorphan 42.6 0.57a 20 mg, q.d., 8 days Nifedipine 1.3 0.62 0.003 Proguanil 1.5 1.2a 0.62 200 mg, q.d., 4 days Lovastatin 36.4 0.94 60 1.1 400 mg, q.d., 7 days Simvastatin 2.9 8.6 1.0 1.4 400 mg, b.i.d., 9 days Metoprolol 1.2 8.6 1.1 1.4 60 mg, q.d., 8 days Desipramine 7.8 1.4a 7.9a 1.6a 0.43 0.30 150 mg, b.i.d., 12 days Desipramine 5.2 0.67 0.032 0.43 0.25 200 mg, b.i.d., 10 days (s)-Warfarin 2.1 1.8 0.019 1.2 0.51 inhibitor dose Victim drug parent drug (D) metabolite (m) auC change of victim drug (observed) [i] (mol/l) [i]/Ki [i] (mol/l) [i]/Ki

parent

metabolite

amiodarone

N-Desethylamiodarone

Bupropion

Erythrohydrobupropion

Hydroxybupropion

Threohydrobupropion

Fluoxetine

Norfluoxetine

Imatinib

N-Desmethylimanitib

Itraconazole

Hydroxyitraconazole

Omeprazole

Omeprazole sulfone

Quinidine

Quinidine N-oxide

3-Hydroxyquinidine

Ranolazine

O-Desmethylranolazine

Sertraline

N-Desmethylsertraline

Sulfinpyrazone Sulfinpyrazone sulfide

Sulfinpyrazone sulfone

Venlafaxine

N-Desmethylvenlafaxine

The Ki values used are the same as those in table 1.

aUC, area under the plasma concentrationtime curve; b.i.d., twice daily; CyP, cytochrome P450; FDa, US Food and Drug administration; L, likely; P, possible; q.d., once daily; R, remote; t.i.d., three times daily.

VOLUME 89 NUMBER 1 | jaNUaRy 2011 | www.nature.com/cpt

aThe concentrations of the inhibitor and its metabolites were calculated with dose normalization. bFDa. Clinical Reviews: Imatinib. Drugs@FDa 2002. cFDa. Clinical Review: Ranolazine. Drugs@FDa 2007.

articles
a
3 Amiodarone Bupropion 2.5 AUC change (observed) Fluoxetine Imatinib (3A4) Itraconazole Quinidine Parent

1.5 1.25 1

Omeprazole (3A4) Ranolazine Sertraline (2D6) Sertraline (3A4) Venlafaxine

Omeprazole (2C19) Sulfinpyrazone

Imatinib (2D6)

0.5 0.01 0.1 [I]/Ki 1 10

b
3

Parent + metabolite Amiodarone 2.5 Bupropion Fluoxetine Imatinib (3A4) Itraconazole Quinidine

AUC change (observed)

2 Omeprazole (3A4) Ranolazine Sertraline (2D6) Venlafaxine Omeprazole (2C19) Sertraline (3A4) Sulfinpyrazone

of the parent drugs effect alone (Table 2). Metabolites of 6 of the 11 compounds were qualitatively predicted to be associated with equal or greater risk of causing DDIs than the parent inhibitor. Five compoundsamiodarone, bupropion, ranolazine, sertraline, and venlafaxinewere classified as representing remote risk for in vivo interactions when only the effect of the parent drug was considered. Two of these in vivo inhibitors, amiodarone (CYP2C9) and sertraline (CYP3A4), were categorized as possible risk and one, bupropion (CYP2D6), as likely risk when inhibition by known circulating metabolites was also considered; the in vivo interactions caused by ranolazine (CYP3A4) and venlafaxine (CYP2D6) were not predicted with the available in vitro inhibition data. Figure 2 shows the correlation between in vitro risk assessment and the magnitude of in vivo DDIs when only the parent drug is considered and when parent and metabolites are analyzed together. As shown in Figure 2, the moderate inhibitor amiodarone and the potent inhibitor bupropion were predicted to be in vivo inhibitors only when their metabolites were also included in the analyses. It is important to note that, for sertraline, the interaction with CYP3A4 could be predicted when the metabolite effect was taken into account, whereas the CYP2D6 interaction could not.
Quantitative in vitrotoin vivo extrapolation of DDIs

1.5 1.25 1

Imatinib (2D6)

0.5 0.01 0.1 [I]/Ki 3 1 10

2.5 AUC change (observed)

False negative

True positive

True positive

2 False negative True positive True positive

1.5 1.25 1

True negative

False positive

False positive

0.5 0.01 0.1 [I]/Ki 1 10

Figure 2 Evaluation of drugdrug interaction (DDI) risk using [I]/Ki data for inhibitor drugs presented in table 2. (a,b) The correlation between in vitro risk assessment and the magnitude of in vivo DDIs when (a) only the parent drugs action is considered and (b) when parent and metabolites are analyzed together. (c) The nine zones for prediction in accordance with a and b, including true negative, true positive, false negative, and false positive. aUC, area under the plasma concentrationtime curve.

were available, allowing establishment of in vivo interaction risk, were included in the analysis. The [I]/Ki ratios were calculated for the inhibitors and metabolites for which sufficient data were available. Of the 11 inhibitor drugs included in the analysis, 6 (54%) were classified as possible or likely interactions on the basis
CliniCal pharmaCology & TherapeuTiCs | VOLUME 89 NUMBER 1 | jaNUaRy 2011

In vivo DDIs are commonly rationalized by the use of static or dynamic models that allow extrapolation of in vitro inhibition potency to in vivo interactions. The existence of circulating inhibitory metabolites could cause a significant underprediction of the existing interactions and, if metabolites are not accounted for, could lead to a poor understanding of the characteristics of specific interactions. In order to determine whether the presence of circulating metabolites leads to significant underpredictions in known, well-characterized DDIs, the effects of inhibitory metabolites in the accuracy of in vitrotoin vivo predictions of interaction magnitude were tested using classic static prediction methods. Table 3 shows the predicted and observed interactions for the 10 inhibitors analyzed. Interactions were included in the analysis wherever in vivo data with a well-characterized probe were available and [I]/Ki ratios could be calculated for both the parent drug and the metabolite. Both maximum plasma concentration (Cmax) and average steady-state plasma concentration (Cav) were used in the analysis, and no apparent differences were observed between the predictions using these two values. This might be because, at steady state, there is minimum fluctuation in the inhibitor concentrations over the dosing interval, and therefore Cmax and Cav values are similar. For amiodarone, bupropion, fluoxetine, and sulfinpyrazone, improved quantitative predictions of the in vivo interactions were obtained when metabolites were accounted for, whereas for the remaining six inhibitors, no significant differences in prediction accuracy were observed. The fact that, in 4 of the 10 inhibitors analyzed, quantitative predictions were possible only after accounting for the metabolites suggests that knowledge of circulating metabolites and their inhibitory potency should be obtained prior to quantitative modeling of the observed interaction.
DIscussIon

This study was aimed at determining the extent to which circulating metabolites contribute to in vivo DDIs in the context
109

articles

table 3 Predicted vs. observed Auc values calculated with Cmax and Cav
auC change of victim drug Parent (P) Metabolite (M) Amiodarone N-Desethylamiodarone Bupropion Erythrohydrobupropion Hydroxybupropion Threohydrobupropion Fluoxetine Norfluoxetine Imatinib N-Desmethylimanitib Itraconazole Hydroxyitraconazole omeprazole Omeprazole sulfone Quinidine Quinidine N-oxide 3-Hydroxyquinidine Ranolazine O-Desmethylranolazine sertraline N-Desmethylsertraline sulfinpyrazone Sulfinpyrazone sulfide Sulfinpyrazone sulfone predicted (Cmax) Cyp 2C9 2D6 inhibitor dose 200 mg, q.d., 6.5 weeks 150 mg, b.i.d., 12 days Victim drug Phenytoin Desipramine observed 1.4 5.2 p 1.01 1.03 p+m 1.35 1.8 predicted (Cav) p 1.01 1.01 p+m 1.31 1.5

2D6 2D6 3a4 3a4 2D6

20 mg, q.d., 21 days 400 mg, b.i.d., 9 days 200 mg, q.d., 4 days 20 mg, q.d., 8 days 50 mg, single

Desipramine Metoprolol Midazolam Nifedipine Desipramine

4.4 1.2 10.8 1.3 3.2

2.41 1.79 10.9 1.0 6.0a

3.35 1.85 12.5 1.0 6.0a

2.47 8.1 1.0

3.61 10.4 1.0

3a4 2D6 2C9

1000 mg, b.i.d., 7 days 50 mg, q.d., 10 days 200 mg, b.i.d., 13 days

Simvastatin Desipramine (s)-Warfarin

1.9 1.4 1.9

1.06 1.0 1.19

1.08 1.01 2.61

1.07 1.0

1.1 1.01

, The Cav was not available; aUC, area under the plasma concentrationtime curve; b.i.d., twice daily; Cav, average steady-state plasma concentration; Cmax, maximum plasma concentration; CyP, cytochrome P450; q.d., once daily.
aThe concentrations of the inhibitor and its metabolites were calculated with dose normalization.

of preclinical risk assessment and development of quantitative methods of predicting in vivo DDIs. The analysis shows that a majority of the clinically important CYP inhibitors have metabolites that circulate at appreciable concentrations and that one-third of the known metabolites are more abundant in the circulation than the parent drug. However, the CYP inhibiting potency has been published only for some metabolites. It is possible that unpublished negative inhibition data that were not available for this analysis exist. The available data for known metabolites were evaluated to determine whether these metabolites would alter the preclinical risk assessment of known clinical CYP inhibitors. Surprisingly, for 3 of the 11 inhibitors evaluated, the risk assessment changed from remote likelihood of DDIs to possible or likely with the inclusion of metabolites in the analyses. This suggests that determination of Ki values for quantitatively important metabolites can be useful when no interactions or only weak interactions are predicted with the parent drug. On the other hand, testing for inhibitory potential of metabolites might be of limited usefulness if the parent drug on its own has an identified risk in the likely category, given that the true magnitude of the in vivo interaction will be determined in specific in vivo DDI studies. Interestingly, all the compounds that were missed in the risk analysis (amiodarone, bupropion, ranolazine, sertraline, and
110

venlafaxine) when only the data related to the parent inhibitor drug were included, had metabolites with IPM/IPP ratios >0.1, suggesting that the metabolites of these compounds may have a critical role in causing in vivo DDIs. On the other hand, omeprazole sulfone, quinidine N-oxide, and 3-hydroxy quinidine, which were predicted in the risk analysis to have only remote potential to cause DDIs, had IPM/IPP ratios <0.1. A possible explanation for the incorrect assignment of risk for the ranolazinesimvastatin interaction could be the involvement of transporters such as P-glycoprotein in the in vivo interaction. Ranolazine has been shown to cause a 1.5-fold increase in in vivo digoxin concentrations, suggesting that it is a P-glycoprotein inhibitor (http://www.accessdata.fda.gov/drugsatfda_docs/ nda/2006/021526_s000_Ranexa.cfm). Given that simvastatin has been shown to be a P-glycoprotein substrate,6 it is possible that the in vivo interaction is due to P-glycoprotein inhibition during simvastatin absorption. Although data for the other probes, inhibitors, and metabolites are not available, it is possible that transporter-mediated interactions contribute to the other underpredictions as well. In addition, metabolites may be sequestered in the hepatocytes, and consequently plasma concentrations may not accurately reflect the intracellular free inhibitor concentrations. Such sequestration is not accounted for in the current analysis.
VOLUME 89 NUMBER 1 | jaNUaRy 2011 | www.nature.com/cpt

articles
A special consideration may be warranted in the case of compounds that are multi-CYP inhibitors in vitro. These compounds are typically analyzed using a rank-order approach.7 Because inhibitor depletion prevents the formation of significant amounts of metabolites in in vitro inhibition assays, rank-order analyses in vitro are likely to reflect the potency of the parent drug. If metabolites do not follow the same rank order, or inhibit only a subset of the CYPs that the parent compound inhibits, the rank-order approach may fail. For the three multi-CYP inhibitors included in this analysis, the rank order for the metabolite was similar to that observed for the parent drug when reversible inhibition was considered, but the data set is not representative of the overall multi-CYP inhibitors. The recently published MIST guidance recommends toxicology review of metabolites that circulate in concentrations >10% of those of their parent drugs. Although the MIST guidance does not address the role of metabolites in DDIs, our study analyzed the available data with reference to the 10% cutoff value of the MIST guidance. On the basis of the [I]/Ki ratios of the metabolites in relation to the parent compounds, all the metabolites that were predicted to contribute to in vivo interactions (with the exception of 4-hydroxypropranolol) had AUCs >10% of those of their respective parent drugs at steady state. Interestingly, there was no correlation between the contributions of the metabolites to interactions (IPM/IPP ratio) and the relative abundance of the metabolites in plasma; this is because many metabolites were found to be more potent inhibitors than the parent drug even though they circulated at lower concentrations than the parent drug. Given these findings, the 10% cutoff value for AUC appears appropriate for inhibition testing of metabolites. On the basis of these analyses, predicting the [I]/Ki ratio for abundant metabolites (as well as for the parent compound) can be helpful in the drug discovery and development process even when the parent compound is not inhibitory. This also requires that the methodology should be established for predicting [I], not only for the parent drug but also for the metabolite. The inhibitory metabolites included in this analysis may not be the sole circulating metabolites of the given inhibitor drug, but were the only ones with published in vitro inhibition data and in vivo AUC values. For example, 5-hydroxyomeprazole is more abundant than omeprazole sulfone in the circulation after the administration of omeprazole. However, no in vitro inhibition data were available for 5-hydroxyomeprazole, and therefore this metabolite could not be analyzed. Similarly, atorvastatin has several metabolites that circulate at concentrations similar to that of atorvastatin itself, but the inhibitory potencies of these metabolites have not been determined. On the other hand, although itraconazole has two other metabolites, keto-itraconazole and N-desalkyl-itraconazole, that inhibit CYP3A4 potently, their steady-state concentrations have not been determined. The existence of multiple inhibitory metabolites for a given drug would further increase the relative contributions of metabolites as a whole in comparison to the parent drug. The role of inhibitory metabolites in quantitatively predicting the magnitude of DDIs may be less than would be expected from the abundance of the metabolites. In the analysis of quantitative
CliniCal pharmaCology & TherapeuTiCs | VOLUME 89 NUMBER 1 | jaNUaRy 2011

predictions, inclusion of metabolites did not always improve the accuracy of the prediction, and the increase in average predictive accuracy was only modest. However, in the case of some of the inhibitors, significant improvement in predictions was observed. Given that data related to plasma and microsomal unbound fractions are not available for a majority of these compounds, the quantitative prediction may not accurately reflect the true relative importance of all of the metabolites analyzed. Despite this, the fact that so many inhibitors have circulating inhibitory metabolites may have implications for overall correlation analyses of in vitro-to-in vivo extrapolations because an accurate model should incorporate all inhibitory species. In conclusion, these results suggest that metabolites or candidates for major circulating metabolites in vivo should be investigated, in addition to the parent drug for inhibition potential in vitro unless clinical interaction studies are already warranted because of the parent compounds inhibitory effect. Although the available data in the literature do not allow for comprehensive evaluation, these results suggest that circulating metabolites contribute to in vivo DDIs. Further studies are required for an in-depth understanding of their overall clinical significance.
Identification of the inhibitors and their metabolites. We searched the

MethoDs

University of Washington Metabolism and Transport Drug Interaction Database (MTDI database: http://www.druginteractioninfo.org) to identify known CYP inhibitors. All the inhibitors selected for analysis had to be drugs currently available in the US market; herbal products, combination therapies, and oral contraceptives were excluded from the analysis. The complete list of identified inhibitors and their CYP inhibition characteristics was previously published previously.5 The AUC values for the parent inhibitor drugs and their metabolites were extracted from the MTDI database by searching for results of DDI studies in which the concentrations of the compounds as well as their metabolites were measured. If no data related to metabolites were available in the MTDI database, a PubMed search was conducted to extract data related to circulating metabolites. Finally, product labels and Micromedex were used as references for data related to circulating metabolites. For calculation of metabolite-to-parent AUC ratios, the following criteria were used: 1. The concentration levels of the metabolite and the parent drug were required to be from the same study, and complete AUCs (dosing interval at steady state or 0infinity after single dose) for both had to be available. 2. Steady-state values for the parent drug and its metabolites were used if available. If steady-state pharmacokinetics data were not available, single-dose kinetics were accepted. The ratios between the AUCs of the metabolite and the parent were calculated either by using the steady-state dosing interval AUC or by using the AUC to infinity after a single dose. 3. For compounds for which data from multiple dosing levels were available, the highest AUCM/AUCP ratio was used. All AUC values were converted to molh/ml units using the molecular weight of the parent drug and the metabolite, before calculating the ratio between the AUCs of the metabolites and the respective parent drugs.
Determination of the magnitude of in vivo DDIs and of the potency of inhibition of parent drug and metabolites. The MTDI database was que-

ried to retrieve all reported in vivo interactions (defined as resulting in a 20% increase in the AUC or decrease in clearance of the victim drug) for the inhibitors that were previously identified as having circulating
111

articles
metabolites. From the resulting list of in vivo interaction studies, the ones conducted with known marker substrates (http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/ ucm072101.pdf) were retained for further analysis. All data related to in vivo AUC changes determined using marker substrates were extracted, and the change in the AUC of the marker of interest was recorded for analysis. The concentration levels of the inhibitor drug and its metabolites as measured in the interaction study were recorded if available. For studies that did not measure the plasma concentrations of the inhibitor drug and its metabolites, data related to the same dosing regimen was accessed from the literature in order to calculate steady-state Cmax and Cav values for the inhibitor drug and its metabolites. If data pertaining to the inhibitor drug and its metabolites were not available at the dose level used in the in vivo interaction studies, the concentrations were dosenormalized for the purposes of the risk analysis. The MTDI database was queried to extract the in vitro Ki values for the identified metabolites and the parent inhibitor drugs. Inhibitor drugs or their metabolites that were known inactivators of a given enzyme in vitro were excluded from further analysis for that enzyme. In case there were multiple reversible Ki values for the parent inhibitor or its metabolites, the Ki value used in the analysis was selected according to the following criteria: (i) the metabolite and the parent drug had been tested in the same study against the same CYP and same probe; (ii) if data related to both metabolite and parent drug were not available in the same study, two different studies that had used the same probe were chosen; (iii) in the absence of both these categories of data, results were taken from two studies that had used the same microsomal protein concentration. For equivalent data, the study with the lowest microsomal protein concentration was used.
Relative importance of circulating metabolites. In order to predict the Quantitative prediction of in vivo DDIs. In vivo DDI data for 10 inhibitor

drugs, obtained with a probe substrate having a known fmCYP (fraction metabolized via specific CYP), were used to estimate the role of metabolites in quantitative in vitro-to-in vivo extrapolations of inhibitory interactions. In vivo studies were included in the quantitative analysis only if they had been conducted with a reliable probe with a defined fm (fraction metabolized) and if the Cmax and Cav were available in the DDI study or in another study with a similar dosing regimen. If several DDI studies were available, the study with the highest AUC change was selected so as to consider the hypothetical worst-case scenario. The total Cmax and Cav values for the inhibitor drug as well as for its metabolites were used in the predictions as [I]. In vitro-to-in vivo extrapolations were calculated from the equation: AUC i 1 = AUC fmCYP + (1 fmCYP ) n 1 + j [I] j K ij (1)

where the fmCYP value for each marker substrate was included according to the following list: fmCYP3A4: midazolam 0.94, nifedipine 0.71, simvastatin 0.99; fmCYP2D6: desipramine 0.877, metoprolol 0.828; f mCYP2C9: phenytoin 0.75, (S)-warfarin 0.87 (refs. 8,9). The effect of multiple inhibitors was accounted for by summing the [I]/Ki ratios.
suPPleMentARY MAteRIAl is linked to the online version of the paper at http://www.nature.com/cpt AcknowleDgMents This study was supported in part by National Institutes of Health grant P01 GM32165 and the Elmer M. and joy B. Plein Fellowship for Excellence in Pharmacy Education (C.K.y.) at the School of Pharmacy, University of Washington. The European Medicines agencys draft guidance on Guideline on the Investigation of Drug Interactions (CPMP/EWP/560/95/Rev. 1-Corr.) includes guidance on testing of metabolites in drugdrug interactions. conFlIct oF InteRest The authors declared no conflict of interest.
2010 american Society for Clinical Pharmacology and Therapeutics

relative contributions of the metabolites in in vivo DDIs, the [I]/Ki ratios were calculated for all inhibitors and metabolites for which there were sufficient in vitro data. The clinically relevant steady-state average total concentrations of the drug and its metabolites (measured in the same study and at the same dose level) were used as [I], and the Ki values were obtained as described above. The [I]/Ki ratio was calculated for the metabolites and parent drugs separately, and the ratio between the [I]/Ki of the metabolite and the relevant parent drug (IPM/IPP) was calculated. Because the [I]/Ki ratios for the metabolite (IPM) and for the parent (IPP) are indicative of their in vivo inhibitory potential (IP), the ratio of the two can be used as a predictor of the relative in vivo contribution of metabolites to CYP inhibition. The predicted contribution of the metabolites was classified as follows: (i) if the IPM/IPP ratio was 0.1, the metabolites were not predicted to contribute to in vivo interactions; (ii) if the IPM/IPP ratio was 1, the metabolite was predicted to be the main inhibitory species causing the in vivo interaction; and (iii) if 0.1 < IPM/IPP < 1, the parent drug was predicted to be the main inhibitor causing the DDI, with the metabolite also categorized as likely to make a contribution. A risk analysis for inhibitory interactions was conducted in accordance with the US Food and Drug Administration guidance for testing parent drugs (http://www.fda.gov/downloads/Drugs/ GuidanceComplianceRegulatoryInformation/Guidances/ucm079266. pdf). Inhibitor drugs were included in this analysis only if there was availability of in vivo interaction data with a known probe. The steady-state concentrations of the parent inhibitor drug (total Cmax) and the in vitro Ki values were used to compute the [I]/Ki ratio for the inhibitors and their metabolites. In order to evaluate the highest risk, the in vivo studies with the most sensitive probes for the inhibited CYP and the highest in vivo interactions were included in the analysis. On the basis of the [I]/Ki ratio, the potential of the drugs to cause in vivo DDIs was classified as likely ([I]/Ki >1), possible (1 > [I]/Ki > 0.1), or remote (0.1 > [I]/Ki) for the parent drug and for the metabolite, separately. For the purpose of predicting the combined risk associated with the metabolites and the parent drug, the [I]/Ki ratios were summed, and the sum was classified similar to the individual compounds.
112

1. 2. 3. 4. 5.

Risk analysis using data related to metabolites and parent drugs.

6. 7.

8.

9.

Zhang, X., jones, D.R. & Hall, S.D. Prediction of the effect of erythromycin, diltiazem, and their metabolites, alone and in combination, on CyP3a4 inhibition. Drug Metab. Dispos. 37, 150160 (2009). Zhang, X., Quinney, S.K., Gorski, j.C., jones, D.R. & Hall, S.D. Semiphysiologically based pharmacokinetic models for the inhibition of midazolam clearance by diltiazem and its major metabolite. Drug Metab. Dispos. 37, 15871597 (2009). Isoherranen, N., Kunze, K.L., allen, K.E., Nelson, W.L. & Thummel, K.E. Role of itraconazole metabolites in CyP3a4 inhibition. Drug Metab. Dispos. 32, 11211131 (2004). Stevens, j.C. & Wrighton, S.a. Interaction of the enantiomers of fluoxetine and norfluoxetine with human liver cytochromes P450. J. Pharmacol. Exp. Ther. 266, 964971 (1993). Isoherranen, N., Hachad, H., yeung, C.K. & Levy, R.H. Qualitative analysis of the role of metabolites in inhibitory drug-drug interactions: literature evaluation based on the metabolism and transport drug interaction database. Chem. Res. Toxicol. 22, 294298 (2009). Hochman, j.H. et al. Interactions of human P-glycoprotein with simvastatin, simvastatin acid, and atorvastatin. Pharm. Res. 21, 16861691 (2004). Obach, R.S., Walsky, R.L., Venkatakrishnan, K., Houston, j.B. & Tremaine, L.M. In vitro cytochrome P450 inhibition data and the prediction of drug-drug interactions: qualitative relationships, quantitative predictions, and the rankorder approach. Clin. Pharmacol. Ther. 78, 582592 (2005). Brown, H.S., Ito, K., Galetin, a. & Houston, j.B. Prediction of in vivo drug-drug interactions from in vitro data: impact of incorporating parallel pathways of drug elimination and inhibitor absorption rate constant. Br. J. Clin. Pharmacol. 60, 508518 (2005). Ito, K., Hallifax, D., Obach, R.S. & Houston, j.B. Impact of parallel pathways of drug elimination and multiple cytochrome P450 involvement on drug-drug interactions: CyP2D6 paradigm. Drug Metab. Dispos. 33, 837844 (2005).
VOLUME 89 NUMBER 1 | jaNUaRy 2011 | www.nature.com/cpt

articles
10. Robinson, K., johnston, a., Walker, S., Mulrow, j.P., McKenna, W.j. & Holt, D.W. The digoxin-amiodarone interaction. Cardiovasc. Drugs Ther. 3, 2528 (1989). 11. Ohyama, K., Nakajima, M., Suzuki, M., Shimada, N., yamazaki, H. & yokoi, T. Inhibitory effects of amiodarone and its N-deethylated metabolite on human cytochrome P450 activities: prediction of in vivo drug interactions. Br. J. Clin. Pharmacol. 49, 244253 (2000). 12. Reese, M.j., Wurm, R.M., Muir, K.T., Generaux, G.T., St john-Williams, L. & McConn, D.j. an in vitro mechanistic study to elucidate the desipramine/ bupropion clinical drug-drug interaction. Drug Metab. Dispos. 36, 11981201 (2008). 13. He, M., Kunze, K.L. & Trager, W.F. Inhibition of (S)-warfarin metabolism by sulfinpyrazone and its metabolites. Drug Metab. Dispos. 23, 659663 (1995). 14. Otton, S.V., Ball, S.E., Cheung, S.W., Inaba, T., Rudolph, R.L. & Sellers, E.M. Venlafaxine oxidation in vitro is catalysed by CyP2D6. Br. J. Clin. Pharmacol. 41, 149156 (1996). 15. jacobsen, W. et al. Lactonization is the critical first step in the disposition of the 3-hydroxy-3-methylglutaryl-Coa reductase inhibitor atorvastatin. Drug Metab. Dispos. 28, 13691378 (2000). 16. Keskitalo, j.E., Kurkinen, K.j., Neuvoneni, P.j. & Niemi, M. aBCB1 haplotypes differentially affect the pharmacokinetics of the acid and lactone forms of simvastatin and atorvastatin. Clin. Pharmacol. Ther. 84, 457461 (2008). 17. Bapiro, T.E., Egnell, a.C., Hasler, j.a. & Masimirembwa, C.M. application of higher throughput screening (HTS) inhibition assays to evaluate the interaction of antiparasitic drugs with cytochrome P450s. Drug Metab. Dispos. 29, 3035 (2001). 18. Skjelbo, E. & Brsen, K. Inhibitors of imipramine metabolism by human liver microsomes. Br. J. Clin. Pharmacol. 34, 256261 (1992). 19. von Moltke, L.L., Greenblatt, D.j., Cotreau-Bibbo, M.M., Duan, S.X., Harmatz, j.S. & Shader, R.I. Inhibition of desipramine hydroxylation in vitro by serotoninreuptake-inhibitor antidepressants, and by quinidine and ketoconazole: a model system to predict drug interactions in vivo. J. Pharmacol. Exp. Ther. 268, 12781283 (1994). 20. Prakash, C., Kamel, a., Cui, D., Whalen, R.D., Miceli, j.j. & Tweedie, D. Identification of the major human liver cytochrome P450 isoform(s) responsible for the formation of the primary metabolites of ziprasidone and prediction of possible drug interactions. Br. J. Clin. Pharmacol. 49 Suppl 1, 35S42S (2000). 21. von Moltke, L.L. et al. Triazolam biotransformation by human liver microsomes in vitro: effects of metabolic inhibitors and clinical confirmation of a predicted interaction with ketoconazole. J. Pharmacol. Exp. Ther. 276, 370379 (1996). 22. alderman, j. Coadministration of sertraline with cisapride or pimozide: an open-label, nonrandomized examination of pharmacokinetics and corrected QT intervals in healthy adult volunteers. Clin. Ther. 27, 10501063 (2005). 23. Shin, j.G., Kane, K. & Flockhart, D.a. Potent inhibition of CyP2D6 by haloperidol metabolites: stereoselective inhibition by reduced haloperidol. Br. J. Clin. Pharmacol. 51, 4552 (2001). 24. Otton, S.V., Wu, D., joffe, R.T., Cheung, S.W. & Sellers, E.M. Inhibition by fluoxetine of cytochrome P450 2D6 activity. Clin. Pharmacol. Ther. 53, 401409 (1993). 25. Kobayashi, K., yamamoto, T., Chiba, K., Tani, M., Ishizaki, T. & Kuroiwa, y. The effects of selective serotonin reuptake inhibitors and their metabolites on S-mephenytoin 4-hydroxylase activity in human liver microsomes. Br. J. Clin. Pharmacol. 40, 481485 (1995). 26. Rowland, K. et al. Inhibition of CyP2D6 activity by treatment with propranolol and the role of 4-hydroxy propranolol. Br. J. Clin. Pharmacol. 38, 914 (1994). 27. Crespi, C.L. & Penman, B.W. Use of cDNa-expressed human cytochrome P450 enzymes to study potential drug-drug interactions. Adv. Pharmacol. 43, 171188 (1997). 28. Masimirembwa, C.M., Hasler, j.a. & johansson, I. Inhibitory effects of antiparasitic drugs on cytochrome P450 2D6. Eur. J. Clin. Pharmacol. 48, 3538 (1995). 29. Wang, y. et al. Effects of imatinib (Glivec) on the pharmacokinetics of metoprolol, a CyP2D6 substrate, in Chinese patients with chronic myelogenous leukaemia. Br. J. Clin. Pharmacol. 65, 885892 (2008). 30. Lakehal, F., Wurden, C.j., Kalhorn, T.F. & Levy, R.H. Carbamazepine and oxcarbazepine decrease phenytoin metabolism through inhibition of CyP2C19. Epilepsy Res. 52, 7983 (2002). 31. Sauer, j.M. et al. atomoxetine hydrochloride: clinical drug-drug interaction prediction and outcome. J. Pharmacol. Exp. Ther. 308, 410418 (2004). 32. Chiba, K., Kobayashi, K., Manabe, K., Tani, M., Kamataki, T. & Ishizaki, T. Oxidative metabolism of omeprazole in human liver microsomes: cosegregation with S-mephenytoin 4-hydroxylation. J. Pharmacol. Exp. Ther. 266, 5259 (1993). 33. Zomorodi, K. & Houston, j.B. Diazepam-omeprazole inhibition interaction: an in vitro investigation using human liver microsomes. Br. J. Clin. Pharmacol. 42, 157162 (1996). 34. Ching, M.S. et al. Potent inhibition of yeast-expressed CyP2D6 by dihydroquinidine, quinidine, and its metabolites. Biochem. Pharmacol. 50, 833837 (1995). 35. OReilly, R.a., Trager, W.F., Rettie, a.E. & Goulart, D.a. Interaction of amiodarone with racemic warfarin and its separated enantiomorphs in humans. Clin. Pharmacol. Ther. 42, 290294 (1987). 36. Shoaf, S.E. et al. Tolvaptan administration does not affect steady state amiodarone concentrations in patients with cardiac arrhythmias. J. Cardiovasc. Pharmacol. Ther. 10, 165171 (2005). 37. Bergstrom, R.F., Peyton, a.L. & Lemberger, L. Quantification and mechanism of the fluoxetine and tricyclic antidepressant interaction. Clin. Pharmacol. Ther. 51, 239248 (1992). 38. Harvey, a.T. & Preskorn, S.H. Fluoxetine pharmacokinetics and effect on CyP2C19 in young and elderly volunteers. J. Clin. Psychopharmacol. 21, 161166 (2001). 39. Neuvonen, P.j. & jalava, K.M. Itraconazole drastically increases plasma concentrations of lovastatin and lovastatin acid. Clin. Pharmacol. Ther. 60, 5461 (1996). 40. Funck-Brentano, C., Becquemont, L., Lenevu, a., Roux, a., jaillon, P. & Beaune, P. Inhibition by omeprazole of proguanil metabolism: mechanism of the interaction in vitro and prediction of in vivo results from the in vitro experiments. J. Pharmacol. Exp. Ther. 280, 730738 (1997). 41. Stuht, H., Lode, H., Koeppe, P., Rost, K.L. & Schaberg, T. Interaction study of lomefloxacin and ciprofloxacin with omeprazole and comparative pharmacokinetics. Antimicrob. Agents Chemother. 39, 10451049 (1995). 42. Soons, P.a. et al. Influence of single- and multiple-dose omeprazole treatment on nifedipine pharmacokinetics and effects in healthy subjects. Eur. J. Clin. Pharmacol. 42, 319324 (1992). 43. Capon, D.a., Bochner, F., Kerry, N., Mikus, G., Danz, C. & Somogyi, a.a. The influence of CyP2D6 polymorphism and quinidine on the disposition and antitussive effect of dextromethorphan in humans. Clin. Pharmacol. Ther. 60, 295307 (1996). 44. Damkier, P., Hansen, L.L. & Brosen, K. Effect of diclofenac, disulfiram, itraconazole, grapefruit juice and erythromycin on the pharmacokinetics of quinidine. Br. J. Clin. Pharmacol. 48, 829838 (1999). 45. Kurtz, D.L., Bergstrom, R.F., Goldberg, M.j. & Cerimele, B.j. The effect of sertraline on the pharmacokinetics of desipramine and imipramine. Clin. Pharmacol. Ther. 62, 145156 (1997). 46. alderman, j. et al. Desipramine pharmacokinetics when coadministered with paroxetine or sertraline in extensive metabolizers. J. Clin. Psychopharmacol. 17, 284291 (1997). 47. OReilly, R.a. Stereoselective interaction of sulfinpyrazone with racemic warfarin and its separated enantiomorphs in man. Circulation 65, 202207 (1982). 48. Rosenkranz, B., Fischer, C., jakobsen, P., Kirstein Pedersen, a. & Frlich, j.C. Plasma levels of sulfinpyrazone and of two of its metabolites after a single dose and during the steady state. Eur. J. Clin. Pharmacol. 24, 231235 (1983). 49. albers, L.j. et al. Effect of venlafaxine on imipramine metabolism. Psychiatry Res. 96, 235243 (2000). 50. Lindh, j.D., annas, a., Meurling, L., Dahl, M.L. & aL-Shurbaji, a. Effect of ketoconazole on venlafaxine plasma concentrations in extensive and poor metabolisers of debrisoquine. Eur. J. Clin. Pharmacol. 59, 401406 (2003).

CliniCal pharmaCology & TherapeuTiCs | VOLUME 89 NUMBER 1 | jaNUaRy 2011

113

Vous aimerez peut-être aussi